cancers

Review
Deep Learning Techniques to Diagnose Lung Cancer
Lulu Wang

Biomedical Device Innovation Center, Shenzhen Technology University, Shenzhen 518118, China;
[email protected] or [email protected]

Simple Summary: This study investigates the latest achievements, challenges, and future research
directions of deep learning techniques for lung cancer and pulmonary nodule detection. Hopefully,
these research findings will help scientists, investigators, and clinicians develop new and effective
medical imaging tools to improve lung nodule diagnosis accuracy, sensitivity, and specificity.

Abstract: Medical imaging tools are essential in early-stage lung cancer diagnostics and the mon-
itoring of lung cancer during treatment. Various medical imaging modalities, such as chest X-ray,
magnetic resonance imaging, positron emission tomography, computed tomography, and molecular
imaging techniques, have been extensively studied for lung cancer detection. These techniques
have some limitations, including not classifying cancer images automatically, which is unsuitable
for patients with other pathologies. It is urgently necessary to develop a sensitive and accurate
approach to the early diagnosis of lung cancer. Deep learning is one of the fastest-growing topics in
medical imaging, with rapidly emerging applications spanning medical image-based and textural
data modalities. With the help of deep learning-based medical imaging tools, clinicians can detect
Citation: Wang, L. Deep Learning
Techniques to Diagnose Lung Cancer.
and classify lung nodules more accurately and quickly. This paper presents the recent development
Cancers 2022, 14, 5569. https:// of deep learning-based imaging techniques for early lung cancer detection.
doi.org/10.3390/cancers14225569
Keywords: lung cancer; medical images; segmentation; classification; deep learning; convolutional
Academic Editor: Andreas Stadlbauer
neural network
Received: 21 October 2022
Accepted: 11 November 2022
Published: 13 November 2022
1. Introduction
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
Lung cancer is the most frequent cancer and the cause of cancer death, with the
published maps and institutional affil- highest morbidity and mortality in the United States [1]. In 2018, GLOBOCAN estimated
iations. approximately 2.09 million new cases and 1.76 million lung cancer-related deaths [2]. Lung
cancer cases and deaths have increased significantly globally [2]. Approximately 85–88%
Correction Statement: This article of lung cancer cases are non-small cell lung carcinoma (NSCLS), and about 12–15% of
has been republished with a minor lung cancer cases are small cell lung cancer (SCLC) [3]. Early lung cancer diagnosis and
change. The change does not affect
intervention are crucial to increase the overall 5-year survival rate due to the invasiveness
the scientific content of the article and
and heterogeneity of lung cancer [4].
further details are available within the
Over the past two decades, various medical imaging techniques, such as chest X-ray,
backmatter of the website version of
positron emission tomography (PET), magnetic resonance imaging (MRI), computed to-
this article.
mography (CT), low-dose CT (LDCT), and chest radiograph (CRG), have been extensively
investigated for lung nodule detection. Although CT is the golden standard imaging tool
for lung nodule detection, it can only detect apparent lung cancer with high false-positive
Copyright: © 2022 by the author. rates and produces harmful X-ray radiation [5]. LDCT has been proposed to reduce harm-
Licensee MDPI, Basel, Switzerland. ful radiation to detect lung cancer [6]. However, cancer-related deaths were concentrated
This article is an open access article in subjects undergoing LDCT. 2-deoxy-18F-fluorodeoxyglucose (18F-FDG) PET has been
distributed under the terms and developed to improve the detection performance of lung cancer [7]. 18F-FDG PET produces
conditions of the Creative Commons semi-quantitative parameters of tumor glucose metabolism, which is helpful in the diag-
Attribution (CC BY) license (https:// nosis of NSCLC [8]. However, 18F-FDG PET requires further evaluation of patients with
creativecommons.org/licenses/by/ NSCLC. Some new imaging techniques, such as magnetic induction tomography (MIT),
4.0/).

Cancers 2022, 14, 5569. https://doi.org/10.3390/cancers14225569 https://www.mdpi.com/journal/cancers

Cancers 2022, 14, 5569 2 of 24

have been developed for early-stage cancer cell detection [9]. However, this technique lacks
clinical validation of human subjects.
Many computer-aided detection (CAD) systems have been extensively studied for
lung cancer detection and classification [10,11]. Compared to trained radiologists, CAD
systems provide better lung nodules and cancer detection performance in medical images.
Generally, the CAD-based lung cancer detection system includes four steps: image process-
ing, extraction of the region of interest (ROI), feature selection, and classification. Among
these steps, feature selection and classification play the most critical roles in improving the
accuracy and sensitivity of the CAD system, which relies on image processing to capture
reliable features. However, benign and malignant nodule classification is a challenge.
Many investigators have applied deep learning techniques to help radiologists make more
accurate diagnoses [12–15]. Previous studies have confirmed that deep learning-based
CAD systems can effectively improve the efficiency and accuracy of medical diagnosis,
especially for diagnosing various common cancers, such as lung and breast cancers [16,17].
Deep learning-based CAD systems can automatically extract high-level features from orig-
inal images using different network structures than traditional CAD systems. However,
deep learning-based CAD systems have some limitations, such as low sensitivity, high
FP, and time consumption. Therefore, a rapid, cost-effective, and highly sensitive deep
learning-based CAD system for lung cancer prediction is urgently needed.
The deep learning-based lung imaging techniques research mainly includes pulmonary
nodule detection, segmentation, and classification of benign and malignant pulmonary
nodules. Researchers mainly focus on developing new network structures and loss func-
tions to improve the performance of deep learning models. Several research groups have
recently published review papers on deep learning techniques [18–20]. However, deep
learning techniques have developed rapidly, and many new methods and applications
have emerged every year. This research has appeared with content that previous studies
cannot cover.
This paper presents recent achievements in lung cancer segmentation, detection, and
classification using deep learning methods. This study highlights current state-of-the-art
deep learning-based lung cancer detection methods. This paper also highlights recent
achievements, relevant research challenges, and future research directions. The rest of the
paper is structured as follows. Section 2 describes the currently available medical lung
imaging techniques for lung cancer detection; Section 3 reviews some recently developed
deep learning-based imaging techniques; Section 4 presents lung cancer prediction using
deep learning techniques; Section 5 describes the current challenges and future research
directions of deep learning-based lung imaging methods; and Section 6 concludes this study.

2. Lung Imaging Techniques

Medical imaging tools help radiologists diagnose lung disease. Among these medical
imaging approaches, CT offers more advantages, including size, location, characterization,
and lesion growth, which could identify lung cancer and nodule information. 4D CT pro-
vides more precise targeting of the administered radiation, which significantly impacts lung
cancer management [21]. Lakshmanaprabu et al. [22] developed an automatic detection
system based on linear discriminate analysis (LDA) and an optimal deep neural network
(ODNN) to classify lung cancer in CT lung images. The LDA reduced the extracted image
features to minimize the feature dimension. The ODNN was applied and optimized by
a modified gravitational search algorithm to provide a more accurate classification result.
Compared to CT, LDCT is more sensitive to early-stage lung nodules and cancer detection
with reduced radiation. However, it does not help reduce lung cancer mortality. It is
recommended that LDCT be carried out annually for high-risk smokers aged 55 to 74 [23].
PET produces much higher sensitivity and specificity for lung nodule detection than
CT due to reactive or granulomatous nodal disease [24]. PET offers a good correlation with
longer progression times and overall survival rates. 18F-FDG PET has been applied to
diagnose solitary pulmonary nodules [25]. 18F-FDG PET is a crucial in-patient selection
 Cancers 2022, 14, x FOR PEER REVIEW 3 of
Cancers 2022, 14, 5569 3 of 24

with longer progression times and overall survival rates. 18F-FDG PET has been applie
and advanced to diagnose
NSCLC forsolitary
radicalpulmonary
radiotherapy. nodules [25]. 18F-FDG
PET-assisted PET is a crucial
radiotherapy offersin-patient
more selectio
accuracy [26] andand manages
advancedabout NSCLC 32% forofradical radiotherapy.
patients with stage PET-assisted
IIIA lung cancer radiotherapy
[27]. 18F- offers mo
FDG PET provides accuracy [26] and response
a significant manages assessment
about 32% of inpatients
patientswith withstage
NSCLC IIIAundergoing
lung cancer [27]. 18
FDG PET provides a significant response assessment in patients with NSCLC undergoin
induction chemotherapy.
MRI is theinduction
most potent chemotherapy.
lung imaging tool without ionizing radiation, but it provides
MRI with
insufficient information is thehigh
mostcosts
potent and lung imaging tool without
time-consuming ionizing
limitations. radiation,
It failed but it provid
to detect
insufficient information with high costs and time-consuming
about 10% of small lung nodules (4–8 mm in diameter) [28]. MRI with ultra-short echo limitations. It failed to dete
about 10% of small lung nodules (4–8 mm in diameter)
time (UTE) can improve signal intensity and reduce lung susceptibility artifacts. MRI with [28]. MRI with ultra-short ech
time (UTE) can improve signal intensity and reduce
UTE is sensitive for detecting small lung nodules (4–8 mm) [29]. MRI achieves a higherlung susceptibility artifacts. MRI wi
UTE is sensitive
lung nodule detection rate thanfor detecting
LDCT. MRI small lung nodules
with different pulse(4–8 mm) [29].
sequences alsoMRI achieves a high
improved
lung nodule detection rate than LDCT. MRI with different
lung nodule detection sensitivity. The authors investigated T1-weighted and T2-weighted pulse sequences also improve
lung nodule detection sensitivity. The authors investigated
MRI to detect small lung nodules [30,31]. Compared to 3T 1.5 MRI, 1.5T MRI is much easier T1-weighted and T2-weighte
MRI to detect small lung nodules [30,31]. Compared to 3T 1.5 MRI, 1.5T MRI is muc
to identify ground glass opacities [32]. Ground glass opacities were successfully detected
easier to identify ground glass opacities [32]. Ground glass opacities were successful
in 75% of subjects with lung fibrosis who received 1.5T MRI with SSFP sequences [33]. MRI
detected in 75% of subjects with lung fibrosis who received 1.5T MRI with SSFP sequenc
with T2-weighted fast spin echo provides similar or even better performance for ground
[33]. MRI with T2-weighted fast spin echo provides similar or even better performance f
glass infiltrate detection in immunocompromised subjects [34].
ground glass infiltrate detection in immunocompromised subjects [34].
Several research groups have recently investigated the feasibility of using MIT for lung
Several research groups have recently investigated the feasibility of using MIT f
disease detection [35,36]. However, due to the lack of measurement systems, expensive
lung disease detection [35,36]. However, due to the lack of measurement systems, expe
computational siveelectromagnetic models, low image resolution, and some other challenges,
computational electromagnetic models, low image resolution, and some other cha
MIT technology still has a long
lenges, MIT technology way to gohas
still before
a longit can
waybe to widely
go before used as be
it can a commercial
widely used as a com
imaging tool inmercial
clinicalimaging
conditions.
tool in clinical conditions.
Medical imaging approaches
Medical imaging play an essential
approaches play anstrategy
essentialin early-stage
strategy lung cancer
in early-stage lung cancer d
detection and tection
improve and improve the survival rate. However, these techniques have limita-
the survival rate. However, these techniques have some some limitation
tions, includingincluding
high false positives,
high and cannot
false positives, and detect
cannotlesions automatically.
detect lesions SeveralSeveral
automatically. CAD CAD sy
systems have been developed for lung cancer detection [37,38]. As shown
tems have been developed for lung cancer detection [37,38]. As shown in Figure in Figure 1, a 1, a CAD
CAD-based lung nodule detection system [14] usually consists of three main
based lung nodule detection system [14] usually consists of three main phases: data co phases: data
collection and pre-processing, training, and
lection and pre-processing, testing.
training, andThere areThere
testing. two types
are two oftypes
CADofsystems:
CAD systems: th
the detection system identifies specific anomalies according to interest
detection system identifies specific anomalies according to interest regions, regions, and the anddi- the dia
agnostic systemnostic
analyses
systemlesion information,
analyses such as type,
lesion information, suchseverity,
as type,stage, andstage,
severity, progression.
and progression

Figurelung
Figure 1. CAD-based 1. CAD-based lung cancer
cancer detection detection
system system
[14]. The [14].
figure The figure
is reused fromis reference
reused from reference
[14]; no [14]; n
special
special permission permission
is required is required
to reuse to reuse
all or part all or part
of articles of articles
published by published by MDPI,
MDPI, including including figur
figures
and tables. For articles published under an open-access Creative Common CC BY license.
and tables. For articles published under an open-access Creative Common CC BY license.
3. Deep Learning-Based
3. Deep Learning-Based Imaging Techniques
Imaging Techniques
A deep learning-based CAD system has been reported as a promising tool for the
automatic diagnosis of lung disease in medical imaging with significant accuracy [34–36].
The deep learning model is a neural network model with multiple levels of data represen-
Cancers 2022, 14, 5569 4 of 24

tation. The deep learning approaches can be grouped into unsupervised, reinforcement,
and supervised learning.
Unsupervised learning does not require user guidance, which analyzes the data and
then sorts inherent similarities between the input data. Therefore, semi-supervised learning
is a mixed model that can provide a win-win situation, even with different challenges.
Semi-supervised learning techniques use both labeled and unlabeled data. With the help of
labeled and unlabeled data, the accuracy of the decision boundary becomes much higher.
Auto-Encoders (AE), Restricted Boltzmann Machines (RBM), and Generative Adversarial
Networks (GAN) are good at clustering and nonlinear dimensionality reduction. A large
amount of labeled data is usually required during training, which increases cost, time, and
difficulty. Researchers have applied deep clustering to reduce labeling and make a more
robust model [39,40].
Convolutional neural networks (CNN), deep convolutional neural networks (DCNN),
and recurrent neural networks (RNN) are the most widely used unsupervised learning
algorithms in medical images. CNN architecture is one of the most widely used supervised
deep learning approaches for lesion segmentation and classification because less pre-
processing is required. CNN architectures have recently been applied to medical images for
image segmentation (such as Mask R-CNN [41]) and classification (such as AlexNet [42] and
VGGNet [43]). DCNN architectures usually contain more layers with complex nonlinear
relationships, which have been used for classification and regression with reasonable
accuracy [44–46]. RNN architecture is a higher-order neural network that can accommodate
the network output to re-input [47]. RNN applies the Elman network with feedback links
from the hidden layer to the input layer, which has the potential to capture and exploit
cross-slice variations to incorporate volumetric patterns of nodules. However, RNN has a
vanishing gradient problem.
The reinforcement learning technique was first applied in Google Deep Mind in
2013 [48]. Since then, reinforcement learning approaches have been extensively investigated
to improve lung cancer detection accuracy, sensitivity, and specificity. Semi-supervised
learning approaches, such as deep reinforcement learning and generative adversarial
networks, use labeled datasets.
Supervised learning usually involves a learning algorithm, and labels are assigned
to the input data according to the labeling data during training. Various supervised
deep learning approaches have been applied to CT images to identify abnormalities with
anatomical localization. These approaches have some drawbacks, such as the large amount
of labeled data required during training, the assumption of fixed network weights upon
training completion, and the inability to be improved after training. Thus, a few-shot
learning (FSL) model is developed to reduce data requirements during training.

4. Lung Cancer Prediction Using Deep Learning

This section presents recent achievements in lung cancer and nodule prediction using
deep learning techniques. The processing includes image pre-processing, lung nodule
segmentation, detection, and classification.

4.1. Imaging Pre-Processing Techniques and Evaluation

4.1.1. Pre-Processing Techniques
The pre-processed images are injected into a deep learning algorithm with specific
architecture and training and tested on the image datasets. The image noise affects the pre-
cision of the final classifier. Several noise reduction approaches, such as median filter [48],
Wiener filter [49], and non-local means filter [50], have been developed for pre-processing
to improve accuracy and generalization performance. After denoising, a normalization
method, such as min-max normalization, is required to rescale the images and reduce the
complexity of image datasets.
Cancers 2022, 14, 5569 5 of 24

4.1.2. Performance Metrics

Several performance metrics have been used to evaluate the performance of deep learn-
ing algorithms, including accuracy, precision, sensitivity, specificity, F1_score, error, mean
squared error (MSE), receiver operation characteristic (ROC) curve, over-segmentation
rate (OR), under-segmentation rate (UR), Dice similarity coefficient (DSC), Jaccard Score
(JS), average symmetric surface distance (ASD), modified Hausdorff distance (MHD), and
intersection over union (IoU).
Accuracy assesses the capability concerning the results with the existing information
features. Sensitivity is helpful for evaluation when FN is high. Precision is an effective
measurement index when FP is high. The F1_score is applied when the class distribution is
uneven. ROC can tune detection sensitivity. The area under the receiver operating charac-
teristic curve (AUC) has been used to evaluate the proposed deep learning model. Larger
values of accuracy, precision, sensitivity, specificity, AUC, DSC, and JS, and smaller values
of Error, UR, OR, and MHD indicate better performance of a deep learning-based algorithm.
These performance metrics can be computed using the following equations [51,52]:

TP + TN
Accuracy = (1)
TP + TN + FP + FN
TP
Sensitivity = (2)
TP + FN
TN
Specificity = (3)
TN + FP
TP
Precision = (4)
TP + FP
2TP
F1_score = (5)
2TP + FP + FN
FP + FN
Error = (6)
TP + TN + FP + FN
2TP
DSC = (7)
2TP + FP + FN
DSC
JS = (8)
2 − DSC
1
MHD(A, B) =
Na ∑ min||a − b|| (9)
a∈ A b∈ B

TP
IoU = (10)
TP + FP + TN
where TP (true positive) denotes the number of correct positives; TN (true negative)
indicates the number of correct negatives; FP (false positive) means the number of incorrect
positives; FN (false negative) denotes the number of incorrect negatives; B is the target
object region, A denotes ground truth dataset, and Na is the number of pixels in A; IoU
refers to the percentage of the intersection to the union of the ground truth and predicted
areas and is a metric for various object detection and semantic segmentation problems.

4.2. Datasets
Lung image datasets play an essential role in evaluating the performance of deep
learning-based algorithms for lung nodule classification and detection. Table 1 shows
publicly available lung images and clinical datasets for assessing nodule classification and
detection performance.
Cancers 2022, 14, 5569 6 of 24

Table 1. Lung image dataset.

Reference Dataset Sample Number

[53] Lung image database consortium (LIDC) 399 CT images
Lung image database consortium and image database
[54] 1018 CT images from 1010 patients
resource initiative (LIDC-IDRI)
[55] Lung nodule analysis challenge 2016 (LUNA16) 888 CT images from LIDC-IDRI dataset
50 LDCT lung images &
[56] Early lung cancer action program (ELCAP)
379 unduplicated lung nodule CT images
294 CT images from Centro Hospitalar e
[57] Lung Nodule Database (LNDb)
Universitario de São Joãao
[58] Indian Lung CT Image Database (ILCID) CT images from 400 patients
[59] Japanese Society of Radiological Technology (JSRT) 154 nodules & 93 nonnodules with labels
Nederland-Leuvens Longkanker Screenings Onderzoek
[60] CT images from 15,523 human subjects
(NELSON)
[61] Automatic nodule detection 2009 (ANODE09) 5 examples & 50 test images
[62] Shanghai Zhongshan hospital database CT images from 350 patients
Society of Photo-Optical Instrumentation Engineers
in conjunction with the American Association of Physicists
[63] 60 thoracic CT scans with 73 nodules
in Medicine and the National Cancer Institute
(SPIE-AAPM-NCI) LungX
General Hospital of Guangzhou military command
[64] 180 benign & 120 malignant lung nodules
(GHGMC) dataset
First Affiliated Hospital of Guangzhou Medical University
[65] 142 T2-weighted MR images
(FAHGMU) dataset
[66] Non-small cell lung cancer (NSCLC)-Radiomics database 13,482 CT images from 89 patients
[67] Danish lung nodule screening trial (DLCST) CT images from 4104 subjects
CT images from 1058 patients with lung cancer &
[68] U.S. National Lung Screening Trial (NLST)
9310 patients with benign lung nodules

4.3. Lung Image Segmentation

Image segmentation aims to recognize the voxel information and external contour
of the region of interest. In medical imaging, segmentation is mainly used to segment
organs or lesions to quantitatively analyze relevant clinical parameters and provide further
guidance for follow-up diagnosis and treatment. For example, target delineation is crucial
for surgical image navigation and tumor radiotherapy guidance.
Lung segmentation plays a crucial role in medical images for lesion detection, in-
cluding thorax extraction (removes artifacts) and lung extraction (identifies the left and
right lungs). Several threshold techniques, such as the threshold method [69], iterative
threshold [70], Otsu threshold [71], and adaptive threshold [72,73], have been investigated
for lung segmentation. Few research groups have investigated segmentation methods
based on region and 3D region growth [74,75]. Kass et al. [76] first introduced the active
contour model, and Lan et al. [77] applied the active contour model for lung segmentation.
These techniques are manual segmentation and have many disadvantages, such as being
relatively slow, prone to human error, scarcity of ground truth, and class imbalance.
Several deep learning approaches have been investigated for lung segmentation.
Wang et al. [78] developed a multi-view CNN (MV-CNN) for lung nodule segmentation,
with an average DSC of 77.67% and an average ASD of 0.24 for the LIDC-IDRI dataset.
Unlike conventional CNN, MV-CNN integrates multiple input images for lung nodule
identification. However, it is difficult for MV-CNN to process 3D CT scans. Thus, a
3D CNN was developed to process volumetric patterns of cancerous nodules [79]. Sun
Cancers 2022, 14, 5569 7 of 24

et al. [80] designed a two-stage CAD system to segment lung nodules and FP reduction
automatically. The first stage aims to identify and segment the nodules, and the second
stage aims to reduce FP. The system was tested on the LIDC-IDRI dataset and evaluated by
four experienced radiologists. The system obtained an average F1_score of 0.8501 for lung
nodule segmentation.
In 2020, Cao et al. [81] developed a dual-branch residual network (DB-ResNet) that
simultaneously captures the multi-view and multi-scale features of nodules. The pro-
posed DB-ResNet was evaluated on the LIDC-IDRI dataset and achieved a DSC of 82.74%.
Compared to trained radiologists, DB-ResNet provides a higher DSC.
In 2021, Banu et al. [82] proposed an attention-aware weight excitation U-Net (AWEU-
Net) architecture in CT images for lung nodule segmentation. The architecture contains
two stages: lung nodule detection based on fine-tuned Faster R-CNN and lung nodule seg-
mentation based on the U-Net with position attention-aware weight excitation (PAWE) and
channel attention-aware weight excitation (CAWE). The AWEU-Net obtained DSC of 89.79%
and 90.35%, IoU of 82.34%, and 83.21% for the LUNA16 and LIDC-IDRI datasets, respectively.
Dutta [83] developed a dense recurrent residual CNN (Dense R2Unet) based on the
U-Net and dense interconnections. The proposed method was tested on a lung segmen-
tation dataset, and the results showed that the Dense R2UNet offers better segmentation
performance than U-Net and ResUNet.
Table 2 shows the recently developed lung nodule segmentation techniques. Among
these approaches, SVM systems obtained an accuracy range of 92.6–98.1%, CNN-based
systems obtained a specificity range of 77.67–91%, ResNet models obtained a DSC range of
82.74–98.1%, and U-Net segmentation systems achieved an accuracy range of 82.2–99.27%,
precision range of 46.61–98.2%, recall range of 21.43–96.33%, and F1_score range of 24.64–
99.1%, respectively. The DenseNet201 system obtained an accuracy of 97%, a sensitivity of
96.2%, a specificity of 97.5%, an AUC of 0.968, and an F1_score of 96.1%. Several segmenta-
tion methods, including SVM, Dense R2UNet, 3D Attention U-Net, Dense R2UNet, Res
BCDU-Net, U-Net FSL, U-Net CT, U-Net PET, U-Net PET/CT, CNN, and DenseNet201,
achieved high accuracy results (over 94%).

Table 2. Lung nodule segmentation approaches.

Reference Year Method Imaging Datasets Results

Shiraz University of
[84] 2013 Support vector machine (SVM) CT images Accuracy: 98.1%
Medical Sciences
Lung nodule
[85] 2014 CT images 85 patients Accuracy: >90%
segmentation
Accuracy: 94.67% for
benign tumors;
[86] 2015 SVM CT images 193 CT images
Accuracy: 96.07% for
adhesion tumor
Bidirectional chain coding
[87] 2015 CT images LIDC Accuracy: 92.6%
combined with SVM
Convolutional networks
[88] 2015 CT images 82 patients DSC: 68% ± 10%
(ConvNets)
Multi-view convolutional neural
[77] 2017 CT images LIDC-IDRI DSC: 77.67%
networks (MV-CNN)
[80] 2017 Two-stage CAD CT images LIDC-IDRI F1-score: 85.01%
3D Slicer chest imaging platform
[89] 2017 CT images LIDC median DSC: 99%
(CIP)
Deep computer aided detection
[90] 2017 CT images LIDC-IDRI Sensitivity: 88%
(CAD)
[91] 2018 3D deep multi-task CNN CT images LUNA16 DSC: 91%
[92] 2018 Improved U-Net CT images LUNA16 DSC: 73.6%
Cancers 2022, 14, 5569 8 of 24

Table 2. Cont.

Reference Year Method Imaging Datasets Results

TCIA DSC: 74% ± 0.13
Incremental-multiple resolution
[93] 2018 residually connected network CT images MSKCC DSC: 75%±0.12
(MRRN)
LIDC DSC: 68%±0.23
712 lung cancer
hematoxylin-eosin- patients operated in
[94] 2018 U-Net Precision: 80%
stained slides Uppsala Hospital,
Stanford TMA cores
[95] 2019 Mask R-CNN CT images LIDC-IDRI Average precision:78%
[96] 2020 3D-UNet CT images LUNA16 DSC: 95.30%
Dual-branch Residual Network
[81] 2020 CT images LIDC-IDRI DSC: 82.74%
(DB-ResNet)
End-to-end 1916 lung tumors in
[97] 2021 CT images Sensitivity: 93.2%
deep learning 1504 patients
Fifth Medical Center
COVID-19
[98] 2021 3D Attention U-Net of the PLA General Accuracy: 94.43%
CT images
Hospital
[99] 2021 Improved U-Net CT images LIDC-IDRI Precision: 84.91%
Attention-aware weight excitation LUNA16 DSC: 89.79%
[82] 2021 CT images
U-Net (AWEU-Net)
LIDC-IDRI DSC: 90.35%
Dense Recurrent Residual
[83] 2021 Convolutional Neural CT images LUNA Sensitivity: 99.4% ± 0.2%
Network(Dense R2U CNN)
Modified U-Net in which the
encoder is replaced with a
[100] 2021 CT images LIDC-IDRI Accuracy: 97.58%
pre-trained ResNet-34 network
(Res BCDU-Net)
Hybrid COVID-19 segmentation
COVID-19 dataset
[101] 2021 and recognition framework X-Ray images Accuracy: 99.30%
from 8 sources *
(HMB-HCF)
First Affiliated
Clinical image radionics DL Hospital of
[102] 2021 CT Images Sensitivity: 0.8763
(CIRDL) Guangzhou Medical
University
260 patients with
[103] 2021 2D & 3D hybrid CNN CT scans Median DSC: 0.73
lung cancer treated
Few-shot learning U-Net (U-Net
Accuracy: 99.27% ± 0.03
FSL)
U-Net CT Lung-PET-CT-DX Accuracy: 99.08% ± 0.05
[104] 2022 PET/CT images
TCIA
U-Net PET Accuracy: 98.78% ± 0.06
U-Net PET/CT Accuracy: 98.92% ± 0.09
CNN Accuracy: 98.89% ± 0.08
Co-learning Accuracy: 99.94% ± 0.09
Seoul St. Mary’s
[105] 2022 DenseNet201 CT images Sensitivity: 96.2%
Hospital dataset
COVID-19 dataset from 8 sources *: COVID-19 Radiography Database, Pneumonia (virus) vs. COVID-19 Dataset,
Covid-19 X-Ray images using CNN Dataset, COVID-19 X-ray Images5 Dataset, COVID-19 Patients Lungs X-Ray
Images 10,000 Dataset, COVID-19 Chest X-Ray Dataset, COVID-19 Dataset, Curated Chest X-Ray Image Dataset
for COVID-19.

4.4. Lung Nodule Detection

Lung nodule detection is challenging because its shape, texture, and size vary greatly,
and some non-nodules, such as blood vessels and fibrosis, have a similar appearance to
lung nodules that often appear in the lungs. The processing includes two main steps:
lung nodule detection and false-positive nodule reduction. Over the past few decades,
Cancers 2022, 14, 5569 9 of 24

researchers worldwide have extensively investigated machine learning and deep learning-
based approaches for lung nodule detection. Chang et al. [106] applied the support vector
machine (SVM) for nodules classification in ultrasound images. Nithila et al. [107] de-
veloped a lung nodule detection model based on heuristic search and particle clustering
algorithms for network optimization. In 2005, Zhang et al. [108] developed a discrete-time
cellular neural network (DTCNN) to detect small (2–10 mm) juxtapleural and non-pleural
nodules in CT images. The method obtained a sensitivity of 81.25% at 8.29 FPs per scan for
juxtapleural nodule detection and a sensitivity of 83.9% at 3.47 FPs per scan for non-pleural
nodule detection.
Hwang et al. [109] investigated the relationship between CT and commercial CAD to
detect lung nodules. They also studied LDCT images with three reconstruction kernels (B,
C, and L) from 36 human subjects. The sensitivities of 82%, 88%, and 82% for the nodules of
B, C, and L were obtained for all images. Experimental results showed that CAD sensitivity
could be elevated by combining data from 2 different kernels without radiation exposure.
Young et al. [110] studied the effects on the performance of a CAD-based nodule detection
model by reducing the CT dose. The CAD system was evaluated on the NLST dataset and
obtained sensitivities of 35%, 20%, and 42.5% at the initial dose, 50% dose, and 25% dose,
respectively. Tajbakhsh et al. [111] studied massive training ANN (MTANN) and CNN
for lung nodule detection and classification. MTANN and CNN obtained AUCs of 0.8806
and 0.7755, respectively. MTANN performs better than CNN for lung nodule detection
and classification.
Liu et al. [112] developed a cascade CNN for lung nodule detection. The transfer
learning model was applied to train the network to detect nodules, and a non-nodule
filter was introduced to the detection network to reduce false positives (FP). The proposed
architecture effectively reduces FP in the lung nodule detection system. Li et al. [65]
developed a lung nodule detection method based on a faster R-CNN network and an FP
reduction model in thoracic MR images. In this study, a faster R-CNN was developed to
detect lung nodules, and an FP reduction model was developed to reduce FP. The method
was tested on the FAHGMU dataset and obtained a sensitivity of 85.2%, with 3.47 FP
per scan. Cao et al. [113] developed a two-stage CNN (TSCNN) model for lung nodule
detection. In the first stage, a U-Net based on ResDense was applied to detect lung nodules.
A 3D CNN-based ensemble learning architecture was proposed in the second stage to
reduce false-positive nodules. The proposed model was compared with three existing
models, including 3DDP-DenseNet, 3DDP-SeResNet, and 3DMBInceptionNet.
Several 3D CNN models have been developed for lung nodule detection [114–116].
Perez et al. [117] developed a 3D CNN to automatically detect lung cancer and tested the
model on the LIDC-IDRI dataset. The experimental results showed that the proposed
method provides a recall of 99.6% and an AUC of 0.913. Vipparla et al. [118] proposed a
multi-patched 3D CNN with a hybrid fusion architecture for lung nodule detection with
reduced FP. The method was tested on the LUNA16 dataset and achieved a competition
performance metric (CPM) of 0.931. Dutande et al. [119] developed a 2D–3D cascaded CNN
architecture and compared it with existing lung nodule detection and segmentation meth-
ods. The results showed that the 2D–3D cascaded CNN architecture obtained a DCM of 0.80
for nodule segmentation and a sensitivity of 90.01% for nodule detection. Luo et al. [120]
developed a 3D sphere representation-based center-point matching detection network
(SCPM-Net) consisting of sphere representation and center-point matching components.
The SCPM-Net was tested on the LUNA16 dataset and achieved an average sensitivity of
89.2% at 7 FPs per image for lung nodule detection. Franck et al. [121] investigated the
effects on the performance of deep learning image reconstruction (DLIR) techniques on
lung nodule detection in chest CT images. In this study, up to 6 artificial nodules were
located within the lung phantom. Images were generated using 50% ASIR-V and DLIR
with low (DL-L), medium (DL-M), and high (DL-H) strengths. No statistically significant
difference was obtained between these methods (p = 0.987, average AUC: 0.555, 0.561, 0.557,
and 0.558 for ASIR-V, DL-L, DL-M, and DL-H).
Cancers 2022, 14, 5569 10 of 24

Table 3 shows recently developed lung nodule detection approaches using deep
learning techniques. Among these approaches, the co-learning feature fusion CNN obtained
the best accuracy of 99.29%, which is higher than other lung nodule detection approaches.
Several networks, including 3D Faster R-CNN with U-Net-like encoder, YOLOv2, YOLOv3,
VGG-16, DTCNN-ELM, U-Net++, MIXCAPS, and ProCAN, obtained good accuracy (>90%)
of lung nodule detection.

Table 3. Lung nodule detection approaches.

Reference Year Method Imaging Datasets Results

Sensitivity: >87% at
[122] 2016 3D CNN CT images LUNA16
4 FPs/scan
Sensitivity: 85.4% at
2D multi-view convolutional networks
[123] 2016 CT images LIDC-IDRI 1 FPs/scan, 90.1% at
(ConvNets)
4 FPs/scan
[124] 2016 Thresholding method CT images JSRT Accuracy: 96%
Mean sensitivity:
[110] 2017 Computer aided detection (CAD) LDCT NLST
74.1%
[125] 2017 3D CNN LDCT KDSB17 Accuracy: 87.5%
LUNA16 Accuracy: 81.41%;
[126] 2017 3D Faster R-CNN with U-Net-like encoder CT scans
LIDC-IDRI Accuracy: 90.44%
[127] 2018 Single-view 2D CNN CT scans LUNA16 metric score: 92.2%
[128] 2018 DetectNet CT scans LIDC Sensitivity: 89%
[129] 2018 3D CNN CT scans KDSB17 Sensitivity: 87%;

Novel pulmonary nodule detection LUNA16 CPM score: 94.7%

[130] 2018 CT scans
algorithm (NODULe) based on 3D CNN LIDC-IDRI Sensitivity: 94.9%
PET images 50 lung cancer patients, Sensitivity: 95.9%
[131] 2018 Deep neural networks (DNN) & 50 patients without
ultralow dose PET lung cancer Sensitivity: 91.5%
FissureNet AUC: 0.98
[132] 2018 U-Net 3DCT COPDGene AUC: 0.963
Hessian AUC: 0.158
DFCN-based cosegmentation CT scans Score: 0.865 ± 0.034;
[133] 2018 60 NSCLC patients
(DFCN-CoSeg)
PET images Score: 0.853 ± 0.063;
LUNA16,
V1 dataset includes CPM: 0.908 for the V1
Multi-scale Gradual Integration CNN
[134] 2018 CT scans 551,065 subjects; dataset, 0.942 for the
(MGI-CNN)
V2 dataset includes V2 dataset;
754,975 subjects
Deep fully CNN (DFCNet) Accuracy: 84.58%
[135] 2018 CT scans LIDC-IDR
CNN Accuracy: 77.6%
Deep learning–based automatic detection Seoul National
[136] 2018 CT scans Sensitivity: 69.9%
algorithm (DLAD) University Hospital
SVM classifier coupled with a least
[137] 2018 absolute shrinkage and selection operator CT scans LIDC-IDRI Accuracy: 84.6%
(SVM-LASSO)
Sensitivity: 88% at
[138] 2019 CNN CT scans LIDC-IDR 1.9 FPs/scan; 94.01%
at 4.01 FPs/scan
LUNA16 and Kaggle Average metric:
[139] 2019 3D CNN LDCT
datasets 92.1%
3500 CXRs contain lung
Deep learning model (DLM) based on Chest radiographs
[140] 2019 nodules & 13,711 Sensitivity: 76.8%
DCNN (CXRs)
normal CXRs
Cancers 2022, 14, 5569 11 of 24

Table 3. Cont.

Reference Year Method Imaging Datasets Results

Sensitivity of 79.6%
Nagasaki University with sizes ≤0.6 mm;
[141] 2019 Two-Step Deep Learning CT scans
Hospital Sensitivity of 75.5%
with sizes ≤0.7 mm;
Faster R-CNN network and false positive
[142] 2019 CT scans FAHGMU Sensitivity: 85.2%
(FP)
YOLOv2 with Asymmetric Convolution
[143] 2019 CT scans LIDC-IDRI Sensitivity: 94.25%
Kernel
[144] 2019 VGG-16 network CT scans LIDC-IDRI Accuracy: 92.72%
[145] 2019 Noisy U-Net (NU-Net) CT scans LUNA16 Sensitivity: 97.1%
CAD using a multi-scale dot
[146] 2019 CT scans LIDC Sensitivity: 87.81%
nodule-enhancement filter
[147] 2019 Co-Learning Feature Fusion CNN PET-CT scans 50 NSCLC patients Accuracy: 99.29%
Convolution networks with attention
[148] 2019 Chest radiographs 430,000 CXRs Sensitivity: 78%
feedback (CONAF)
Recurrent attention model with
[148] 2019 Chest radiographs 430,000 CXRs Sensitivity: 74%
annotation feedback (RAMAF)
[113] 2020 Two-Stage CNN (TSCNN) CT scans LUNA16 & LIDC-IDRI CPM: 0.911
Deep Transfer CNN and Extreme LIDC-IDRI &
[149] 2020 CT scans Sensitivity: 93.69%;
Learning Machine (DTCNN-ELM) FAH-GMU
Sensitivity: 94.2% at
[150] 2020 U-Net++ CT scans LIDC-IDRI 1 FP/scan, 96% at
2 FPs/scan
[151] 2020 MSCS-DeepLN CT scans LIDC-IDRI & DeepLN
Accuracy:
[152] 2020 Multi-scale CNN (MCNN) CT scans LIDC-IDRI
93.7% ± 0.3
[153] 2021 Lung Cancer Prediction CNN (LCP-CNN) CT scans U.S. NLST Sensitivity: 99%;
mean sensitivity: 82%
for second-reading
150 images include 340
[154] 2021 Automatic AI-powered CAD CT scans mode, 80% for
nodules
concurrent-reading
mode
Detection accuracy:
DNA-derived phage nose (D2pNose) Pusan National >75%;
[155] 2021 CT scans
using machine learning and ANN University Classification
accuracy: >86%
Capsule network-based mixture of experts
[156] 2021 CT scans LIDC-IDRI Sensitivity: 89.5%;
(MIXCAPS)
[157] 2021 CNN with attention mechanism CT scans LUNA16 Specificity: 98.9%
AUC of 0.555, 0.561,
Deep learning image reconstruction 0.557, 0.558 for
[121] 2021 CT scans LIDC-IDRI
(DLIR) ASIR-V, DL-L, DL-M,
DL-H
[58] 2021 2D-3D cascaded CNN CT scans LIDC-IDRI Sensitivity: 90.01%
3D sphere representation-based
Average sensitivity:
[120] 2022 center-points matching detection network CT scans LUNA16
89.2%
(SCPM-Net)
[158] 2022 YOLOv3 CT scans RIDER Accuracy: 95.17%
[118] 2022 3D Attention CNN CT scans LUNA16 CPM: 0.931
Progressive Growing Channel Attentive
[159] 2022 CT scans LIDC-IDRI Accuracy: 95.28%
Non-Local (ProCAN) network
Cancers 2022, 14, 5569 12 of 24

4.5. Lung Nodule Classification

In recent years, investigators have studied various deep learning techniques to im-
prove the performance of lung nodule classification [160–173]. The sensitivity and speci-
ficity of the SIFT-based classifier and SVM in the classification of pulmonary nodules
reached 86% and 97% [160], 91.38%, and 89.56% [163], respectively. The accuracy, sensitiv-
ity, and specificity of multi-scale CNN and multi-crop CNN in lung nodule classification
were 90.63%, 92.30%, and 89.47% [164], respectively, and 87%, 77%, and 93% [170], re-
spectively. The accuracy of deep-level semantic networks and multi-scale CNN in lung
nodule classification were 84.2% [167] and 86.84% [168], respectively. The CAD system
developed by Cheng et al. [169] achieved the best accuracy of 95.6%, sensitivity of 92.4%,
and specificity of 98.9% in the classification of pulmonary nodules.
The comparative study results showed that the sensitivity and specificity of CNN
and DBN for pulmonary nodule classification are 73.40% and 73.30%, 82.20%, and 78.70%,
respectively [165]. Another comparative study showed that the sensitivity and specificity
of CNN and ResNet in the classification of nodules are 76.64% and 89.50%, 81.97%, and
89.38%, respectively [171]. The combined application of CNN and RNN achieved accu-
racy, sensitivity, and specificity of 94.78%, 94.66%, and 95.14%, respectively, in classifying
pulmonary nodules [172].
In 2019, Zhang et al. [174] used an ensemble learner of multiple deep CNN in CT
images and obtained a classification accuracy of 84% for the LIDC-IDRI dataset. The
proposed classifier achieved better performance than other algorithms, such as SVM, multi-
layer perceptron, and random forests.
Sahu et al. [175] proposed a lightweight multi-section CNN with a classification accu-
racy of 93.18% for the LIDC-IDRI dataset to improve accuracy. The proposed architecture
could be applied to select the representative cross sections determining malignancy that
facilitate the interpretation of the results.
Ali et al. [176] developed a system based on transferable texture CNN that consists
of nine layers to extract features automatically and classify lung nodules. The proposed
method achieved an accuracy of 96.69% ± 0.72%, with an error of 3.30% ± 0.72% and a
recall of 97.19% ± 0.57%, respectively.
Marques et al. [177] developed a multi-task CNN to classify malignancy nodules with
an AUC of 0.783. Thamilarasi et al. [178] proposed an automatic lung nodule classifier
based on CNN with an accuracy of 86.67% for the JSRT dataset. Kawathekar et al. [179]
developed a lung nodule classifier using a machine-learning technique with an accuracy of
94% and an F1_score of 92% for the LNDb dataset.
More recently, Radford et al. [180] proposed deep convolution GAN (DCGAN),
Chuquicusma et al. [181] applied DCGAN to generate realistic lung nodules, and
Zhao et al. [182] applied Forward and Backward GAN (F&BGAN) to classify lung nodules.
The F&BGAN was evaluated on the LIDC-IDRI dataset and obtained the best accuracy of
95.24%, a sensitivity of 98.67%, a specificity of 92.47%, and an AUC of 0.98.
Table 4 shows the recently developed traditional and deep learning-based tech-
niques for classifying lung nodules. Among these methods, CNN variants obtained an
accuracy range of 83.4–99.6%, a specificity range of 73.3–95.17%, a sensitivity range of
73.3–96.85%, and an AUC range of 0.7755–0.9936, respectively. Several methods achieved
high classification accuracy (>95%), including F&BGAN, Inception_ResNet_V2, ResNet152V2,
ResNet152V2+GRU, CSO-CADLCC, ProCAN, Net121, ResNet50, DITNN, and optimal
DBN with an opposition-based pity beetle algorithm. DCNN systems obtained a sensitivity
of 89.3% [183] and an accuracy of 97.3% [184]. The classifier was developed based on the
VGG19 and CNN models and achieved accuracy, sensitivity, specificity, recall, F1_score,
AUC, and MCC above 98%.
Cancers 2022, 14, 5569 13 of 24

Table 4. Lung nodule classification approaches.

Reference Year Method Imaging Datasets Results

[185] 2014 FF-BPNN CT scans LIDC Sensitivity: 91.4%
[168] 2015 Multi-scale CNN CT scans LIDC-IDRI Accuracy: 86.84%
[166] 2015 CAD using deep features CT scans LIDC-IDRI Sensitivity: 83.35%
[165] 2015 Deep belief network (DBN) CT scans LIDC Sensitivity: 73.4%
[165] 2015 CNN CT scans LIDC Sensitivity:73.3%
[165] 2015 Fractal CT scans LIDC Sensitivity:50.2%
Scale-invariant feature transform
[165] 2015 CT scans LIDC Sensitivity: 75.6%
(SIFT)
[186] 2016 Intensity features +SVM CT scans DLCST Accuracy: 27.0%
[186] 2016 Unsupervised features+SVM CT scans DLCST Accuracy: 39.9%
[186] 2016 ConvNets 1 scale CT scans DLCST Accuracy: 84.4%
[186] 2016 ConvNets 2 scale CT scans DLCST Accuracy: 85.6%
[186] 2016 ConvNets 3 scale CT scans DLCST Accuracy: 85.6%
[171] 2017 Multi-crop CNN CT scans LIDC-IDRI Accuracy: 87.14%
[171] 2017 Deep 3D DPN CT scans LIDC-IDRI Accuracy: 88.74%
[171] 2017 Deep 3D DPN+ GBM CT scans LIDC-IDRI Accuracy: 90.44%
[111] 2017 Massive-training ANN (MTANN) CT scans LDCT AUC: 0. 8806
[111] 2017 CNN CT scans LDCT AUC: 0.7755
Japanese Society
Wavelet Recurrent Neural
[187] 2017 Chest X-Ray Radiology and Sensitivity: 88.24%
Network
Technology
Multi-crop convolutional neural
[171] 2017 CT scans LIDC-IDRI Sensitivity: 77%
network (MC-CNN)
Topology-based phylogenetic
[188] 2018 diversity index classification CT scans LIDC Sensitivity: 90.70%
CNN
[189] 2018 Transfer learning deep 3D CNN CT scans Institution records Accuracy: 71%
Kaggle Data
[128] 2018 CNN CT scans Sensitivity: 87%
Science Bowl 2017
Feature Representation Using
[190] 2018 CT scans ELCAP Accuracy: 93.9%
Deep Autoencoder
overall classification
rates: 92.3% for
LIDC-IDRI &
[112] 2018 Multi-view multi-scale CNN CT scans LIDC-IDRI; overall
ELCAP
classification rates:
90.3% for ELCAP
448 images include
[191] 2018 Wavelet-Based CNN CT scans Accuracy: 91.9%
four categories
[192] 2018 Deep ConvNets CT scans LIDC-IDRI Accuracy: 98%
Forward and Backward GAN
[182] 2018 CT scans LIDC-IDRI Sensitivity: 98.67%
(F&BGAN)
Ensemble learner of multiple
[174] 2019 CT scans LIDC-IDRI Accuracy: 84.0%
deep CNN
[175] 2019 Lightweight Multi-Section CNN CT scans LIDC-IDRI Accuracy: 93.18%
Cancers 2022, 14, 5569 14 of 24

Table 4. Cont.

Reference Year Method Imaging Datasets Results

Deep hierarchical semantic CNN
[167] 2019 CT scans LIDC Sensitivity: 70.5%
(HSCNN)
Multi-view knowledge-based
[193] 2019 CT scans LIDC-IDRI Accuracy: 91.60%
collaborative (MV-KBC)
[167] 2019 3D CNN CT scans LIDC Sensitivity: 66.8%
46 images from
[183] 2019 DCNN CT scans interventional Sensitivity: 89.3%
cytology
LIDC-IDRI &
[194] 2019 3D MixNet CT scans Accuracy: 88.83%
LUNA16
LIDC-IDRI &
[194] 2019 3D MixNet +GBM CT scans Accuracy: 90.57%
LUNA16
LIDC-IDRI &
[194] 2019 3D CMixNet+ GBM CT scans Accuracy: 91.13
LUNA16
LIDC-IDRI &
[194] 2019 3D CMixNet+ GBM+Biomarkers CT scans Accuracy: 94.17%
LUNA16
Deep Learning with
Cancer imaging
[195] 2019 Instantaneously Trained Neural CT scans Accuracy: 98.42%
Archive (CIA)
Networks (DITNN)
[184] 2020 DCNN CT scans LIDC Accuracy: 97.3%
[196] 2020 CNN CT scans LIDC Sensitivity: 93.4%
[197] 2020 2.75D CNN CT scans LUNA16 AUC: 0.9842
[198] 2020 Two-step Deep Network (TsDN) CT scans LIDC-IDRI Sensitivity: 88.5%
LIDC-IDRI & Accuracy: 96.69% ±
[176] 2020 Transferable texture CNN CT scans
LUNGx 0.72%
[199] 2020 Taguchi-Based CNN X-ray & CT images 245,931 images Accuracy: 99.6%
Optimal Deep Belief Network
[200] 2021 with Opposition-based Pity Beetle CT scans LIDC-IDRI Sensitivity: 96.86%
Algorithm
[177] 2021 Multi-task CNN CT scans LIDC-IDRI AUC: 0.783
[178] 2021 CNN CT scans JSRT Accuracy: 86.67%
[201] 2021 Inception_ResNet_V2 CT scans LC25000 Accuracy: 99.7%
[201] 2021 VGG19 CT scans LC25000 Accuracy: 92%
[201] 2021 ResNet50 CT scans LC25000 Accuracy: 99%
[201] 2021 Net121 CT scans LC25000 Accuracy: 99.4%
Improved Faster R-CNN and Heilongjiang
[202] 2021 CT scans Accuracy: 89.7%
transfer learning Provincial Hospital
[203] 2021 Three-stream network CT scans LIDC-IDRI Accuracy: 98.2%
[204] 2021 FractalNet CT scans LUNA 16 Sensitivity: 96.68%
[205] 2021 VGG19+CNN X-ray & CT images GitHub Specificity: 99.5%
[205] 2021 ResNet152V2 X-ray & CT images GitHub Specificity: 98.4%
[205] 2021 ResNet152V2+GRU X-ray & CT images GitHub Specificity: 98.7%
[205] 2021 ResNet152V2+Bi-GRU X-ray & CT images GitHub Specificity: 97.8%
[179] 2022 Machine learning CT scans LNDb Accuracy: 94%
Cancers 2022, 14, 5569 15 of 24

Table 4. Cont.

Reference Year Method Imaging Datasets Results

Progressively Growing Channel
[159] 2022 CT scans LIDC-IDRI Accuracy: 95.28%
Attentive Non-Local (ProCAN)
CNN-based multi-task learning
[206] 2022 CT scans LIDC-IDRI Sensitivity: 96.2%
(CNN-MTL)
Cat swarm optimization-based
[207] 2022 CAD for lung cancer classification CT scans Benchmark Specificity: 99.17%
(CSO-CADLCC)
2-Pathway Morphology-based
[208] 2022 CT scans LIDC-IDRI Sensitivity: 96.85%
CNN (2PMorphCNN)

Forte et al. [209] recently conducted a systematic review and meta-analysis of the
diagnostic accuracy of current deep learning approaches for lung cancer diagnosis. The
pooled sensitivity and specificity of deep learning approaches for lung cancer detection
were 93% and 68%, respectively. The results showed that AI plays an important role in
medical imaging, but there are still many research challenges.

5. Challenges and Future Research Directions

This study extensively surveys papers published between 2014 and 2022. Tables 2–4
demonstrate that deep learning-based lung imaging systems have achieved high efficiency
and state-of-the-art performance for lung nodule segmentation, detection, and classification
using existing medical images. Compared to reinforcement and supervised learning
techniques, unsupervised deep learning techniques (such as CNN, Faster R-CNN, Mask R-
CNN, and U-Net) are more popular methods that have been used to develop convolutional
networks for lung cancer detection and false-positive reduction.
Previous studies have shown that CT is the most widely used imaging tool in the CAD
system for lung cancer diagnosis. Compared to 2D CNN, 3D CNN architectures provide
more promising usefulness in obtaining representative features of malignant nodules. To
this day, only a few works on 3D CNN for lung cancer diagnosis have been reported.
Deep learning techniques have achieved good performance in segmentation and
classification. However, deep learning techniques still have many unsolved problems
in lung cancer detection. First, clinicians have not fully acknowledged deep learning
techniques for everyday clinical exercise due to the lack of standardized medical image
acquisition protocols. The unification of the acquisition protocols could minimize it.
Second, deep learning techniques usually require massive annotated medical images
by experienced radiologists to complete training tasks. However, it is costly and time
consuming to collect an enormous annotated image dataset, even performed by experi-
enced radiologists. Several methods were applied to overcome the scarcity of annotated
data. For example, transfer learning is a possible way to solve the training problem of
small samples. Another possible method is the computer synthesis of images, such as the
generation of confrontation networks. Inadequate data will inevitably affect the accuracy
and stability of predictions. Therefore, improving prediction accuracy using weak supervi-
sion, transfer learning, and multi-task learning with small labeled data is one of the future
research directions.
Third, the clinical application of deep learning requires high interpretability, but
current deep learning techniques cannot effectively explain the learned features. Many
researchers have applied visualization and parameter analysis methods to explain deep
learning models. However, there is still a certain distance from the interpretable imaging
markers required by clinical requirements. Therefore, investigating the interpretable deep
learning method will be a hot spot in the medical image field.
Cancers 2022, 14, 5569 16 of 24

Fourth, developing the robustness of the prediction model is a challenging task. Most
deep learning techniques work well only for a single dataset. The image of the same
disease may vary significantly due to different acquisition parameters, equipment, time,
and other factors. This led to poor robustness and generalization of existing deep learning
models. Thus, improving the model structure and training methods by combining brain
cognitive ideas and improving the generalization ability of deep learning is one of the key
future directions.
Finally, some of the current literature has little translation into applicability in clinical
practice due to the lack of experience of non-medical investigators in choosing more
relevant clinical outcomes. Most deep learning techniques were developed by non-medical
professionals with little or no oversight of radiologists, who, in practice, will use these
resources when they become more widely available. As a result, some performance metrics,
such as accuracy, AUC, and precision, which have little meaningful clinical application,
continue to be used and are often the only summary outcomes reported by some studies.
Instead, investigators should always strive to report more relevant clinical parameters,
such as sensitivity and specificity, because they are independent of the prevalence of the
disease and can be more easily translated into practice.
In the future, investigators should pay more attention to the following research di-
rections: (1) develop new convolutional networks and loss functions to improve the per-
formance; (2) weak supervised learning, using a large number of incomplete, inaccurate,
and ambiguous annotation data in the existing medical records to achieve model training;
(3) bring prior clinical knowledge into model training; (4) radiologists, computer scien-
tists, and engineers need to work more closely to develop more realistic and sensitive
models and add more meaning to the research field; (5) single disease identification to
complete disease identification. In clinical examination, only a few cases need to solve
one well-defined problem. For example, clinicians can detect pulmonary nodules in LDCT
and check whether there are other abnormalities, such as emphysema. Solving multiple
problems with one network will not reduce performance in specific tasks. In addition, deep
learning can be explored in some areas where the medical mechanism is not precise, such as
large-scale lung image analysis using deep learning, which is expected to make diagnosing
lung diseases more objective.

6. Conclusions
This paper reviewed recent achievements in deep learning-based approaches for
lung nodule segmentation, detection, and classification. CNN is one of the most widely
used deep learning techniques for lung disease detection and classification, and CT image
datasets are the most frequently used imaging datasets for training networks. The article
review was based on recent publications (published in 2014 and later). Experimental and
clinical trial results demonstrate that deep learning techniques can be superior to trained
radiologists. Deep learning is expected to effectively improve lung nodule segmentation,
detection, and classification. With this powerful tool, radiologists can interpret images more
accurately. Deep learning algorithm has shown great potential in a series of tasks in the
radiology department and has solved many medical problems. However, it still faces many
difficulties, including large-scale clinical verification, patient privacy protection, and legal
accountability. Despite these limitations, with the current trend and rapid development of
the medical industry, deep learning is expected to generate a greater demand for accurate
diagnosis and treatment in the medical field.

Funding: This research was funded by the International Science and Technology Cooperation Project
of the Shenzhen Science and Technology Commission (GJHZ20200731095804014).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Cancers 2022, 14, 5569 17 of 24

Acknowledgments: The author would like to thank the reviewers for their critical comments to
improve the manuscript significantly.
Conflicts of Interest: The author declares no conflict of interest.

References
1. Siegel, R.L.; Miller, K.D.; Fuchs, H.E.; Jemal, A. Cancer statistics. A Cancer J. Clin. 2022, 72, 7–33. [CrossRef] [PubMed]
2. Bade, B.C.; Cruz, C. Lung cancer. Clin. Chest Med. 2020, 41, 1–24. [CrossRef]
3. Stamatis, G.; Eberhard, W.; Pöttgen, C. Surgery after multimodality treatment for non-small-cell lung cancer. Lung Cancer 2004, 45,
S107–S112. [CrossRef] [PubMed]
4. Chiang, T.A.; Chen, P.H.; Wu, P.F.; Wang, T.N.; Chang, P.Y.; Ko, A.M.; Huang, M.S.; Ko, Y.C. Important prognostic factors for the
long-term survival of lung cancer subjects in Taiwan. BMC Cancer 2008, 8, 324. [CrossRef] [PubMed]
5. Journy, N.; Rehel, J.L.; Pointe, H.D.L.; Lee, C.; Brisse, H.; Chateil, J.F.; Caer-Lorho, S.; Laurier, D.; Bernier, M.O. Are the studies on
cancer risk from ct scans biased by indication? Elements of answer from a large-scale cohort study in France. Br. J. Cancer 2015,
112, 1841–1842. [CrossRef]
6. National Lung Screening Trial Research Team. Reduced lung-cancer mortality with low-dose computed tomographic screening.
N. Engl. J. Med. 2011, 365, 395–409. [CrossRef]
7. Ippolito, D.; Capraro, C.; Guerra, L.; De Ponti, E.; Messa, C.; Sironi, S. Feasibility of perfusion CT technique integrated into
conventional (18) FDG/PET-CT studies in lung cancer patients: Clinical staging and functional information in a single study. Eur.
J. Nucl. Med. Mol. Imaging 2013, 40, 156–165. [CrossRef]
8. Park, S.Y.; Cho, A.; Yu, W.S.; Lee, C.Y.; Lee, J.G.; Kim, D.J.; Chung, K.Y. Prognostic value of total lesion glycolysis by F-18-FDG
PET/CT in surgically resected stage IA non-small cell lung cancer. J. Nucl. Med. 2015, 56, 45–49. [CrossRef]
9. Griffiths, H. Magnetic induction tomography. Meas. Sci. Technol. 2011, 12, 1126–1131. [CrossRef]
10. Brown, M.S.; Lo, P.; Goldin, J.G.; Barnoy, E.; Kim, G.H.J.; Mcnitt-Gray, M.F.; Aberle, D.R. Toward clinically usable CAD for lung
cancer screening with computed tomography. Eur. Radiol. 2020, 30, 1822. [CrossRef]
11. Roberts, H.C.; Patsios, D.; Kucharczyk, M.; Paul, N.; Roberts, T.P. The utility of computer-aided detection (CAD) for lung cancer
screening using low-dose CT. Int. Congr. Ser. 2005, 1281, 1137–1142. [CrossRef]
12. Abdul, L.; Rajasekar, S.; Lin, D.S.Y.; Venkatasubramania Raja, S.; Sotra, A.; Feng, Y.; Liu, A.; Zhang, B. Deep-lumen assay-human lung
epithelial spheroid classification from brightfield images using deep learning. Lab A Chip 2021, 21, 447–448. [CrossRef] [PubMed]
13. Armato, S.G.I. Deep learning demonstrates potential for lung cancer detection in chest radiography. Radiology 2020, 297, 697–698.
[CrossRef] [PubMed]
14. Ali, S.; Li, J.; Pei, Y.; Khurram, R.; Rehman, K.U.; Rasool, A.B. State-of-the-Art Challenges and Perspectives in Multi-Organ Cancer
Diagnosis via Deep Learning-Based Methods. Cancers 2021, 13, 5546. [CrossRef]
15. Riquelme, D.; Akhloufi, M.A. Deep Learning for Lung Cancer Nodules Detection and Classification in CT Scans. AI 2020, 1,
28–67. [CrossRef]
16. Zhukov, T.A.; Johanson, R.A.; Cantor, A.B.; Clark, R.A.; Tockman, M.S. Discovery of distinct protein profiles specific for lung
tumors and pre-malignant lung lesions by SELDI mass spectrometry. Lung Cancer 2003, 40, 267–279. [CrossRef]
17. Zeiser, F.A.; Costa, C.; Ramos, G.; Bohn, H.C.; Santos, I.; Roehe, A.V. Deepbatch: A hybrid deep learning model for interpretable
diagnosis of breast cancer in whole-slide images. Expert Syst. Appl. 2021, 185, 115586. [CrossRef]
18. Mandal, M.; Vipparthi, S.K. An empirical review of deep learning frameworks for change detection: Model design, experimental
frameworks, challenges and research needs. IEEE Trans. Intell. Transp. Syst. 2022, 23, 6101–6122. [CrossRef]
19. Alireza, H.; Cheikh, M.; Annika, K.; Jari, V. Deep learning for forest inventory and planning: A critical review on the remote
sensing approaches so far and prospects for further applications. Forestry 2022, 95, 451–465.
20. Highamcatherine, F.; Highamdesmond, J. Deep learning. SIAM Rev. 2019, 32, 860–891.
21. Latifi, K.; Dilling, T.J.; Feygelman, V.; Moros, E.G.; Stevens, C.W.; Montilla-Soler, J.L.; Zhang, G.G. Impact of dose on lung
ventilation change calculated from 4D-CT using deformable image registration in lung cancer patients treated with SBRT. J. Radiat.
Oncol. 2015, 4, 265–270. [CrossRef]
22. Lakshmanaprabu, S.K.; Mohanty, S.N.; Shankar, K.; Arunkumar, N.; Ramirez, G. Optimal deep learning model for classification
of lung cancer on CT images. Future Gener. Comput. Syst. 2019, 92, 374–382.
23. Shim, S.S.; Lee, K.S.; Kim, B.T.; Chung, M.J.; Lee, E.J.; Han, J.; Choi, J.Y.; Kwon, O.J.; Shim, Y.M.; Kim, S. Non-small cell lung
cancer: Prospective comparison of integrated FDG PET/CT and CT alone for preoperative staging. Radiology 2005, 236, 1011–1019.
[CrossRef] [PubMed]
24. Ab, G.D.C.; Domínguez, J.F.; Bolton, R.D.; Pérez, C.F.; Martínez, B.C.; García-Esquinas, M.G.; Carreras Delgado, J.L. PET-CT
in presurgical lymph node staging in non-small cell lung cancer: The importance of false-negative and false-positive findings.
Radiologia 2017, 59, 147–158.
25. Yaturu, S.; Patel, R.A. Metastases to the thyroid presenting as a metabolically inactive incidental thyroid nodule with stable size
in 15 months. Case Rep. Endocrinol. 2014, 2014, 643986. [CrossRef]
Cancers 2022, 14, 5569 18 of 24

26. Eschmann, S.M.; Friedel, G.; Paulsen, F.; Reimold, M.; Hehr, T.; Budach, W.; Langen, H.J.; Bares, R. 18F-FDG PET for assessment
of therapy response and preoperative re-evaluation after neoadjuvant radio-chemotherapy in stage III non-small cell lung cancer.
Eur. J. Nucl. Med. Mol. Imaging 2007, 34, 463–471. [CrossRef]
27. Lee, W.K.; Lau, E.W.; Chin, K.; Sedlaczek, O.; Steinke, K. Modern diagnostic and therapeutic interventional radiology in lung
cancer. J. Thorac. Dis. 2013, 5, 511–523.
28. Zurek, M.; Bessaad, A.; Cieslar, K.; Crémillieux, Y. Validation of simple and robust protocols for high-resolution lung proton MRI
in mice. Magn. Reson. Med. 2010, 64, 401–407. [CrossRef]
29. Burris, N.S.; Johnson, K.M.; Larson, P.E.Z.; Hope, M.D.; Nagle, S.K.; Behr, S.C.; Hope, T.A. Detection of small pulmonary nodules
with ultrashort echo time sequences in oncology patients by using a PET/MR system. Radiology 2016, 278, 239–246. [CrossRef]
30. Fink, C.; Puderbach, M.; Biederer, J.; Fabel, M.; Dietrich, O.; Kauczor, H.U.; Reiser, M.F.; Schönberg, S.O. Lung MRI at 1.5 and 3 tesla:
Observer preference study and lesion contrast using five different pulse sequences. Investig. Radiol. 2007, 42, 377–383. [CrossRef]
31. Cieszanowski, A.; Anyszgrodzicka, A.; Szeszkowski, W.; Kaczynski, B.; Maj, E.; Gornicka, B.; Grodzicki, M.; Grudzinski, I.P.;
Stadnik, A.; Krawczyk, M.; et al. Characterization of focal liver lesions using quantitative techniques: Comparison of apparent
diffusion coefficient values and T2 relaxation times. Eur. Radiol. 2012, 22, 2514–2524. [CrossRef] [PubMed]
32. Hughes, D.; Tiddens, H.; Wild, J.M. Lung imaging in cystic fibrosis. Imaging Decis. MRI 2009, 13, 28–37. [CrossRef]
33. Groth, M.; Henes, F.O.; Bannas, P.; Muellerleile, K.; Adam, G.; Regier, M. Intraindividual comparison of contrast-enhanced MRI and
unenhanced SSFP sequences of stenotic and non-stenotic pulmonary artery diameters. Rofo 2011, 183, 47–53. [CrossRef] [PubMed]
34. Chong, A.L.; Chandra, R.V.; Chuah, K.C.; Roberts, E.L.; Stuckey, S.L. Proton density MRI increases detection of cervical spinal cord
multiple sclerosis lesions compared with T2-weighted fast spin-echo. Am. J. Neuroradiol. 2016, 37, 180–184. [CrossRef] [PubMed]
35. Alzeibak, S.; Saunders, N.H. A feasibility study of in vivo electromagnetic imaging. Phys. Med. Biol. 1993, 38, 151–160. [CrossRef]
[PubMed]
36. Merwa, R.; Hollaus, K.; Brunner, P.; Scharfetter, H. Solution of the inverse problem of magnetic induction tomography (MIT).
Physiol. Meas. 2006, 26, 241–250. [CrossRef] [PubMed]
37. Fernandes, S.L.; Gurupur, V.P.; Lin, H.; Martis, R.J. A novel fusion approach for early lung cancer detection using computer aided
diagnosis techniques. J. Med. Imaging Health Inform. 2017, 7, 1841–1850. [CrossRef]
38. Lu, H. Computer-aided diagnosis research of a lung tumor based on a deep convolutional neural network and global features.
BioMed Res. Int. 2021, 2021, 5513746. [CrossRef] [PubMed]
39. Standford.edu. Deep Learning Tutorial. Available online: http://deeplearning.stanford.edu/tutorial/ (accessed on
5 October 2022).
40. Dauphin, Y.N.; Fan, A.; Auli, M.; Grangier, D. Language Modeling with Gated Convolutional Networks. In Proceedings of the
34th International Conference on Machine Learning, Sydney, Australia, 6–11 August 2017; Volume 70, pp. 933–941.
41. Jeong, J.; Lei, Y.; Shu, H.K.; Liu, T.; Wang, L.; Curran, W.; Shu, H.-K.; Mao, H.; Yang, X. Brain tumor segmentation using 3D mask
R-CNN for dynamic susceptibility contrast enhanced perfusion imaging. Med. Imaging Biomed. Appl. Mol. Struct. Funct. Imaging
2020, 65, 185009.
42. Krizhevsky, A.; Sutskever, I.; Hinton, G. ImageNet classification with deep convolutional neural networks. Commun. ACM 2017,
60, 6. [CrossRef]
43. Kim, J.; Lee, J.K.; Lee, K.M. Accurate image super-resolution using very deep convolutional networks. In Proceedings of the 2016
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 27–30 June 2016.
44. Liu, W.; Chen, W.; Wang, C.; Mao, Q.; Dai, X. Capsule Embedded ResNet for Image Classification. In Proceedings of the 2021 5th
International Conference on Computer Science and Artificial Intelligence (CSAI 2021), Beijing, China, 4–6 December 2021.
45. Guan, X.; Gao, W.; Peng, H.; Shu, N.; Gao, D.W. Image-based incipient fault classification of electrical substation equipment by
transfer learning of deep convolutional neural network. IEEE Can. J. Electr. Comput. Eng. 2021, 45, 1–8. [CrossRef]
46. Warin, K.; Limprasert, W.; Suebnukarn, S.; Jinaporntham, S.; Jantana, P. Performance of deep convolutional neural network for
classification and detection of oral potentially malignant disorders in photographic images. Int. J. Oral Maxillofac. Surg. 2022, 51,
699–704. [CrossRef] [PubMed]
47. Magge, A.; Weissenbacher, D.; Sarker, A.; Scotch, M.; Gonzalez-Hernandez, G. Bi-directional recurrent neural network models for
geographic location extraction in biomedical literature. Pac. Symp. Biocomput. 2019, 24, 100–111. [PubMed]
48. Garg, J.S. Improving segmentation by denoising brain MRI images through interpolation median filter in ADTVFCM. Int. J.
Comput. Trends Technol. 2013, 4, 187–188.
49. Siddeq, M. De-noise color or gray level images by using hybred dwt with wiener filter. Hepato-Gastroenterology 2014, 61, 1308–1312.
50. Rajendran, K.; Tao, S.; Zhou, W.; Leng, S.; Mccollough, C. Spectral prior image constrained compressed sensing reconstruction for
photon-counting detector based CT using a non-local means filtered prior (NLM-SPICCS). Med. Phys. 2018, 6, 45.
51. Powers, D.M.W. Evaluation: From precision, recall and f-measure to roc., informedness, markedness & correlation. J. Mach. Learn.
Technol. 2011, 2, 37–63.
52. Das, A.; Rajendra Acharya, U.; Panda, S.S.; Sabut, S. Deep learning-based liver cancer detection using watershed transform and
Gaussian mixture model techniques. Cogn. Syst. Res. 2019, 54, 165–175. [CrossRef]
53. Lung Image Database Consortium (LIDC). Available online: https://imaging.nci.nih.gov/ncia/login.jsf (accessed on
5 October 2022).
Cancers 2022, 14, 5569 19 of 24

54. Armato Samuel, G.; McLennan, G.; Bidaut, L.; McNitt-Gray, M.F.; Meyer, C.R.; Reeves, A.P.; Zhao, B.; Aberle, D.R.; Henschke, C.I.;
Hoffman, E.A.; et al. Data from LIDC-IDRI. 2015. Available online: https://wiki.cancerimagingarchive.net/display/Public/
LIDC-IDRI (accessed on 5 October 2022).
55. Setio, A.A.A.; Traverso, A.; de Bel, T.; Berens, M.S.N.; van den Bogaard, C.; Cerello, P.; Chen, H.; Dou, Q.; Fantacci, M.E.; Geurts,
B.; et al. Validation, Comparison, and Combination of Algorithms for Automatic Detection of Pulmonary Nodules in Computed
Tomography Images: The LUNA16 Challenge. Med. Image Anal. 2017, 42, 1–13. [CrossRef]
56. ELCAP Public Lung Image Database. 2014. Available online: http://www.via.cornell.edu/lungdb.html (accessed on
5 October 2022).
57. Pedrosa, J.; Aresta, G.; Ferreira, C.; Rodrigues, M.; Leito, P.; Carvalho, A.S.; Rebelo, J.; Negrao, E.; Ramos, I.; Cunha, A.; et al.
LNDb: A Lung Nodule Database on Computed Tomography. arXiv 2019, arXiv:1911.08434.
58. Prasad, D.; Ujjwal, B.; Sanjay, T. LNCDS: A 2D-3D cascaded CNN approach for lung nodule classification, detection and
segmentation. Biomed. Signal Process. Control. 2021, 67, 102527.
59. Shiraishi, J.; Katsuragawa, S.; Ikezoe, A.; Matsumoto, T.; Kobayashi, T.; Komatsu, K.; Matsiu, M.; Fujita, H.; Kodera, Y.; Doi, K.
Development of a digital image database for chest radiographs with and without a lung nodule: Receiver operating characteristic
analysis of radiologists’ detection of pulmonary nodules. Am. J. Roentgen. 2000, 174, 71–74. [CrossRef] [PubMed]
60. Costa, D.D.; Broodman, I.; Hoogsteden, H.; Luider, T.; Klaveren, R.V. Biomarker identification for early detection of lung cancer
by proteomic techniques in the NELSON lung cancer screening trial. Cancer Res. 2008, 68, 3961.
61. Van Ginneken, B.; Armato, S.G.; de Hoop, B.; van Amelsvoort-van de Vorst, S.; Duindam, T.; Niemeijer, M.; Murphy, K.; Schilham,
A.; Retico, A.; Fantacci, M.E.; et al. Comparing and Combining Algorithms for Computer-Aided Detection of Pulmonary Nodules
in Computed Tomography Scans: The ANODE09 Study. Med. Image Anal. 2010, 14, 707–722. [CrossRef] [PubMed]
62. Hospital, S.Z. A Trial to Evaluate the Impact of Lung-Protective Intervention in Patients Undergoing Esophageal Cancer Surgery; US
National Library of Medicine: Bethesda, MD, USA, 2013.
63. Armato Samuel, G., III; Hadjiiski, L.; Tourassi, G.D.; Drukker, K.; Giger, M.L.; Li, F.; Redmond, G.; Farahani, K.; Kirby, J.S.; Clarke,
L.P. SPIE-AAPM-NCI Lung Nodule Classification Challenge Dataset. 2015. Available online: https://wiki.cancerimagingarchive.
net/display/Public/SPIE-AAPM+Lung+CT+Challenge (accessed on 5 October 2022).
64. Li, T.Y.; Li, S.P.; Zhang, Q.L. Protective effect of ischemic preconditioning on lung injury induced by intestinal is-
chemia/reperfusion in rats. Mil. Med. J. South China 2011, 25, 107–110.
65. Li, Y.; Zhang, L.; Chen, H.; Yang, N. Lung nodule detection with deep learning in 3D thoracic MR images. IEEE Access 2019, 7,
37822–37832. [CrossRef]
66. Aerts, H.; Velazquez, E.; Leijenaar, R.; Parmar, C.; Grossman, P.; Carvalho, S.; Bussink, J.; Monshouwer, R.; Haibe-Kains, B.;
Rietveld, D. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 2014,
5, 4006. [CrossRef]
67. Danish Lung Cancer Screening Trial (DLCST)—Full Text View—ClinicalTrials.Gov. Available online: https://clinicaltrials.gov/
ct2/show/NCT00496977 (accessed on 5 October 2022).
68. Trial Summary—Learn—NLST—The Cancer Data Access System. Available online: https://biometry.nci.nih.gov/cdas/learn/
nlst/trial-summary/ (accessed on 5 October 2022).
69. Hu, S.; Hoffman, E.A.; Reinhardt, J.M. Accurate lung segmentation for accurate quantization of volumetric X-ray CT images.
IEEE Trans. Med. Imaging 2001, 20, 490–498. [CrossRef]
70. Dawoud, A. Lung segmentation in chest radiographs by fusing shape information in iterative thresholding. Comput. Vis. IET
2011, 5, 185–190. [CrossRef]
71. Otsu, N. A threshold selection method from gray-level histograms. IEEE Trans. Sys. Man Cyber. 1979, 9, 62–66. [CrossRef]
72. Peng, T.; Wang, C.; Zhang, Y.; Wang, J. H-SegNet: Hybrid segmentation network for lung segmentation in chest radiographs
using mask region-based convolutional neural network and adaptive closed polyline searching method. Phys. Med. Biol. 2022, 67,
075006. [CrossRef] [PubMed]
73. Tseng, L.Y.; Huang, L.C. An adaptive thresholding method for automatic lung segmentation in CT images. In Proceedings of the
IEEE Africon, Nairobi, Kenya, 23–25 September 2009; pp. 1–5.
74. Dehmeshki, J.; Amin, H.; Valdivieso, M.; Ye, X. Segmentation of pulmonary nodules in thoracic CT scans: A region growing
approach. IEEE Trans. Med. Imaging 2008, 27, 467–480. [CrossRef] [PubMed]
75. Fabijacska, A. The influence of pre-processing of CT images on airway tree segmentation using 3D region growing. In Proceedings
of the 5th International Conference on Perspective Technologies and Methods in MEMS Design, Lviv, Ukraine, 22–24 April 2009.
76. Kass, M.; Witkin, A.; Terzopoulos, D. Snakes: Active contour models. Int. J. Comput. Vis. 1988, 1, 321–331. [CrossRef]
77. Lan, Y.; Xu, N.; Ma, X.; Jia, X. Segmentation of Pulmonary Nodules in Lung CT Images based on Active Contour Model. In
Proceedings of the 14th IEEE International Conference on Intelligent Human-Machine Systems and Cybernetics (IHMSC),
Hangzhou, China, 20–21 August 2022.
78. Wang, S.; Zhou, M.; Olivier, G.; Tang, Z.C.; Dong, D.; Liu, Z.Y.; Tian, J. A Multi-view Deep Convolutional Neural Networks
for Lung Nodule Segmentation. In Proceedings of the 2017 39th Annual International Conference of the IEEE Engineering in
Medicine and Biology Society (EMBC), Jeju, Korea, 11–15 July 2017; pp. 1752–1755.
79. Hamidian, S.; Sahiner, B.; Petrick, N.; Pezeshk, A. 3D Convolutional Neural Network for Automatic Detection of Lung Nodules
in Chest CT. Proc. SPIE Int. Soc. Opt. Eng. 2017, 10134, 1013409.
Cancers 2022, 14, 5569 20 of 24

80. Sun, X.F.; Lin, H.; Wang, S.Y.; Zheng, L.M. Industrial robots sorting system based on improved faster RCNN. Comput. Syst. Appl.
2019, 28, 258–263.
81. Cao, H.C.; Liu, H.; Song, E.; Hung, C.C.; Ma, G.Z.; Xu, X.Y.; Jin, R.C.; Jianguo Lu, J.G. Dual-branch residual network for lung
nodule segmentation. Appl. Soft Comput. 2020, 86, 105934. [CrossRef]
82. Banu, S.F.; Sarker, M.; Abdel-Nasser, M.; Puig, D.; Raswan, H.A. AWEU-Net: An attention-aware weight excitation u-net for lung
nodule segmentation. arXiv 2021, arXiv:2110.05144. [CrossRef]
83. Dutta, K. Densely connected recurrent residual (DENSE R2UNET) convolutional neural network for segmentation of lung CT
images. arXiv 2021, arXiv:2102.00663.
84. Keshani, M.; Azimifar, Z.; Tajeripour, F.; Boostani, R. Lung nodule segmentation and recognition using SVM classifier and active
contour modeling: A complete intelligent system. Comput. Biol. Med. 2013, 43, 287–300. [CrossRef]
85. Qi, S.L.; Si, G.L.; Yue, Y.; Meng, X.F.; Cai, J.F.; Kang, Y. Lung nodule segmentation based on thoracic CT images. Beijing Biomed.
Eng. 2014, 33, 29–34.
86. Wang, X.P.; Wen, Z.; Ying, C. Tumor segmentation in lung CT images based on support vector machine and improved level set.
Optoelectron. Lett. 2015, 11, 395–400. [CrossRef]
87. Shen, S.; Bui, A.; Cong, J.J.; Hsu, W. An automated lung segmentation approach using bidirectional chain codes to improve
nodule detection accuracy. Comput. Biol. Med. 2015, 57(C), 139–149. [CrossRef] [PubMed]
88. Roth, H.R.; Farag, A.; Le, L.; Turkbey, E.B.; Summers, R.M. Deep convolutional networks for pancreas segmentation in CT
imaging. Proc. SPIE 2015, 9413, 94131G.
89. Yip, S.; Chintan, P.; Daniel, B.; Jose, E.; Steve, P.; John, K.; Aerts, H.J.W.L. Application of the 3D slicer chest imaging platform
segmentation algorithm for large lung nodule delineation. PLoS ONE 2017, 12, e0178944. [CrossRef]
90. Firdouse, M.J.; Balasubramanian, M. A survey on lung segmentation methods. Adv. Comput. Sci. Technol. 2017, 10, 2875–2885.
91. Khosravan, N.; Bagci, U. Semi-Supervised Multi-Task Learning for Lung Cancer Diagnosis. Annu. Int. Conf. IEEE Eng. Med. Biol.
Soc. 2018, 2018, 710–713.
92. Tong, G.; Li, Y.; Chen, H.; Zhang, Q.; Jiang, H. Improved U-NET network for pulmonary nodules segmentation. Optik 2018, 174,
460–469. [CrossRef]
93. Jiang, J.; Hu, Y.C.; Liu, C.J.; Darragh, H.; Hellmann, M.D.; Deasy, J.O.; Mageras, G.; Veeraraghavan, H. Multiple resolution
residually connected feature streams for automatic lung tumor segmentation from CT images. IEEE Trans. Med. Imaging 2019, 38,
134–144. [CrossRef]
94. Burlutskiy, N.; Gu, F.; Wilen, L.K.; Backman, M.; Micke, P. A deep learning framework for automatic diagnosis in lung cancer.
In Proceedings of the 1st Conference on Medical Imaging with Deep Learning (MIDL 2018), Amsterdam, The Netherlands,
4–6 July 2018.
95. Yan, H.; Lu, H.; Ye, M.; Yan, K.; Jin, Q. Improved Mask R-CNN for Lung Nodule Segmentation. In Proceedings of the 2019 10th
International Conference on Information Technology in Medicine and Education (ITME), Qingdao, China, 23–25 August 2019; pp.
137–147.
96. Xiao, Z.; Liu, B.; Geng, L.; Zhang, F.; Liu, Y. Segmentation of lung nodules using improved 3D-Unet neural network. Symmetry
2020, 12, 1787. [CrossRef]
97. Kashyap, M.; Panjwani, N.; Hasan, M.; Huang, C.; Bush, K.; Dong, P.; Zaky, S.; Chin, A.; Vitzthum, L.; Loo, B.; et al. Deep learning
based identification and segmentation of lung tumors on computed tomography images. Int. J. Radiat. Oncol. Biol. Phys. 2021,
111(3S), E92–E93. [CrossRef]
98. Chen, C.; Zhou, K.; Zha, M.; Qu, X.; Xiao, R. An effective deep neural network for lung lesions segmentation from COVID-19 CT
images. IEEE Trans. Ind. Inform. 2021, 17, 6528–6538. [CrossRef]
99. Zhang, M.; Li, H.; Pan, S.; Lyu, J.; Su, S. Convolutional neural networks based lung nodule classification: A surrogate-assisted
evolutionary algorithm for hyperparameter optimization. IEEE Trans. Evol. Comput. 2021, 25, 869–882. [CrossRef]
100. Jalali, Y.; Fateh, M.; Rezvani, M.; Abolghasemi, V.; Anisi, M.H. ResBCDU-Net: A deep learning framework for lung ct image
segmentation. Sensors 2021, 21, 268. [CrossRef]
101. Balaha, H.; Balaha, M.; Ali, H. Hybrid COVID-19 segmentation and recognition framework (HMB-HCF) using deep learning and
genetic algorithms. Artif. Intell. Med. 2021, 119, 102156. [CrossRef]
102. Lin, X.; Jiao, H.; Pang, Z.; Chen, H.; Wu, W.; Wang, X.; Xiong, L.; Chen, B.; Huang, Y.; Li, S.; et al. Lung cancer and granuloma
identification using a deep learning model to extract 3-dimensional radiomics features in CT imaging. Clin. Lung Cancer 2021, 22,
e756–e766. [CrossRef]
103. Gan, W.; Wang, H.; Gu, H.; Duan, Y.; Xu, Z. Automatic segmentation of lung tumors on CT images based on a 2D & 3D hybrid
convolutional neural network. Br. J. Radiol. 2021, 94, 20210038.
104. Protonotarios, N.E.; Katsamenis, I.; Sykiotis, S.; Dikaios, N.; Kastis, G.A.; Chatziioannou, S.N.; Metaxas, M.; Doulamis, N.;
Doulamis, A. A FEW-SHOT U-NET deep learning model for lung cancer lesion segmentation via PET/CT imaging. Biomed. Phys.
Eng. Express 2022, 8, 025019. [CrossRef]
105. Kim, H.M.; Ko, T.; Young, C.I.; Myong, J.P. Asbestosis diagnosis algorithm combining the lung segmentation method and deep
learning model in computed tomography image. Int. J. Med. Inform. 2022, 158, 104667. [CrossRef]
106. Chang, C.Y.; Chen, S.J.; Tsai, M.F. Application of support-vector-machine-based method for feature selection and classification of
thyroid nodules in ultrasound images. Pattern Recognit. 2010, 43, 3494–3506. [CrossRef]
Cancers 2022, 14, 5569 21 of 24

107. Nithila, E.E.; Kumar, S.S. Segmentation of lung nodule in CT data using active contour model and Fuzzy C-mean clustering. Alex.
Eng. J. 2016, 55, 2583–2588. [CrossRef]
108. Zhang, X. Computer-Aided Detection of Pulmonary Nodules in Helical CT Images. Ph.D. Dissertation, The University of Iowa,
Iowa City, IA, USA, 2005.
109. Hwang, J.; Chung, M.J.; Bae, Y.; Shin, K.M.; Jeong, S.Y.; Lee, K.S. Computer-aided detection of lung nodules. J. Comput. Assist.
Tomogr. 2010, 34, 31–34. [CrossRef] [PubMed]
110. Young, S.; Lo, P.; Kim, G.; Brown, M.; Hoffman, J.; Hsu, W.; Wahi-Anwar, W.; Flores, C.; Lee, G.; Noo, F.; et al. The effect of
radiation dose reduction on computer-aided detection (CAD) performance in a low-dose lung cancer screening population. Med.
Phys. 2017, 44, 1337–1346. [CrossRef] [PubMed]
111. Tajbakhsh, N.; Suzuki, K. Comparing two classes of end-to-end machine-learning models in lung nodule detection and classifica-
tion: MTANNS VS. CNNS. Pattern Recognit. 2017, 63, 476–486. [CrossRef]
112. Liu, X.; Hou, F.; Hong, Q.; Hao, A. Multi-view multi-scale CNNs for lung nodule type classification from CT images. Pattern
Recognit. 2018, 77, 262–275. [CrossRef]
113. Cao, H.C.; Liu, H.; Song, E.; Ma, G.Z.; Xu, X.Y.; Jin, R.C.; Liu, T.Y.; Hung, C.C. A Two-Stage Convolutional Neural Networks for
Lung Nodule Detection. IEEE J. Biomed. Health Inform. 2020, 24, 2006–2015. [CrossRef]
114. Alakwaa, W.; Nassef, M.; Badr, A. Lung cancer detection and classification with 3D convolutional neural network (3D-CNN).
Lung Cancer 2017, 8, 409–417. [CrossRef]
115. Anirudh, R.; Thiagarajan, J.J.; Bremer, T.; Kim, H. Lung nodule detection using 3D convolutional neural networks trained on
weakly labeled data. In Proceedings of the Medical Imaging 2016: Computer-Aided Diagnosis, International Society for Optics
and Photonics, San Diego, CA, USA, 27 February–3 March 2016; Volume 9785, pp. 1–6.
116. Feng, Y.; Hao, P.; Zhang, P.; Liu, X.; Wu, F.; Wang, H. Supervoxel based weakly-supervised multi-level 3D CNNs for lung nodule
detection and segmentation. J. Ambient. Intell. Humaniz. Comput. 2019. [CrossRef]
117. Perez, G.; Arbelaez, P. Automated lung cancer diagnosis using three-dimensional convolutional neural networks. Med. Biol. Eng.
Comput. 2020, 58, 1803–1815. [CrossRef]
118. Vipparla, V.K.; Chilukuri, P.K.; Kande, G.B. Attention based multi-patched 3D-CNNs with hybrid fusion architecture for reducing
false positives during lung nodule detection. J. Comput. Commun. 2021, 9, 1–26. [CrossRef]
119. Dutande, P.; Baid, U.; Talbar, S. Deep residual separable convolutional neural network for lung tumor segmentation. Comput. Biol.
Med. 2022, 141, 105161. [CrossRef] [PubMed]
120. Luo, X.; Song, T.; Wang, G.; Chen, J.; Chen, Y.; Li, K.; Metaxas, D.N.; Zhang, S. SCPM-Net: An anchor-free 3D lung nodule detection
network using sphere representation and center points matching. Med. Image Anal. 2022, 75, 102287. [CrossRef] [PubMed]
121. Franck, C.; Snoeckx, A.; Spinhoven, M.; Addouli, H.E.; Zanca, F. Pulmonary nodule detection in chest CT using a deep
learning-based reconstruction algorithm. Radiat. Prot. Dosim. 2021, 195, 158–163. [CrossRef] [PubMed]
122. Dou, Q.; Chen, H.; Yu, L.; Qin, J.; Heng, P.A. Multi-level contextual 3D CNNs for false positive reduction in pulmonary nodule
detection. IEEE Trans. Biomed. Eng. 2016, 64, 1558–1567. [CrossRef]
123. Setio, A.; Ciompi, F.; Litjens, G.; Gerke, P.; Jacobs, C.; Riel, S.; Wille, M.M.; Naqibullah, M.; Sanchez, C.I.; van Ginneken, B.
Pulmonary nodule detection in CT images: False positive reduction using multi-view convolutional networks. IEEE Trans. Med.
Imaging 2016, 35, 1160–1169. [CrossRef] [PubMed]
124. Mercy Theresa, M.; Subbiah Bharathi, V. CAD for lung nodule detection in chest radiography using complex wavelet transform
and shearlet transform features. Indian J. Sci. Technol. 2016, 9, 1–12. [CrossRef]
125. Jin, T.; Hui, C.; Shan, Z.; Wang, X. Learning Deep Spatial Lung Features by 3D Convolutional Neural Network for Early Cancer
Detection. In Proceedings of the 2017 International Conference on Digital Image Computing: Techniques and Applications
(DICTA), Sydney, Australia, 29 November–1 December 2017.
126. Zhu, W.; Liu, C.; Fan, W.; Xie, X. DeepLung: Deep 3D dual path nets for automated pulmonary nodule detection and classification.
arXiv 2017, arXiv:1709.05538.
127. Eun, H.Y.; Kim, D.Y.; Jung, C.; Kim, C. Single-view 2D CNNs with Fully Automatic Non-nodule Categorization for False Positive
Reduction in Pulmonary Nodule Detection. Comput. Methods Programs Biomed. 2018, 165, 215–224. [CrossRef]
128. Ramachandran, S.; George, J.; Skaria, S.; Varun, V.V. Using yolo based deep learning network for real time detection and
localization of lung nodules from low dose CT scans. In Proceedings of the SPIE 10575, Medical Imaging 2018: Computer-Aided
Diagnosis, 105751I (2018). Houston, TX, USA, 27 February 2018. [CrossRef]
129. Serj, M.F.; Lavi, B.; Hoff, G.; Valls, D.P. A deep convolutional neural network for lung cancer diagnostic. arXiv 2018,
arXiv:1804.08170.
130. Zhang, J.; Xia, Y.; Zeng, H.; Zhang, Y. Nodule: Combining constrained multi-scale log filters with densely dilated 3D deep
convolutional neural network for pulmonary nodule detection. Neurocomputing 2018, 317, 159–167. [CrossRef]
131. Schwyzer, M.; Ferraro, D.A.; Muehlematter, U.J.; Curioni-Fontecedro, A.; Messerli, M. Automated detection of lung cancer at
ultralow dose PET/CT by deep neural networks—Initial results. Lung Cancer 2018, 126, 170–173. [CrossRef] [PubMed]
132. Gerard, S.E.; Patton, T.J.; Christensen, G.E.; Bayouth, J.E.; Reinhardt, J.M. Fissurenet: A deep learning approach for pulmonary
fissure detection in CT images. IEEE Trans. Med. Imaging 2018, 38, 156–166. [CrossRef] [PubMed]
133. Zhong, Z.S.; Kim, Y.S.; Plichta, K.; Allen, B.G.; Zhou, L.X.; Buatti, J.; Wu, X.D. Simultaneous cosegmentation of tumors in PET-CT
images using deep fully convolutional networks. Med. Phys. 2019, 2, 619–633. [CrossRef] [PubMed]
Cancers 2022, 14, 5569 22 of 24

134. Kim, B.C.; Choi, J.S.; Suk, H.I. Multi-scale gradual integration CNN for false positive reduction in pulmonary nodule detection.
Neural Netw. 2019, 115, 1–10. [CrossRef] [PubMed]
135. Masood, A.; Sheng, B.; Li, P.; Hou, X.; Wei, X.; Qin, J.; Feng, D. Computer-assisted decision support system in pulmonary cancer
detection and stage classification on CT images. J. Biomed. Inform. 2018, 79, 117–128. [CrossRef]
136. Nam, J.G.; Park, S.; Hwang, E.J.; Lee, J.H.; Jin, K.-N.; Lim, K.Y.; Vu, T.H.; Sohn, J.H.; Hwang, S.; Goo, J.M.; et al. Development and
validation of deep learning-based automatic detection algorithm for malignant pulmonary nodules on chest radiographs. Clin.
Infect. Dis. 2019, 69, 739–747. [CrossRef]
137. Choi, W.; Oh, J.H.; Riyahi, S.; Liu, C.J.; Lu, W. Radiomics analysis of pulmonary nodules in low-dose CT for early detection of
lung cancer. Med. Phys. 2018, 45, 1537–1549. [CrossRef]
138. Tan, J.X.; Huo, Y.M.; Liang, Z.; Li, L. Expert knowledge-infused deep learning for automatic lung nodule detection. J. X-Ray Sci.
Technol. 2019, 27, 17–35. [CrossRef]
139. Ozdemir, O.; Russell, R.L.; Berlin, A.A. A 3D probabilistic deep learning system for detection and diagnosis of lung cancer using
low-dose CT scans. IEEE Trans. Med. Imaging 2020, 39, 1419–1429. [CrossRef]
140. Cha, M.J.; Chung, M.J.; Lee, J.H.; Lee, K.S. Performance of deep learning model in detecting operable lung cancer with chest
radiographs. J. Thorac. Imaging 2019, 34, 86–91. [CrossRef]
141. Pham, H.; Futakuchi, M.; Bychkov, A.; Furukawa, T.; Fukuoka, J. Detection of lung cancer lymph node metastases from whole-slide
histopathologic images using a two-step deep learning approach. Am. J. Pathol. 2019, 189, 2428–2439. [CrossRef] [PubMed]
142. Li, D.; Vilmun, B.M.; Carlsen, J.F.; Albrecht-Beste, E.; Lauridsen, C.A.; Nielsen, M.B.; Hansen, K.L. The performance of deep
learning algorithms on automatic pulmonary nodule detection and classification tested on different datasets that are not derived
from LIDC-IDRI: A systematic review. Diagnostics 2019, 9, 207. [CrossRef] [PubMed]
143. Li, X.; Jin, W.; Li, G.; Yin, C. Yolo v2 network with asymmetric convolution kernel for lung nodule detection of CT image. Chin. J.
Biomed. Eng. 2019, 38, 401–408.
144. Guo, T.; Xie, S.P. Automated segmentation and identification of pulmonary nodule images. Comput. Eng. Des. 2019, 40, 467–472.
145. Huang, W.; Hu, L. Using a noisy U-net for detecting lung nodule candidates. IEEE Access 2019, 7, 67905–67915. [CrossRef]
146. Gu, Y.; Lu, X.; Zhang, B.; Zhao, Y.; Zhou, T. Automatic lung nodule detection using multi-scale dot nodule-enhancement filter
and weighted support vector machines in chest computed tomography. PLoS ONE 2019, 14, e0210551. [CrossRef]
147. Kumar, A.; Fulham, M.J.; Feng, D.; Kim, J. Co-learning feature fusion maps from PET-CT images of lung cancer. IEEE Trans. Med.
Imaging 2019, 39, 204–217. [CrossRef]
148. Pesce, E.; Withey, S.; Ypsilantis, P.P.; Bakewell, R.; Goh, V.; Montana, G. Learning to detect chest radiographs containing pulmonary
lesions using visual attention networks. Med. Image Anal. 2019, 53, 26–38. [CrossRef]
149. Huang, X.; Lei, Q.; Xie, T.; Zhang, Y.; Hu, Z.; Zhou, Q. Deep transfer convolutional neural network and extreme learning machine
for lung nodule diagnosis on CT images. Knowl. Based Syst. 2020, 204, 105230. [CrossRef]
150. Zheng, S.; Cornelissen, L.J.; Cui, X.; Jing, X.; Ooijen, P. Deep convolutional neural networks for multiplanar lung nodule detection:
Improvement in small nodule identification. Med. Phys. 2021, 48, 733–744. [CrossRef]
151. Xu, X.; Wang, C.; Guo, J.; Gan, Y.; Yi, Z. MSCS-DEEPLN: Evaluating lung nodule malignancy using multi-scale cost-sensitive
neural networks. Med. Image Anal. 2020, 65, 101772. [CrossRef] [PubMed]
152. Yektai, H.; Manthouri, M. Diagnosis of lung cancer using multi-scale convolutional neural network. Biomed. Eng. Appl. Basis
Commun. 2020, 32, 2050030. [CrossRef]
153. Heuvelmans, M.A.; Ooijen, P.; Ather, S.; Silva, C.F.; Oudkerk, M. Lung cancer prediction by deep learning to identify benign lung
nodules. Lung Cancer 2021, 154, 1–4. [CrossRef]
154. Hsu, H.H.; Ko, K.H.; Chou, Y.C.; Wu, Y.C.; Chiu, S.H.; Chang, C.K.; Chang, W.C. Performance and reading time of lung nodule
identification on multidetector CT with or without an artificial intelligence-powered computer-aided detection system. Clin.
Radiol. 2021, 76, 626.e23. [CrossRef] [PubMed]
155. Lee, J.M.; Choi, E.J.; Chung, J.H.; Lee, K.W.; Oh, J.W. A DNA-derived phage nose using machine learning and artificial neural
processing for diagnosing lung cancer. Biosens. Bioelectron. 2021, 194, 113567. [CrossRef] [PubMed]
156. Afshar, P.; Naderkhani, F.; Oikonomou, A.; Rafiee, M.J.; Plataniotis, K.N. MIXCAPS: A capsule network-based mixture of experts
for lung nodule malignancy prediction. Pattern Recognit. 2021, 116(August 2021), 107942. [CrossRef]
157. Lai, K.D.; Nguyen, T.T.; Le, T.H. Detection of lung nodules on ct images based on the convolutional neural network with attention
mechanism. Ann. Emerg. Technol. Comput. 2021, 5, 78–89. [CrossRef]
158. Bu, Z.; Zhang, X.; Lu, J.; Lao, H.; Liang, C.; Xu, X.; Wei, Y.; Zeng, H. Lung nodule detection based on YOLOv3 deep learning with
limited datasets. Mol. Cell. Biomech. 2022, 19, 17–28. [CrossRef]
159. Al-Shabi, M.; Shak, K.; Tan, M. ProCAN: Progressive growing channel attentive non-local network for lung nodule classification.
Pattern Recognit. 2022, 122, 108309. [CrossRef]
160. Farag, A.; Ali, A.; Graham, J.; Farag, A.; Elshazly, S.; Falk, R. Evaluation of geometric feature descriptors for detection and
classification of lung nodules in low dose CT scans of the chest. In Proceedings of the 2011 IEEE International Symposium on
Biomedical Imaging: From Nano to Macro, Chicago, IL, USA, 30 March–2 April 2011; pp. 169–172.
161. Orozco, H.M.; Villegas, O.O.V.; Domínguez, H.J.O.; Domínguez, H.D.J.O.; Sanchez, V.G.C. Lung nodule classification in CT
thorax images using support vector machines. In Proceedings of the 2013 12th Mexican International Conference on Artificial
Intelligence (MICAI), Mexico City, Mexico, 24–30 November 2013; pp. 277–283.
Cancers 2022, 14, 5569 23 of 24

162. Krewer, H.; Geiger, B.; Hall, L.O.; Goldgof, D.B.; Gu, Y.; Tockman, M.; Gillies, R.J. Effect of texture features in computer aided
diagnosis of pulmonary nodules in low-dose computed tomography. In Proceedings of the 2013 IEEE International Conference
on Systems, Man, and Cybernetics (SMC), Manchester, UK, 13–16 October 2013; pp. 3887–3891.
163. Parveen, S.S.; Kavitha, C. Classification of lung cancer nodules using SVM Kernels. Int. J. Comput. Appl. 2014, 95, 975–8887.
164. Dandıl, E.; Çakiroğlu, M.; Ekşi, Z.; Özkan, M.; Kurt, Ö.K.; Canan, A. Artificial neural network-based classification system for lung
nodules on computed tomography scans. In Proceedings of the 2014 6th International Conference of Soft Computing and Pattern
Recognition (SoCPar), Tunis, Tunisia, 11–14 August 2014; pp. 382–386.
165. Hua, K.L.; Hsu, C.H.; Hidayati, S.C.; Hidayati, S.C.; Cheng, W.H.; Chen, Y.J. Computer-aided classification of lung nodules on
computed tomography images via deep learning technique. OncoTargets Ther. 2015, 8, 2015–2022.
166. Kumar, D.; Wong, A.; Clausi, D.A. Lung nodule classification using deep features in CT images. In Proceedings of the 2015 12th
Conference on Computer and Robot Vision (CRV), Halifax, NS, Canada, 3–5 June 2015; pp. 133–138.
167. Shen, S.; Han, S.X.; Aberle, D.R.; Bui, A.A.; Hsu, W. An interpretable deep hierarchical semantic convolutional neural network for
lung nodule malignancy classification. Expert Syst. Appl. 2019, 128, 84–95. [CrossRef] [PubMed]
168. Shen, W.; Zhou, M.; Yang, F.; Yang, C.; Tian, J. Muti-scale convolutional neural networks for lung nodule Classification. Inf.
Process. Med. Imaging 2015, 24, 588–599. [PubMed]
169. Cheng, J.Z.; Ni, D.; Chou, Y.H.; Qin, J.; Tiu, C.M.; Chang, Y.C.; Huang, C.S.; Shen, D.; Chen, C.M. Computer-aided diagnosis with
deep learning architecture: Applications to breast lesions in US images and pulmonary nodules in CT scans. Sci. Rep. 2016, 6,
24454. [CrossRef]
170. Kwajiri, T.L.; Tezukam, T. Classification of Lung Nodules Using Deep Learning. Trans. Jpn. Soc. Med. Biol. Eng. 2017, 55, 516–517.
171. Shen, W.; Zhou, M.; Yang, F.; Yu, D.; Dong, D.; Yang, C.; Tian, J.; Zang, Y. Multi-crop convolutional neural networks for lung
nodule malignancy suspiciousness classification. Pattern Recognit. 2017, 61, 663–673. [CrossRef]
172. Abbas, Q. Lung-deep: A computerized tool for detection of lung nodule patterns using deep learning algorithms detection of
lung nodules patterns. Int. J. Adv. Comput. Sci. Appl. 2017, 8, 112–116. [CrossRef]
173. Da Silva, G.L.F.; da Silva Neto, O.P.; Silva, A.C.; Gattass, M. Lung nodules diagnosis based on evolutionary convolutional neural
network. Multimed. Tools Appl. 2017, 76, 19039–19055. [CrossRef]
174. Zhang, B.; Qi, S.; Monkam, P.; Li, C.; Qian, W. Ensemble learners of multiple deep CNNs for pulmonary nodules classification
using CT images. IEEE Access 2019, 7, 110358–110371. [CrossRef]
175. Sahu, P.; Yu, D.; Dasari, M.; Hou, F.; Qin, H. A lightweight multi-section CNN for lung nodule classification and malignancy
estimation. IEEE J. Biomed. Health Inform. 2019, 23, 960–968. [CrossRef]
176. Ali, I.; Muzammil, M.; Ulhaq, D.I.; Khaliq, A.A.; Malik, S. Efficient lung nodule classification using transferable texture
convolutional neural network. IEEE Access 2020, 8, 175859–175870. [CrossRef]
177. Marques, S.; Schiavo, F.; Ferreira, C.A.; Pedrosa, J.; Cunha, A.; Campilho, A. A multi-task CNN approach for lung nodule
malignancy classification and characterization. Expert Syst. Appl. 2021, 184, 115469.1–115469.9. [CrossRef]
178. Thamilarasi, V.; Roselin, R. Automatic classification and accuracy by deep learning using CNN methods in lung chest x-ray
images. IOP Conf. Ser. Mater. Sci. Eng. 2021, 1055, 012099. [CrossRef]
179. Kawathekar, I.D.; Areeckal, A.S. Performance analysis of texture characterization techniques for lung nodule classification. J. Phys.
Conf. Ser. 2022, 2161, 012045. [CrossRef]
180. Radford, A.; Metz, L.; Chintala, S. Unsupervised representation learning with deep convolutional generative adversarial networks.
arXiv 2016, arXiv:1511.06434.
181. Chuquicusma, M.J.M.; Hussein, S.; Burt, J.; Bagci, U. How to fool radiologists with generative adversarial networks? A visual
turing test for lung cancer diagnosis. In Proceedings of the 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI
2018), Washington, DC, USA, 4–7 April 2018; pp. 240–244.
182. Zhao, D.; Zhu, D.; Lu, J.; Luo, Y.; Zhang, G. Synthetic Medical Images Using F&BGAN for Improved Lung Nodules Classification
by Multi-Scale VGG16. Symmetry 2018, 10, 519.
183. Teramoto, A.; Yamada, A.; Kiriyama, Y.; Tsukamoto, T.; Fujita, H. Automated classification of benign and malignant cells from
lung cytological images using deep convolutional neural network. Inform. Med. Unlocked 2019, 16, 100205. [CrossRef]
184. Rani, K.V.; Jawhar, S.J. Superpixel with nanoscale imaging and boosted deep convolutional neural network concept for lung
tumor classification. Int. J. Imaging Syst. Technol. 2020, 30, 899–915. [CrossRef]
185. Kuruvilla, J.; Gunavathi, K. Lung cancer classification using neural networks for CT images. Comput. Methods Programs Biomed
2014, 113, 202–209. [CrossRef]
186. Ciompi, F.; Chung, K.; Riel, S.V.; Setio, A.; Gerke, P.K.; Jacobs, C.; Scholten, E.T.; Schaefer-Prokop, C.; Wille, M.M.W.; Marchianò,
A.; et al. Corrigendum: Towards automatic pulmonary nodule management in lung cancer screening with deep learning. Sci.
Rep. 2017, 7, 46878. [CrossRef]
187. Nurtiyasari, D.; Rosadi, D.; Abdurakhman. The application of Wavelet Recurrent Neural Network for lung cancer classification.
In Proceedings of the 2017 3rd International Conference on Science and Technology—Computer (ICST), Yogyakarta, Indonesia,
11–12 July 2017; pp. 127–130.
188. De Carvalho Filho, A.O.; Silva, A.C.; de Paiva, A.C.; Nunes, R.A.; Gattass, M. Classification of patterns of benignity and
malignancy based on CT using topology-based phylogenetic diversity index and convolutional neural network. Pattern Recognit.
2018, 81, 200–212. [CrossRef]
Cancers 2022, 14, 5569 24 of 24

189. Lindsay, W.; Wang, J.; Sachs, N.; Barbosa, E.; Gee, J. Transfer learning approach to predict biopsy-confirmed malignancy of
lung nodules from imaging data: A pilot study. In Image Analysis for Moving Organ, Breast, and Thoracic Images; Springer:
Berlin/Heidelberg, Germany, 2018; pp. 295–301.
190. Keming, M.; Renjie, T.; Xinqi, W.; Weiyi, Z.; Haoxiang, W. Feature representation using deep autoencoder for lung nodule image
classification. Complexity 2018, 3078374. [CrossRef]
191. Matsuyama, E.; Tsai, D.Y. Automated classification of lung diseases in computed tomography images using a wavelet based
convolutional neural network. J. Biomed. Sci. Eng. 2018, 11, 263–274. [CrossRef]
192. Sathyan, H.; Panicker, J.V. Lung Nodule Classification Using Deep ConvNets on CT Images. In Proceedings of the 2018 9th
International Conference on Computing, Communication and Networking Technologies, Bengaluru, India, 10–12 July 2018;
p. 18192544. [CrossRef]
193. Xie, Y.; Xia, Y.; Zhang, J.; Song, Y.; Feng, D.; Fulham, M.; Cai, W. Knowledge-based collaborative deep learning for benign-
malignant lung nodule classification on chest CT. IEEE Trans. Med. Imaging 2019, 38, 991–1004. [CrossRef]
194. Nasrullah, N.; Sang, J.; Alam, M.S.; Mateen, M.; Cai, B.; Hu, H. Automated lung nodule detection and classification using deep
learning combined with multiple strategies. Sensors 2019, 19, 3722. [CrossRef]
195. Shakeel, P.M.; Burhanuddin, M.A.; Desa, M.I. Lung cancer detection from CT image using improved profuse clustering and deep
learning instantaneously trained neural networks. Measurement 2019, 145, 702–712. [CrossRef]
196. Suresh, S.; Mohan, S. Roi-based feature learning for efficient true positive prediction using convolutional neural network for lung
cancer diagnosis. Neural Comput. Appl. 2020, 32, 15989–16009. [CrossRef]
197. Su, R.; Xie, W.; Tan, T. 2.75D convolutional neural network for pulmonary nodule classification in chest CT. arXiv 2020,
arXiv:2002.04251.
198. Zia, M.B.; Zhao, J.J.; Ning, X. Detection and classification of lung nodule in diagnostic CT: A TSDN method based on improved
3D-FASTER R-CNN and multi-scale multi-crop convolutional neural network. Int. J. Hybrid Inf. Technol. 2020, 13, 45–56. [CrossRef]
199. Lin, C.J.; Li, Y.C. Lung nodule classification using taguchi-based convolutional neural networks for computer tomography images.
Electronics 2020, 9, 1066. [CrossRef]
200. Mmmap, A.; Sjj, B.; Gjm, C. Optimal deep belief network with opposition based pity beetle algorithm for lung cancer classification:
A DBNOPBA approach. Comput. Methods Programs Biomed 2021, 199, 105902.
201. Baranwal, N.; Doravari, P.; Kachhoria, R. Classification of histopathology images of lung cancer using convolutional neural
network (CNN). arXiv 2021, arXiv:2112.13553.
202. Shiwei, L.I.; Liu, D. Automated classification of solitary pulmonary nodules using convolutional neural network based on transfer
learning strategy. J. Mech. Med. Biol. 2021, 21, 2140002.
203. Arumuga Maria Devi, T.; Mebin Jose, V.I. Three Stream Network Model for Lung Cancer Classification in the CT Images. Open
Comput. Sci. 2021, 11, 251–261. [CrossRef]
204. Naik, A.; Edla, D.R.; Kuppili, V. Lung Nodule Classification on Computed Tomography Images Using Fractalnet. Wireless Pers
Commun 2021, 119, 1209–1229. [CrossRef]
205. Ibrahim, D.M.; Elshennawy, N.M.; Sarhan, A.M. Deep-chest: Multi-classification deep learning model for diagnosing COVID-19,
pneumonia, and lung cancer chest diseases. Comput. Biol. Med. 2021, 132, 104348. [CrossRef]
206. Fu, X.; Bi, L.; Kumar, A.; Fulham, M.; Kim, J. An attention-enhanced cross-task network to analyse lung nodule attributes in CT
images. Pattern Recognit. 2022, 126, 108576. [CrossRef]
207. Vaiyapuri, T.; Liyakathunisa; Alaskar, H.; Parvathi, R.; Pattabiraman, V.; Hussain, A. CAT Swarm Optimization-Based Computer-
Aided Diagnosis Model for Lung Cancer Classification in Computed Tomography Images. Appl. Sci. 2022, 12, 5491. [CrossRef]
208. Halder, A.; Chatterjee, S.; Dey, D. Adaptive morphology aided 2-pathway convolutional neural network for lung nodule
classification. Biomed. Signal Process. Control 2022, 72, 103347. [CrossRef]
209. Forte, G.C.; Altmayer, S.; Silva, R.F.; Stefani, M.T.; Libermann, L.L.; Cavion, C.C.; Youssef, A.; Forghani, R.; King, J.;
Mohamed, T.-L.; et al. Deep Learning Algorithms for Diagnosis of Lung Cancer: A Systematic Review and Meta-Analysis.
Cancers 2022, 14, 3856. [CrossRef]