Wang et al.

BMC Oral Health () : BMC Oral Health

https://doi.org/10.1186/s12903-023-03039-2

RESEARCH Open Access

New classication for bone type at dental

implant sites: a dental computed tomography
study
Shiuan‑Hui Wang1, Jui‑Ting Hsu1,2*, Lih‑Jyh Fuh1,3, Shin‑Lei Peng4, Heng‑Li Huang1,2 and Ming‑Tzu Tsai5

Abstract
Objective This study proposed a new classication method of bone quantity and quality at the dental implant site
using cone‑beam computed tomography (CBCT) image analysis, classifying cortical and cancellous bones separately
and using CBCT for quantitative analysis.
Methods Preoperative CBCT images were obtained from 128 implant patients (315 sites). First, measure the crestal
cortical bone thickness (in mm) and the cancellous bone density [in grayscale values (GV) and bone mineral den‑
sity (g/cm3)] at the implant sites. The new classication for bone quality at the implant site proposed in this study is
a “nine‑square division” bone classication system, where the cortical bone thickness is classied into A: > 1.1 mm,
B:0.7–1.1 mm, and C: < 0.7 mm, and the cancellous bone density is classied into 1: > 600 GV (= 420 g/cm3), 2:300–600
GV (= 160 g/cm3–420 g/cm3), and 3: < 300 GV (= 160 g/cm3).
Results The results of the nine bone type proportions based on the new jawbone classication were as follows: A1
(8.57%,27/315), A2 (13.02%), A3 (4.13%), B1 (17.78%), B2 (20.63%), B3 (8.57%) C1 (4.44%), C2 (14.29%), and C3 (8.57%).
Conclusions The proposed classication can complement the parts overlooked in previous bone classication meth‑
ods (bone types A3 and C1).
Trial registration The retrospective registration of this study was approved by the Institutional Review Board of
China Medical University Hospital, No. CMUH 108‑REC2‑181.
Keywords Bone classication, Dental CBCT, Cortical bone thickness, Cancellous bone density, Jawbone

Introduction
e number of patients with missing teeth has been
increasing along with the aging population in recent
*Correspondence: years. e use of dental implants is one of the most
Jui‑Ting Hsu common treatment methods for restoring the normal
[email protected]; [email protected]
1
occlusal function of patients with missing teeth [1].
School of Dentistry, China Medical University, Taichung 404, Taiwan
2
Department of Biomedical Engineering, China Medical University, erefore, how to increase the dental implant success
Taichung 404, Taiwan rate is a critical issue. e success rate of dental implants
3
Department of Dentistry, China Medical University and Hospital, depends on several factors, including the surgeon’s surgi-
Taichung 404, Taiwan
4
Department of Biomedical Imaging and Radiological Science, China cal skills, the patient’s postoperative oral hygiene habits,
Medical University, Taichung 404, Taiwan and the thread design or surface treatment of the den-
5
Department of Biomedical Engineering, Hungkuang University, tal implant. Notably, the bone quality of the jawbone is
Taichung 433, Taiwan
one of the essential inuencing factors [2–6]. Previous

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Wang et al. BMC Oral Health () : Page 2 of 13

literature has reported that good jawbone quality can site using CBCT image analysis, classifying cortical and
provide better initial stability, allowing for better osse- cancellous bones separately, and using CBCT for quan-
ointegration in future recovery and ensuring a more sta- titative analysis. All of the jawbones that may be encoun-
ble implant, thereby increasing the success rate of dental tered clinically could be covered by this classication
implant surgeries [7]. method, thereby providing a reference basis for dentists
e jawbone structure is composed of two layers. to classify the jawbone quality before dental implant
Dense cortical bone forms the outer layer, and a porous surgery.
cancellous bone with trabecular bone structures forms
the inner layer. e success rate of dental implant sur- Materials and methods
gery is highly dependent on the jawbone quality, and for Dental CBCT examinations of patients and implant sites
its assessment, numerous scholars have proposed various is study was approved by the Institutional Review
classication methods. ese methods can be grouped Board of China Medical University Hospital, No. CMUH
into three types; in the following sections, they will be 108-REC2-181. We conrmed that all the methods were
referred to as Type I, Type II, and Type III classication performed in accordance with relevant guidelines and
methods. Type I refers to Lekholm and Zarb’s jawbone regulations. Informed consent was waived by CMUH
quality classication method and its extensions. At the 108-REC2-181 owing to the retrospective nature of the
time of writing, the bone quality classication method study. Samples for this experiment were collected for
they proposed in 1985 [8] is the most widely used. When implant-planning purposes.
using this method, the observer subjectively ranks jaw- is retrospective study was conducted at the Den-
bones into four types according to the proportions of tal Division of China Medical University Hospital from
cortical and cancellous bone, namely Bone Type 1–Type August 2018 to March 2020. Promax 3D Max (Planmeca,
4. Several scholars then began using computed tomog- Helsinki, Finland) was used for dental CBCT imaging,
raphy (CT) or cone-beam CT (CBCT) to explore this and the scanning parameters were set as follows: voxel
method in greater depth [9–11]. However, the subjective size 200 μm, voltage 96 kV, and current 12.5 mA. In this
nature of this method and the absence of quantitative study, 315 suitable dental implant sites recorded from a
analysis may lead to dierent results depending on the total of 128 implant patients (66 males and 62 females)
observer. e Type II method is based on the tactile of were collected, including 42 anterior mandible sites, 127
bone drilling during implant surgery and image Houns- anterior maxilla sites, 39 posterior mandible sites, and
eld unit (HU) for classication and was initially pro- 107 posterior maxilla sites. e study inclusion criteria
posed by Misch et al. in 1989 [12]. e Type III method were as follows: (a) CBCT images were taken before the
refers to the system that classies the cortical bone and dental implant surgery, and (b) a dental surgical stent
the cancellous bone separately proposed by Tomaso Ver- with a radiographic guide was used during the CBCT
cellotti in 2009 [13]. e lack of quantitative analysis is scan. e exclusion criteria were as follows: (a) CBCT
the con of the two methods mentioned above. images with motion artifacts due to patient movement
Typically, cortical bone thickness and cancellous bone during the scanning process, (b) CBCT images with
density are used as quantitative measures for jawbone metal artifacts due to the presence of dental implants,
quality. Many papers have been published on the use of amalgam lling, or orthodontic appliances (e.g., bracket,
CT and CBCT in measuring cancellous bone density in archwire, and miniscrew).
jawbones [10, 14, 15]. It can be seen from the aforemen- In addition, to prevent factors such as dierent den-
tioned literature that the order of cortical bone thickness tal CBCT brands and models from aecting the repro-
and cancellous bone density diers in the four jawbone ducibility of this experiment, the QRM-MicroCT-HA
regions. phantom (QRM GmbH; Moehrendorg, Germany) was
Most of the past classication methods of jawbone also scanned to converse GV to BMD to standardize the
quality and bone quantity [8–10, 12] have assumed that research results.
better bone quality is characterized by thicker corti-
cal bone and denser cancellous bone and that worse Measurement of cancellous bone density and cortical bone
bone quality is characterized by thinner cortical bone thickness at the dental implant sites
and less dense cancellous bone. However, our previous CBCT images were input into the medical imaging soft-
study revealed that cortical bone thickness and cancel- ware Mimics 15.0 (Materialise, Leuven, Belgium) and
lous bone density are ordered dierently in the four jaw- resectioned along the dental arch to generate orthogo-
bone regions, with little correlation between them [16]. nal section images of the potential dental implant sites.
As such, this study aimed to propose a new classication All patients underwent CBCT with a dental surgical
method of bone quantity and bone quality at the implant stent before implantation. According to the radiographic
Wang et al. BMC Oral Health () : Page 3 of 13

guides (on the dental surgical stent), the referenced infor- for each level will be determined by the nal measure-
mation, such as the planned insertion site and angle, was ment results. Finally, through the permutation and
determined. e measurement was performed on a sin- combination of the crestal cortical bone thickness and
gle slice in the center of the radiographic guides, and the cancellous bone density, the jawbones are divided into a
thickness measurement was based on the instructions total of nine bone types (Fig. 2): A1, A2, A3, B1, B2, B3,
of the radiographic guide on CBCT. e corresponding C1, C2, and C3. Among them, A1 represents the jawbone
crestal cortical bone thickness was measured (Fig. 1) in with the thickest cortical bone and the densest cancellous
mm. To measure the cancellous bone density, a three- bone, while C3 represents the one with the thinnest cor-
dimensional cylinder simulated the dental implant at the tical bone and the least dense cancellous bone.
potential implant site, point by the radiographic guides,
was created according to the actual implant size (diam- Statistical analysis
eter: 3.5, 4.1, 5 mm; length: 10, 11.5 mm). e density Measurement accuracy before measuring cortical bone
of the cancellous bone inside the 3D cylinder was meas- thickness and cancellous bone density was validated.
ured on CBCT multi-slices (Fig. 1). e density of bone Intraclass correlation coecients (ICCs) were calcu-
is expressed by its grayscale value (GV). In addition, the lated to determine the reliability of the intra- and inter-
dental CBCT was also used in this study to record images examiner measurements. Twelve CBCT image data were
of phantoms with varying bone mineral densities (BMD), randomly selected from 315 implant sites to assess intra-
thereby establishing the GV/BMD conversion formula. and inter-examiner error. Two experienced examiners,
e measurement method for cortical bone thickness a dentist with 15 years of experience and a dental radia-
and cancellous bone density has been reported in previ- tion technician with ve years of experience interpreting
ous literature [10, 14, 16]. dental CBCT images, were recruited to test the intra-
observer agreement measuring the cortical bone thick-
The new classication for bone quality type at the dental ness and cancellous bone density.
implant site of the jawbone The results of this study involved descriptive statisti-
First, the cortical and cancellous bones are classied sep- cal analysis by determining the mean value, standard
arately; then, according to the measured crestal cortical deviation, and proportion of each bone type in all of
bone thickness, they are classied into A, B, and C from the samples. The data were also grouped into the fol-
thick to thin, respectively. e cancellous bone density is lowing regions: anterior maxilla, posterior maxilla,
classied into 1 (high density), 2 (intermediate density), anterior mandible, and posterior mandible. The results
and 3 (low density). Category 1 is lighter, representing were discussed from two aspects: (a) the percentages
more radioopacity, and category 3 is darker, indicating of each of the nine bone types for the whole jawbone
more radiolucency. e corresponding numerical range based on the new bone classification system; and (b)

Fig. 1 The measurement of cancellous bone density and cortical bone thickness at the dental implant site
Wang et al. BMC Oral Health () : Page 4 of 13

Fig. 2 Schematic diagram of the new bone classication; three dierent thicknesses (A, B, and C) of cortical bone and three dierent densities (1, 2,
and 3) of cancellous bone

the percentages of bone types based on the four jaw- Statistical Package for the Social Sciences (IBM Cor-
bone regions. In addition, a normal distribution analy- poration, Armonk, NY, USA) was used for all statisti-
sis of the Kolmogorov–Smirnov test for cortical bone cal analyses.
thickness and cancellous bone density was performed.
Wang et al. BMC Oral Health () : Page 5 of 13

Results thickness and cancellous bone density, the measurement

Distribution of cortical bone thickness and cancellous result of the mean value and standard deviation for each
bone density at the dental implant sites and the new bone bone type is shown in Table 1.
classication
To calculate the inter-examiner error, cortical bone thick- The percentage of nine bone types for the whole jawbone
ness, and cancellous bone mineral density in CBCT, two based on the new bone classication
examiners measured each of the factors once, and the According to the permutation and combination of dif-
ICC values were found to be 0.986 (cancellous bone) ferent types of cortical bone thicknesses and cancel-
and 0.959 (cortical bone). To account for intra-examiner lous bone densities, the analyzed jawbones were divided
error, cortical bone thickness concavity and cancellous into nine classes: A1, A2, A3, B1, B2, B3, C1, C2, and C3
bone mineral density in CBCT were measured twice by (Table 2), where the number of samples in each type and
a single examiner, and the ICC values were found to be their proportions were detailed. As seen below, most of
0.989 (cancellous bone) and 0.968 (cortical bone). ese the jawbones at the potential implant sites were of inter-
values suggest that the method’s intra- and inter-inspec- mediate quality. Among them, the most expected quality
tor errors were negligible in this study. was B2, which was observed at up to 82 sites (26.03%),
According to the results of this experiment, in the nor- followed by A2 at 41 sites (13.02%), and C2 at 45 sites
mal distribution analysis of the Kolmogorov–Smirnov (14.29%). e least common was A3 at 13 sites (4.13%). In
test, cancellous bone density showed a normal distri- addition, it can be seen from the table that bone types C1
bution (p > 0.05), and cortical bone density showed an and A3 were barely mentioned in the previous literature,
abnormal distribution (p < 0.05), as shown in Fig. 3. Based despite representing 4.44% and 4.13% of the samples col-
on the new jawbone classication system proposed in this lected in this study, respectively.
study (Fig. 4), the crestal cortical bone thickness of the Furthermore, the proportions of all four regions (ante-
jawbone was classied into three levels: A, B, and C, with rior maxilla, anterior mandible, posterior maxilla, poste-
a classication interval of A: > 1.1 mm, B: 0.7–1.1 mm, rior mandible) in each type were counted (Fig. 5). Type
and C: < 0.7 mm. Similarly, cancellous bone density was A1 was mainly found in the mandibular region of the
classied into levels 1, 2, and 3 based on the density from jawbone (30% of the anterior mandible and 59% of the
high to low. Furthermore, the GV value was converted posterior mandible), which exhibits the greatest cortical
into the absolute BMD value by the use of dental CBCT bone thickness and the highest cancellous bone density.
image calibration phantoms (Micro CT-HA phantom, Type A3 has the greatest cortical bone thickness and the
QRM GmbH, Moehrendorg, Germany). e density lowest cancellous bone density, 54% of which was found
classication intervals are 1: > 600 GV (= 420 g/cm3), 2: in the posterior mandible region. Type C1 has the lowest
300–600 GV (= 160 g/cm3–420 g/cm3), and 3: < 300 GV cortical bone thickness and the highest cancellous bone
(= 160 g/cm3). According to categories of cortical bone density, 43% of which was found in the posterior maxilla

Fig. 3 a The distribution of cortical bone thickness. (A: red, B: blue, C: purple); b The distribution of cancellous bone density (1: red, 2: blue, 3:
purple)
Wang et al. BMC Oral Health () : Page 6 of 13

Fig. 4 Example of the nine bone types based on the new bone classication

region. Type C3 represents the worst bone quality, which overall bone types were mostly of intermediate quality
was mostly found in the maxilla region, where the poste- in each region, with the exception of the posterior max-
rior maxilla accounts for 85%. illa region. Grouping the results by region shows that the
bone quality distribution in the anterior maxilla region
The percentage of bone types based on the four jawbone follows the same pattern as the overall proportion. e
regions relatively poor-quality bone types (B3, C2, and C3) in
Among the nine bone types obtained from the new bone the posterior maxilla region accounted for 59.06% of the
classication in the four jawbone regions (Fig. 6), the total, thus exhibiting a lower bone quality in comparison
Wang et al. BMC Oral Health () : Page 7 of 13

Table 1 Results of each bone type measurement in the new classication system
Mean value ± standard deviation Cortical bone thickness
A B C

Cancellous Bone Density 1 1.39 ± 0.36 0.88 ± 0.11 0.57 ± 0.07

781.85 ± 134.33 781.85 ± 134.33 781.85 ± 134.33
2 1.39 ± 0.36 0.88 ± 0.11 0.57 ± 0.07
452.71 ± 134.33 452.71 ± 134.33 452.71 ± 134.33
3 1.39 ± 0.36 0.88 ± 0.11 0.57 ± 0.07
198.79 ± 75.92 198.79 ± 75.92 198.79 ± 75.92
Data presented as mean value ± standard deviation. Note: In each bone type, the upper row represents the cortical bone thickness value (unit: mm), and the lower row
represents the cancellous bone density (unit: grayscale value, GV)

Table 2 The proportion of nine bone types for the whole with the other regions. e bone qualities in the man-
jawbone is based on the new bone classication dibular regions were generally better than those in the
maxillary regions. Most of the bone types present in the
Percent of total (Amount) Cortical bone thickness
anterior mandible region were A1, A2, and B1, and in
A B C this experiment, the C3 bone type was not found in this
Cancellous bone density 1 8.57% (27) 17.78% (56) 4.44% (14) region. However, the posterior mandible region had more
2 13.02% (41) 20.63% (65) 14.29% (45) than 80% of A1, A2, B1, and B2 bone types.
3 4.13% (13) 8.57% (27) 8.57% (27)

Fig. 5 The constructions of the four jawbone regions for each bone type are based on the new bone classication
Wang et al. BMC Oral Health () : Page 8 of 13

Fig. 6 The case number of nine bone types in the four jawbone regions

Discussion by dental CBCT for quantitative analysis so as to pro-

e purpose of this study was to propose a new type of vide a more comprehensive clinical and academic clas-
jawbone classication system. e previously held belief sication of jawbones. In addition, the various jawbone
in the classication of jawbones was that the thicker the structures and the dierent ratios of cortical and can-
cortical bone, the denser the cancellous bone. In real- cellous bones, which may be present in clinical settings,
ity, this concept cannot accommodate all of the jawbone were also simulated. Based on the measurement results,
conditions [16] that may be present in clinical settings. a new type of quantitative jawbone classication system
According to previous research, cortical bone thickness has been established to meet the dental implant patient’s
and cancellous bone density play distinct roles in main- clinical needs and to provide reference information
taining dental implant stability [17, 18]. us, when clas- before and after dental implant surgery on bone quality
sifying jawbone quality, cortical and cancellous bones classication.
must be considered separately. Over the past few years, e survival rate of dental implants depends on many
dierent combinations of jawbone thickness and density factors [19, 20], one of which is bone quality and quan-
have been proposed [13]. However, cortical bone thick- tity. ere have been many studies on the relationship
ness and cancellous bone density are not quantitatively between a patient’s jawbone quality and the success rate
measured but only conceptually categorized by previ- of dental implants [19, 21–23]. In the literature that
ously established classication systems. Given the afore- explores the success rate of dental implants according
mentioned points, this study combined crestal cortical to the dierent bone types, Jan et al. [22] followed the
bone thickness and cancellous bone density measured observation of a thousand dental implants and analyzing
Wang et al. BMC Oral Health () : Page 9 of 13

the results based on Lekholm and Zarb’s bone classica- e Type II bone classication method relies on the tac-
tion standard, they found that the failure rate for dental tile sensation when drilling bones. e technique was
implants in Bone type IV sites was up to 35% but was originally developed by Misch et al., who used the tac-
only about 3% in total for Bone type I–III sites. Jemt et al. tile sensation from logging wood to simulate the tactile
[24] also pointed out that the failure rate for implants sensation of drilling bone during dental implantation
after implantation in bone with better quality was only [12]. However, clinicians and researchers might nd this
7.9%, while that in bone with poor quality could reach description relatively abstract. It was only recently that
28.8%. Besides cancellous bone density, the cortical bone the Type III jawbone classication method was intro-
thickness at the edentulous site is another critical fac- duced. Tomaso Vercellotti suggested considering the cor-
tor aecting the initial stability of a future implant. Miy- tical bone and the cancellous bone separately, where the
amoto et al. [21] examined the initial implant stability in cortical bone thickness is clearly dened, and the can-
225 implant sites using a resonance frequency analyzer. cellous bone density is conceptually classied [13]. e
ey found that the thicker the cortical bone is, the growth thickness of cortical bone is measured at dier-
higher the initial implant stability would be. Song et al. ent time points after tooth extraction in this method of
[23] employed CBCT to assess the bone quality of the classication. Nevertheless, the amount of bone forma-
jawbone and measured the implant stability quotients tion and the growth rate of bone dier in individuals in
(ISQ value) after the implantation of dental implants. clinical settings. is method should be used with cau-
ey found that thicker cortical bone layers delivered tion when describing the bone conditions of “dierent
higher dental implant stability. implant patients at the same time point” due to the slight
Each previous jawbone classication method has its error that may occur between this classication system
own pros and cons. As previously mentioned, the Type I and the actual bone thickness measurements.
method concerns Lekholm and Zarb’s classication sys- In Lekholm and Zarb’s jawbone classication system
tem [8] and its extensions. At present, this is the most and its extensions as aforementioned, cortical bone and
commonly used classication method in dental clinics. cancellous bone are mostly considered as one unit to be
However, this method is subject to subjective judgment; classied, where it is assumed that the thicker the corti-
thus, dierent observers may arrive at dierent conclu- cal bone, the denser the inner cancellous bone. However,
sions and inuence the experimental outcomes. Numer- in Tomaso Vercellotti’s classication method, the corti-
ous scholars have therefore improved and extended the cal bone thickness and the cancellous bone density of the
classication model of Lekholm and Zarb to classify the jawbone are discussed separately, which is very dierent
bone density of each bone type quantitatively. One of the from the previous bone classication approach. In the
representative studies was conducted in 2001 by Nor- system later proposed by Al Ekrish et al. [11], Types 2 and
ton and Gamble [9], who employed CT to measure jaw- 3 are further subdivided according to the cancellous bone
bone density. Despite the use of CT and CBCT images to density from high to low, implying that cortical bone and
quantify the bone density of each bone type in the afore- cancellous bone may need to be treated separately when
mentioned methods, the thickness of the cortical bone classifying bone quality. In the previous studies focused
layer has not been considered in conjunction. Further- on the cancellous bone density at dental implant sites
more, in this type of classication concept, it is assumed [14, 15], generally, it has been noted that cancellous bone
that the cortical bone thickness and the cancellous bone density is the highest in the anterior mandible, followed
density are positively correlated. However, the results of by the anterior maxilla, posterior mandible, and posterior
the present study show that the variation trends of cor- maxilla. However, in the previous studies focused on the
tical bone thickness and cancellous bone density are not cortical bone thickness at dental implant sites [25, 27],
consistent. In comparing previous literature on cortical the posterior mandible had the thickest cortical bone,
bone thickness in the four jawbone regions [25–27] and followed by the anterior mandible, anterior maxilla, and
on cancellous bone density [10, 14, 15, 28, 29], it can be posterior maxilla. is indicates that the cancellous bone
seen that these two parameters exhibit dierent patterns. density and cortical bone thickness in the four jawbone
In addition, our team’s previous research further demon- regions. Moreover, the previous research performed at
strated that there is only a low correlation between the our laboratory also demonstrated a low correlation, or
two [16] which indicates that in jawbones, “the thickest even no correlation in some jawbone regions, between
cortical bone may be paired with the most porous can- the cortical bone thickness and the cancellous bone den-
cellous bone, and the thinnest cortical bone may also be sity [16], which suggests that the two bone structures
paired with the densest cancellous bone.” In other words, play dierent roles in maintaining the stability of dental
Lekholm and Zarb’s classication system cannot accom- implants [17, 18]. erefore, it is believed to be necessary
modate all clinically possible conditions of bone qualities.
Wang et al. BMC Oral Health () : Page 10 of 13

to consider the cortical and cancellous bones separately high-density 36.5%. eir results revealed that bones of
in the classication of jawbones. intermediate density are the most common type, followed
Cortical bone thickness in the potential dental implant by those of high density, and the low-density ones are the
sites was measured in the central cross-section of the least common. Dahiya et al. [39] also obtained similar
radiographic guides, following the measurement pro- results. ey found that in the same molar and premo-
posed by Ko et al. [27] and Gupta et al. [25]. In addition, lar regions, low-density bones made up 21% of the total,
Wang et al. used the same measurement method [16]. intermediate-density 39.5%, and high-density 39.4%; the
From the research above, the thickness of the crestal cor- average jawbone density for women was 580.2 ± 120.22
tical bone is within 2 mm (mostly between 0.7 and 1.2 GV, which was below the average of 690.5 ± 104.12 GV
mm) [16, 25, 27]. is value is signicantly lower than for men. According to the previous literature, when set-
the buccal and lingual cortical bone thicknesses (approxi- ting the CBCT measurement value of > 600 GV as high
mately 1–3 mm) [30, 31] due to the crestal cortical bone density, 300–600 GV as intermediate density, and < 300
growth after tooth extraction. erefore, the interval of GV as low density in this study, the respective propor-
thickness classication is bound to dier from that of tions in the posterior mandible are: low-density 14.01%,
other areas (buccal and lingual) to better represent the intermediate-density 46.73%, and high-density 39.25%.
actual situation in this area. Previous studies have pro- ese results are similar to those reported in the previ-
posed using 1 mm as the boundary, and the cortical bone ous literature (Fig. 7), such that in this region, the pro-
will be considered “thick” when the thickness is greater portion of intermediate-density bones is the greatest, and
than 1 mm [32]. In addition, Salimov et al. scored bone that of low-density bones is the least. According to this
quality by the tactile sensation of cortical and cancellous denition, in the whole jawbone, high density is about
bone by surgeons during dental implants. is study iden- 30%, intermediate density is 45%, and low density is 25%.
tied a thick cortical bone ≥1 mm [33]. Previous research Also, the proportion of each density level is highly con-
has pointed out that resistance in the early drilling stage sistent with the research results of Alkhader et al. [38].
is closely related to the cortical bone [34, 35]. Linck et al. In addition, it can also be found that the average jawbone
[36] classied bones according to tactile sensation during density in the posterior mandible is 535 ± 206 GV, which
bone drilling in the literature on surgeons’ tactile sensa- is slightly lower than that reported by Dahiya et al. [39]
tion and cortical bone. e results showed that the worst is is probably due to the fact that dental CBCT’s value
bones (easy to insert implants) were approximately 23%. is likely to vary depending on the brand, the equipment,
In addition, Alsaadi et al. studied the correlation between etc. Furthermore, the average age of subjects in this study
tactile and stability parameters during surgery [37]. e is 52.1 years, whereas the average age of subjects in the
results showed a high correlation between PTV and the study by Dahiya et al.is 42.5 years. Many previous studies
surgeon’s tactile sensation with the worse cortical bone in have indicated that there is a positive correlation between
the group accounting for about 23% (10/44). To prevent age and the loss of jawbones [40, 41].
inconvenience in clinical applications, this study did not In this quantitative jawbone classication system, the
make detailed divisions of the jawbone density and thick- B2 type is the most common bone type, accounting for
ness. erefore, the cancellous bone density and cortical 20.63% of the total (65 sites). A3 type (i.e., the thickest
bone thickness in the new bone classication system are cortical bone with the loosest cancellous bone) is the
only divided into three categories (Density: 1, 2, 3; ick- least observed bone type, only accounting for 4.13% of
ness: A, B, C; values from high to low). In the present the total (13 sites). In the past, classication systems gen-
study, the Class C cortical bone was about 25%, a result erally believed that the cortical bone thickness and can-
that is similar to the previous research. According to the cellous bone density should be positively correlated; that
research mentioned above, the three categories of corti- is, the thicker the cortical bone is, the denser the cancel-
cal bone thickness in this research are as follows: A, >1.1 lous bone density will be, and vice versa. is concept,
mm; B, 0.7–1.1 mm; and C, <0.7 mm. however, is only partially accurate. ere may be dier-
While setting the boundaries for high, intermediate, ent combinations of cancellous bone density and cortical
and low cancellous bone densities, this study referred bone thickness. In a paper previously published by our
to many previous publications on the jawbone densities laboratory, only a low correlation was observed between
of dental implant sites. Alkhader et al. [38] studied the jawbone thickness and density [16]. Moreover, this nd-
CBCT images of the molar and premolar regions before ing was further conrmed by the results of the current
dental implantation. e jawbones were classied as study. From the actual measurements, C1 (the thinnest
low-, intermediate-, and high-density by two experienced cortical bone with the densest cancellous bone) and A3
observers. ey found that low-density bones made (the thickest cortical bone with the most porous cancel-
up 15.6% of the total, intermediate-density 47.9%, and lous bone) can also be observed at the potential dental
Wang et al. BMC Oral Health () : Page 11 of 13

Fig. 7 The relationship between bone mineral density (BMD) and grayscale value (GV). The transfer formula between BMD and GV is also listed in
this gure

implant sites in patients. Although they are not the most Using dental CBCT GVs to indicate the density of test
prevalent jawbone qualities in clinical settings, they still objects was deemed inaccurate in some previous litera-
represent 4.44% (C1) and 4.13% (A3) of the total, respec- ture [44, 45]. According to them, even though the image
tively. In the commonly used jawbone classication sys- GVs of CBCT are linear, the absolute values can be easily
tems such as Lekhlom and Zarb, the A3 and C1 bone aected by factors such as voltage (kVp), current (mAs),
types have yet to be described, but the results of this and the dierent instrument manufacturers. In this study,
experiment suggest they account for approximately 10% the results of cancellous bone density were standardized
of all imaging samples of potential dental implant sites. from GV to BMD(g/cm3). In this manner, the image GVs
erefore, the purpose of this paper is to develop a new were converted into actual BMD values (g/cm3) (Fig. 8),
type of jawbone classication system to better meet the which are also listed in the research results, thereby pro-
clinical needs and complement the parts overlooked in viding a reference basis for future scholars to cite our
previous bone classication methods (bone quality such method. With this conversion, even with dierent brands
as A3 and C1). Additionally, it can be used in conjunction of CBCT, users of this new jawbone classication system
with postoperative tracking to analyze the initial stability will be able to convert between the two values by scan-
of dental implants in bones of dierent qualities and their ning the BMD phantom to obtain the image GV. is
long-term success rate. way, future clinical use will not encounter errors arising
In the literature related to dental implants, Liu et al. from the use of dierent CBCT machines, thereby facili-
[42] employed dental CBCT to assess jawbone density to tating future clinical applications.
determine whether this technique would be appropriate For the clinical applicability of the new method pro-
for evaluating treatment plans for dental implants. eir posed in this study, dentists should pay more attention
research results demonstrated that CBCT images could to choosing appropriate dental implants or preparing
provide valid information on jawbone density and other the host bone for the A3 (thick cortical bone and low-
bone quality characteristics, making it a very appropriate density cancellous bone) and C1 (thin cortical bone and
assessment tool before dental implant surgery. e GV high-density cancellous bone) implant positions before
obtained from dental CBCT is not a real HU; however, its dental implant surgery. Postoperative follow-up of the
working principle is also based on the linear relationship 315 implants in this study will be studied in the future.
between the radiation absorption and the object density, We hope to provide clearer recommendations to dentists
where objects of dierent densities are presented with in selecting dental implants or surgical techniques cor-
dierent GVs. In this sense, it is also suitable for evaluat- responding to the nine dierent bone states for the new
ing jawbone density. Nomura et al. [43] once pointed out classication method proposed in this study.
that although the GV of dental CBCT imaging would be As far as the limitations of this study are concerned.
higher than the HU of traditional CT imaging, these two Even though the sample size of this study is more sig-
values are still positively correlated with bone density. nicant than that of similar retrospective studies [9, 10,
Wang et al. BMC Oral Health () : Page 12 of 13

Fig. 8 Percentage of low‑, intermediate‑ and high‑density in the posterior mandible region compared with previous studies [42, 43]

14, 15, 28], in the future, if it is necessary to conduct drafting article, funding secured/ LJF: Data acquisition, drafting article/ HLH:
Drafting article/ SLP: Drafting article/ MTT: Drafting article. The author(s) read
subgrouping studies based on dierent sexes, ages, and and approved the nal manuscript.
implant sites, more dental implant samples should be col-
lected to represent the entire population, in addition, this Funding
This study was supported by the Ministry of Science and Technology, Taiwan
study only discussed the dental CBCT images of patients (MOST 110‑2221‑E‑039‑005).
with dental implants before the surgery and did not fol-
low up on these patients after the dental implantation. In Availability of data and materials
The datasets used and/or analysed during the current study are available from
the future, the survival rate of these dental implants in the corresponding author on reasonable request.
these patients should be tracked, and its correlation with
the bone type should be explored. Declarations
Ethics approval and consent to participate
This study was approved by the Institutional Review Board of China Medical
Conclusions University Hospital, No. CMUH 108‑REC2‑181. Informed consent was waived
In conclusion, the proposed bone classication is a new by CMUH 108‑REC2‑181 owing to the retrospective nature of the study. All the
quantitative classication system of jawbone quality and methods were performed in accordance with Declaration of Helsinki.
quantity at the dental implant site developed based on Consent for publication
dental CBCT. In this system, the bone quality and bone Not applicable.
quantity of the jawbone are classied into nine bone
Competing interests
types. at is, the crestal cortical bone thickness is clas- The authors declare no competing interests.
sied into A: > 1.1 mm, B: 0.7–1.1 mm, and C: < 0.7 mm,
and the cancellous bone density is classied into 1: > 600
Received: 21 January 2023 Accepted: 13 May 2023
GV (= 420 g/cm3), 2: 300–600 GV (= 160–420 g/cm3),
and 3: < 300 GV (= 160 g/cm3). e proposed classica-
tion system veried that nine possible types of bone were
found in all maxilla and mandible regions.
References
Acknowledgements 1. Hong DGK, Oh JH. Recent advances in dental implants. Maxillofac Plast
Not applicable Reconstr Surg. 2017;39(1):33.
2. Raikar S, Talukdar P, Kumari S, Panda SK, Oommen VM, Prasad A. Factors
Authors’ contributions aecting the survival rate of dental implants: a retrospective study. J Int
SHW: Concept design, data acquisition, data analysis, data interpretation, Soc Prev Community Dent. 2017;7(6):351.
drafting article/ JTH: Concept design, data acquisition, data interpretation,
Wang et al. BMC Oral Health () : Page 13 of 13

3. Kandasamy B, Kaur N, Tomar G, Bharadwaj A, Manual L, Chauhan M. Long‑ dental implant placement? A cohort study. Int J Environ Res Public
term retrospective study based on implant success rate in patients with Health. 2021;18(8):4284.
risk factor: 15‑year follow‑up. J Contemp Dent Pract. 2018;19(1):90–3. 27. Ko YC, Huang HL, Shen YW, Cai JY, Fuh LJ, Hsu JT. Variations in crestal
4. Chrcanovic BR, Albrektsson T, Wennerberg A. Bone quality and quantity cortical bone thickness at dental implant sites in dierent regions of the
and dental implant failure: a systematic review and meta‑analysis. Int J jawbone. Clin Implant Dent Relat Res. 2017;19(3):440–6.
Prosthodont. 2017;30(3):219–37. 28. Fuh LJ, Huang HL, Chen CS, Fu KL, Shen YW, Tu MG, Shen WC, Hsu JT. Vari‑
5. Hsu Y‑Y, Tsai M‑T, Huang H‑L, Fuh L‑J, Hsu J‑T. Insertion speed aects the ations in bone density at dental implant sites in dierent regions of the
initial stability of dental implants. J Med Biol Eng. 2022;42(4):516–25. jawbone. J Oral Rehabil. 2010;37(5):346–51.
6. Cheng K‑C, Liu P‑H, Chen H‑S, Lan T‑H. Stress distribution of four‑unit 29. Yu J‑H, Huang H‑L, Liu C‑F, Wu J, Li Y‑F, Tsai M‑T, Hsu J‑T. Does orthodontic
implant‑supported xed partial prosthesis with dierent numbers and treatment aect the alveolar bone density? Medicine. 2016;95(10):e3080.
positions of xtures in maxilla anterior Region‑3D FEA. J Med Biol Eng. 30. Katranji A, Misch K, Wang HL. Cortical bone thickness in dentate and
2022;42(4):526–33. edentulous human cadavers. J Periodontol. 2007;78(5):874–8.
7. Esposito M, Hirsch J‑M, Lekholm U, Thomsen P. Biological factors contrib‑ 31. Ge J, Zheng J‑W, Yang C, Qian W‑T. Variations in the buccal‑lingual alveo‑
uting to failures of osseointegrated oral implants. (II) Etiopathogenesis. lar bone thickness of impacted mandibular third molar: our classication
Eur J Oral Sci. 1998;106(3):721. and treatment perspectives. Sci Rep. 2016;6(1):16375.
8. Lekholm U. Tissue integrated prosthesis: Patient selection and prepara‑ 32. Huang HM, Lee SY, Yeh CY, Lin CT. Resonance frequency assessment
tion In: Brånemark PI, Zarb GA, Albrektsson T, eds. 1985. p. 199–209. of dental implant stability with various bone qualities: a numerical
9. Norton MR, Gamble CJ. Bone classication: an objective scale of bone approach. Clin Oral Implants Res. 2002;13(1):65–74.
density using the computerized tomography scan. Clin Oral Implants Res. 33. Salimov F, Tatli U, Kürkçü M, Akoğlan M, Öztunç H, Kurtoğlu C. Evaluation
2001;12(1):79–84. of relationship between preoperative bone density values derived from
10. Hao Y, Zhao W, Wang Y, Yu J, Zou D. Assessments of jaw bone density cone beam computed tomography and implant stability parameters: a
at implant sites using 3D cone‑beam computed tomography. Group. clinical study. Clin Oral Implants Res. 2014;25(9):1016–21.
2014;1:D1. 34. Holmes DC, Loftus JT. Inuence of bone quality on stress distribution for
11. Al‑Ekrish AA, Widmann G, Alfadda SA. Revised, computed tomography‑ endosseous implants. J Oral Implantol. 1997;23(3):104–11.
based Lekholm and Zarb Jawbone quality classication. Int J Prostho‑ 35. Di Stefano DA, Arosio P, Capparè P, Barbon S, Gherlone EF. Stability of den‑
dont. 2018;31(4):342–5. tal implants and thickness of cortical bone: clinical research and future
12. Misch CJ. Bone classication, training keys to implant success. Dent perspectives A systematic review. Materials (Basel). 2021;14(23):7183.
Today. 1989;8(4):39. 36. Linck GKSB, Ferreira GM, De Oliveira RCG, Lindh C, Leles CR, Ribeiro‑Rotta
13. Vercellotti T. Essentials in piezosurgery: Clinical advantages in dentistry: RF. The inuence of tactile perception on classication of bone tissue at
Quintessence; 2009. dental implant insertion. Clin Implant Dent Relat Res. 2016;18(3):601–8.
14. Shapurian T, Damoulis PD, Reiser GM, Grin TJ, Rand WM. Quantitative 37. Alsaadi G, Quirynen M, Michiels K, Jacobs R, Van Steenberghe D. A bio‑
evaluation of bone density using the Hounseld index. Int J Oral Maxil‑ mechanical assessment of the relation between the oral implant stability
lofac Implants. 2006;21(2):290–7. at insertion and subjective bone quality assessment. J Clin Periodontol.
15. de Oliveira RCG, Leles CR, Normanha LM, Lindh C, Ribeiro‑Rotta RF. 2007;34(4):359–66.
Assessments of trabecular bone density at implant sites on CT images. 38. Alkhader M, Hudieb M, Khader Y. Predictability of bone density at poste‑
Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2008;105(2):231–8. rior mandibular implant sites using cone‑beam computed tomography
16. Wang S‑H, Shen Y‑W, Fuh L‑J, Peng S‑L, Tsai M‑T, Huang H‑L, Hsu J‑T. Rela‑ intensity values. Eur J Dent. 2017;11(03):311–6.
tionship between cortical bone thickness and cancellous bone density at 39. Dahiya K, Kumar N, Bajaj P, Sharma A, Sikka R, Dahiya S. Qualitative assess‑
dental implant sites in the jawbone. Diagnostics. 2020;10(9):710. ment of reliability of cone‑beam computed tomography in evaluating
17. Rozé J, Babu S, Saarzadeh A, Gayet‑Delacroix M, Hoornaert A, Layrolle bone density at posterior mandibular implant site. J Contemp Dent Pract.
PJ. Correlating implant stability to bone structure. Clin Oral Implants Res. 2018;19(4):426–30.
2009;20(10):1140–5. 40. Shen X, Chen X, Huang J, Xu R, Cheng J, Jiang H. Age‑dependent role
18. Isoda K, Ayukawa Y, Tsukiyama Y, Sogo M, Matsushita Y, Koyano KJ. Rela‑ of SIRT6 in jawbone via regulating senescence and autophagy of bone
tionship between the bone density estimated by cone‑beam computed marrow stromal cells. J Mol Histol. 2020;51:67–76.
tomography and the primary stability of dental implants. Clin Oral 41. Ko Y‑C, Tsai M‑T, Fuh L‑J, Tsai M‑J, Wang X‑H, Huang H‑L, Hsu J‑T. Associa‑
Implants Res. 2012;23(7):832–6. tion between age of menopause and thickness of crestal cortical bone
19. Pitman J, Christiaens V, Cosyn J, Glibert M. Primary stability of con‑ at dental implant site: a cross‑sectional observational study. IInt J Environ
ventionally tapered versus reverse tapered body shift implants under Res Public Health. 2020;17(16):5868.
varying bone support conditions—an in‑vitro study. J Med Biol Eng. 42. Liu J, Chen H‑Y, DoDo H, Yousef H, Firestone AR, Chaudhry J, Johnston
2022;42(4):429–35. WM, Lee DJ, Emam HA, Kim D‑G. Ecacy of cone‑beam computed
20. Huang H‑L, Lin T‑W, Tsai H‑L, Wu Y‑L, Wu AY‑J. Biomechanical eects of tomography in evaluating bone quality for optimum implant treatment
bone atrophy, implant design, and vertical or tilted of posterior implant planning. Implant Dent. 2017;26(3):405–11.
on all‑on‑four concept implantation: nite element analysis. J Med Biol 43. Nomura Y, Watanabe H, Honda E, Kurabayashi TJ. Reliability of voxel val‑
Eng. 2022;42(4):488–97. ues from cone‑beam computed tomography for dental use in evaluating
21. Miyamoto I, Tsuboi Y, Wada E, Suwa H, Iizuka T. Inuence of cortical bone mineral density. Clin Oral Implants Res. 2010;21(5):558–62.
bone thickness and implant length on implant stability at the time of 44. Silva IMdCC, Freitas DQd, Ambrosano GMB, Bóscolo FN, Almeida SM.
surgery—clinical, prospective, biomechanical, and imaging study. Bone. Bone density: comparative evaluation of Hounseld units in multislice
2005;37(6):776–80. and cone‑beam computed tomography. Braz Oral Res. 2012;26:550–6.
22. Jan RA, Berman CL. The excessive loss of Branemark xtures in type IV 45. Varshowsaz M, Goorang S, Ehsani S, Azizi Z, Rahimian S. Comparison of
bone: a 5‑year analysis. J Periodontol. 1991;62(1):2–4. tissue density in Hounseld units in computed tomography and cone
23. Song Y‑D, Jun S‑H, Kwon J‑JJ, Implants M. Correlation between bone beam computed tomography. J Dent (Tehran). 2016;13(2):108.
quality evaluated by cone‑beam computerized tomography and implant
primary stability. Int J Oral Maxillofac Implants. 2009;24(1):59–64.
24. Jemt T, Lekholm U. Implant treatment in edentulous maxillae: a 5‑year Publisher’s Note
follow‑up report on patients with dierent degrees of jaw resorption. Int Springer Nature remains neutral with regard to jurisdictional claims in pub‑
J Oral Maxillofac Implants. 1995;10(3):303–11. lished maps and institutional aliations.
25. Gupta A, Rathee S, Agarwal J, Pachar RB. Measurement of crestal cortical
bone thickness at implant site: a cone beam computed tomography
study. J Contemp Dent Pract. 2017;18(9):785–9.
26. Wang S‑H, Ko Y‑C, Tsai M‑T, Fuh L‑J, Huang H‑L, Shen Y‑W, Hsu J‑T. Can
male patient’s age aect the cortical bone thickness of jawbone for