Dr. Eva-Maria Hempe

Great article from the superstar Harvard Business School team led by Karim Lakhani on how Generative AI is reshaping cultures and ways of working by lowering the expertise threshold in large organizations! This article is based on a multiple recent studies and both confirms trends we've seen before and reveals hidden patterns that every Leadership team should pay attention to. Here are my key take-aways: Confirmed Trends - Efficiency: Individuals using AI significantly cut down task time and teams matched or exceeded performance vs. control group without AI, especially in cross-functional tasks. - Efficacy: Teams with AI produced top-quality solutions at a higher rate than other groups, showing that human collaboration combined with AI generates increased performance. - Business Integration: AI usage enabled participants to develop solutions outside their expertise, e.g. technical and commercial. AI can help break down silos with broader problem-solving across functions. Hidden patterns - Team dynamics: AI usage reshapes teamwork by flattening hierarchy, especially in knowledge intensive teams as it helps balance contributions from team members with varying degrees of expertise. - Emotional Impact: AI users reported increased positive emotions (excitement, energy) and reduced negative ones (anxiety, frustration). This indicates that AI can significantly enhance the work experience. - Cognitive Dissonance: Despite objectively better performance, AI-assisted participants had lower confidence in their outcomes, revealing a disconnect between reality and perception in AI-augmented work. This is the only clear negative outcome of the study. Potential causes may be culture related, and we should stay ahead with learning and communication: 1. Less effort is tied to lesser outcomes, in spite of precedents such as industrial automation, computing, etc. 2. Work done with AI is (still) felt as "cheating", and therefore taints the experience as a fraud, undermining confidence 3. The outcome feels less personal and relatable. This could be a good thing too if it leads to more fact-based discussions in decision making vs. strong attachment to one's own ideas and preconceptions. As culture influencers and leaders, our focus should be on how we can start shifting the cultural norm regarding AI augmented work. This reminds me of online dating in the early 2000's! 🙄 Time to move on... https://lnkd.in/eyG6zCQk

59

4 Kommentare

Challenges of Using generative AI in medicine: 1. Data Quality and Availability: High-quality, large-scale datasets are essential for training generative models, but such data can be scarce, incomplete, or biased in the medical field. 2. Privacy and Security: Handling sensitive patient data requires stringent privacy measures. Ensuring that generative AI models comply with regulations like HIPAA is a significant challenge. 3. Regulatory Compliance: Generative AI applications in medicine must adhere to strict regulatory standards set by authorities such as the FDA, which can slow down development and deployment. 4. Bias and Fairness: Generative models can inadvertently learn and propagate biases in the training data, leading to unfair or inaccurate outcomes. Interpretability and Transparency: Medical professionals must understand how AI models make decisions. Many generative models, like deep learning networks, are often considered "black boxes." 5. Validation and Testing: It is crucial but challenging to ensure that generative AI models are thoroughly validated and tested in diverse, real-world scenarios. Integration with Existing Systems: Integrating generative AI solutions with existing healthcare IT systems (like EHRs) can be complex and costly. 6. Ethical Concerns: Using generative AI in medicine raises various ethical issues, including the potential for misuse, the impact on doctor-patient relationships, and the implications of AI-generated content. 7. User Trust and Acceptance: Building trust among healthcare providers and patients in AI-generated solutions is essential for widespread adoption. 8. Cost: Developing, implementing, and maintaining generative AI solutions can be expensive, particularly for smaller healthcare providers. 9. Scalability: Scaling generative AI solutions across different healthcare settings and geographies can be challenging due to variability in healthcare practices and infrastructure. 10. Responsibility and Accountability: Determining who is responsible and accountable for AI-generated decisions or errors is a complex issue. 11. Training and Expertise: Healthcare professionals need proper training to effectively use and understand generative AI tools. 12. Real-Time Processing: Generative AI models often require substantial computational resources, hindering real-time clinical applications. 13. Generalizability: It is crucial to ensure that generative AI models generalize well to different patient populations and medical conditions. 14. Data Annotation and Labeling: Annotating medical data for training generative models is labor-intensive and requires expert knowledge. 15. Patient Consent: Obtaining informed consent from patients for using their data in training generative AI models can be complicated.

8

Adoption is the final step...

2

1 Kommentar

📢 New colleague or gimmick hurdle? A user-centric review of the barriers and facilitators of robots (or #healthtech / AI in general) in hospitals Healthcare systems are under immense pressure—facing rising costs, personnel shortages, ageing populations, and a surge in chronic diseases and global health challenges. How can we keep up with these demands while improving care quality? Enter robotic technologies and artificial intelligence (AI). These tools offer immense potential to enhance efficiency, reduce costs, and improve working conditions in healthcare. But their integration into hospitals isn't without challenges. A recent review of 501 studies distilled insights from 40 in-depth, user-focused papers, highlighting key barriers and facilitators in implementing robots in hospitals: 🚧 Barriers: Organisational misalignments Resistance to change Concerns about privacy, confidentiality, and anthropomorphisation 🌟 Facilitators: Improved workflows Reduced workload for staff Enhanced patient outcomes The takeaway? The road to robotic healthcare isn't just about advancing tech; it's about redesigning hospital systems, workflows, and user adoption. To ensure success, future research must explore the delicate interplay between hospital employees, institutions, and emerging robotic technologies. 💡 A deeper understanding of these dynamics could pave the way for innovative healthcare solutions that are more accessible, efficient, and human-centric. Here is the great insights and read by Mathias Kofoed Rasmussen, Anna Schneider-Kamp, Tobias Hyrup, Alessandro Godono for PLOS: https://lnkd.in/e4cSFExi #HealthcareInnovation #AI #Robotics #FutureOfCare #DigitalHealth #HealthcareFuturist #digitalhealth #changemanagement

16

✍ Crafting an Effective AI Council Charter in Healthcare & Life Sciences Continuing our ongoing discussions on AI Council for GenAI acceleration (post 1 - https://lnkd.in/eW2QhTKR; post 2 - https://lnkd.in/egbJmTps), let’s dive deeper into what might be literally the first task for AI Council i.e. establishing and maintaining a well-defined AI Council Charter. A well-crafted charter not only mitigates risks but also drives responsible AI innovation. The healthcare and life sciences sector faces unique challenges, including bias in AI algorithms, data privacy concerns, and the complexities of regulatory compliance. Addressing these challenges, as part of their charter, is critical for the success of the AI Council. Drafting a charter as the AI Council's first deliverable sets the tone for expectations and its effectiveness in leading GenAI adoption. This process may benefit from engaging a partner or consultant for a smooth experience. Below, I am sharing my mental model on charter development, giving insight into what it takes to deliver. As a new AI Council of diverse internal leaders, expect to go through norming, storming, and forming stages while drafting a charter. For a common understanding, I recommend members review key documents from government and NGO entities, such as FDA's AI & Medical Products white paper, White House's AI Bill of Rights, WHO's AI health regulations, and EU's AI Act. Give members ample time to familiarize themselves with these materials. Next, invite them for collaborative charter-defining workshops, ideally offsite to avoid distractions. During this 2-4 day event, use design thinking to gather inputs on purpose, scope, goals, challenges, roles, responsibilities, deliverables, KPIs, structure, and alignment with medical/scientific rigor. This iterative process should allow time for idea generation, sharing, grouping, and prioritizing. Using prioritized ideas, facilitate a session for participants to collaboratively draft the charter, combining key elements identified. Define the team's Purpose, Goals, Roles, Responsibilities, KPIs, Deliverables, and Structure. Allow ample time to ensure alignment and commitment from your cross-functional AI Council. Next, seek council feedback on clear, measurable KPIs (e.g., AI implementations, compliance rates) to gauge effectiveness. Align on audit cadence and PMO support. Consider additional measures: communication strategy, surveys, ethical incident analysis, impact assessments, and benchmark comparisons. These ensure the charter evolves with the AI landscape, remaining a living document. 💬 Have you defined your AI Council's charter? 📢 Subscribe to my newsletter: https://lnkd.in/g3bdneR7 #genai #ai #aws #AIinHealthcare #ResponsibleAI #LifeSciences

15

🔍⚖️ Practitioners Guide To Responsible AI - Part 3 Building on our previous discussions (https://lnkd.in/ekchutB4), today we delve deeper into the first layer of our Responsible AI framework: Model Training & Data Security. In the healthcare and life sciences sectors, the integrity and security of data are paramount. This layer addresses key concerns such as: 1/ Data Privacy: Ensuring patient data is protected and compliant with regulations. 2/ Bias Mitigation: Implementing comprehensive strategies to prevent bias in AI models, ensuring equitable outcomes across different demographics. 3/ Data Integrity: Maintaining the accuracy and consistency of data throughout its lifecycle. and 4/ Secure Data Handling: Safeguarding data during storage and transfer to prevent unauthorized access. So, what does this mean in practice? Let’s look at some examples: If your organization handles unique patient data, it's crucial to ensure that your AI models do not inadvertently favor specific races or genders. But when you want to do this at scale, you will need integrated tools to continuously validate, test, and audit both the data and the model through their lifecycle. Another example, when sharing data across lines of business, the data and its metadata should be self-contained and easily interpretable to prevent misinterpretation. Again, you will need tools to do this at scale – accounting for discovery, authentication and authorization. Well, it's easier said than done! How would one go about supporting these responsible AI practices at scale? AWS, for example, offers a range of integrated tools that allow for organizations to manage these issues at scale. Amazon SageMaker Clarify helps detect and mitigate bias during data preparation, model training, and deployment, providing insights into model predictions for greater transparency. Amazon DataZone facilitates governed data sharing and metadata management, ensuring data privacy and integrity across business lines. Additionally, AWS Identity and Access Management (IAM) and AWS Key Management Service (KMS) ensure secure data handling by managing access and encryption. These tools collectively enable healthcare and life sciences organizations to implement robust data governance and responsible AI practices effectively. In my next post, I will dive deeper and share insights on the next layer of our Responsible AI framework: Model Usage, Infrastructure, and Deployment. 💬 How are you integrating Responsible AI practices and tools into your operations at scale? 📢 Subscribe to my newsletter to get access to strategies and practical guidance on accelerating adoption of generative AI within your organization. Get started here: https://lnkd.in/g3bdneR7 #genai #ai #aws #ResponsibleAI #AIinLifeSciences #HealthcareAI #AIEthics #AIGovernance #AISecurity

16

Network, I need your help! I will be at the health.tech conference in Munich on June 5-6. In the meanwhile, I’m working on some interesting discussion points related to how AI can help improve laboratory and healthcare systems performance. Which of the following topics resonate with you most? I’d love to hear your thoughts: ▶ Diagnostic speed and accuracy ▶ Enhancing patient outcomes ▶ Improving diagnosis and treatment suggestions ▶ Optimizing operations & workflow management ▶ Treatment personalization ▶ Accelerating R&D processes, innovation ▶ Predicting equipment failures & error detection ▶ Lab protocols automation Are there any other topics that come to mind? If you're attending, I’d love to share the best practices we've seen success with for our clients. Shoot me a message or leave a comment, and let’s meet up at the conference. Otherwise, feel free to schedule a time slot here: https://bit.ly/3QL2Nxl And what if you’re not attending, but would still like to discuss best practices? Feel free to DM me, and let’s connect!

20

2 Kommentare

I’m incredibly proud to see cutting-edge AI research from my hometown in Dresden showcased by AI at Meta. 🧠 🔎 This powerful use case, built with #Llama by the team around Isabella Wiest and Jakob Nikolas Kather at Else Kröner Fresenius Center for Digital Health, is driving critical advances in precision oncology and beyond. In simpler terms, the researchers are using #AI to anonymize and sift through enormous amounts of patient data, helping doctors pinpoint the most effective treatments for cancer patients while safeguarding their privacy. This makes healthcare more personalized and ensures life-saving therapies reach the right people, faster. 🩺 💊 Stories like these underscore how vital #OpenSourceAI is for fueling innovation, societal progress, and economic growth across Europe. Moreover, it highlights the importance of fostering a strategic #interdependence with the U.S. in AI development, ensuring we learn from each other, share resources, and jointly push the boundaries of what’s possible. 🧩📈 I’m excited to see Dresden’s contribution to this global effort—and can’t wait to watch how open collaboration continues to unlock new frontiers in AI and healthcare.

21

1 Kommentar

Raindrops are a great metaphor for the #digitaltransformation of Healthcare. What do I mean? The promises of the digital health revolution are many. Decreased workload, improved diagnostics, improved efficiency, greater access, enabling technologies etc. Furthermore, their wide adoption seems imminent! Not only does it make sense, but in every other industry, those same technologies are in place and delivering on their promise. In fact, the digial revolution should hit healthcare like a tsunami and allow us to start over on some version of the medical bay on the Starship Enterprise. It is a cruel experiment: your car drives you to work but when you get there, you are faxing patient referrals. A patient can book an international flight but when they try to book a clinic appointment, the task is insanely complicated. But not everywhere. At Kaiser Permanenet in NCAL, it is faster to book an online appointment than to do it on the phone (I did it). Some patients have the benefit of a sci-fi avatar driven all knowing virtual assistant to help them recover from surgery. A couple of posts back, I talked about the speed at which smaller hospital groups are able to adopt new technology efficiently. Unburdened by legacy tech stacks and outdated leadership models, they can actually identify pain points for which there are mature digital solutions and implement them. Done. Each of these smaller groups, or digitally enabled private practices, represents a rain drop of technological adoption. Over time, adjacent raindrops will coalesce, competition will lead to curiosity, curiosity will lead to familiarity, familiarity to trust, and trust to broad adoption. In the 1990s, many proclaimed that sequencing the human genome would usher in a wave of cancer cures within a decade. Nearly 40 years later, we are starting to see some raindrops. Today, it's AI. We hoped that a computer science tsunami would wipe the healthcare slate clean, but instead, we are getting raindrops of success and people are either surprised or disappointed. We should not be. For each raindrop, in that location, the job is done. Eventually, others join in. The network effect grows. And it grows quickly. So: go ahead, be a raindrop. Lead the pack. Forge the way where it makes sense. You won’t be alone for long. What do you think? Is this a good visual to explain how innovation spreads in healthcare and why it pays to be an innovator? Mohan Nair Logan Plaster Daniel Kraft, MD John Nosta Professor Shafi Ahmed Fabrizio Billi Gil Bashe Mark Froimson MD, MBA Ivan Tornos Todd S. Eury Douglas Fairbanks Peter Birch Ali Hashemi Saul Marquez Michael Havig, MD Rafael J. Grossmann, MD, MSHS, FACS Claudio Díaz Ledezma Daniel Couchman Kendall MPH Sanjeev Agrawal

34

9 Kommentare

This paper discusses the pervasive issue of burnout among healthcare workers and proposes strategies for its mitigation. 1️⃣ Burnout is a global issue in healthcare, exacerbated by increasing patient demands, complex conditions, and administrative burdens. 2️⃣ European healthcare systems face challenges in replacing retiring professionals, leading to increased workloads for remaining staff. 3️⃣ Self-reporting studies often underestimate the real-life burnout levels among healthcare professionals, indicating a more severe issue than literature suggests. 4️⃣ Burnout can transfer to the next generation of healthcare workers, as burned-out professionals may pass on negative attitudes and emotional exhaustion to trainees. 5️⃣ Expanding competencies and sharing tasks across healthcare professions can help manage workloads and delay burnout onset. 6️⃣ AI can aid in reducing routine tasks, but its implementation must be carefully managed with quality education and early engagement of healthcare workers. 7️⃣ Mitigating burnout requires a supportive work environment, teamwork, emotional support resources, and resilience training. 8️⃣ Insufficient appreciation and remuneration, coupled with high administrative burdens, contribute significantly to burnout. Delegating administrative tasks to assistants can enhance job satisfaction. 9️⃣ Developing appropriate work-life balance strategies and building personal resilience are crucial for improving job performance and overall well-being of healthcare professionals. ✍🏻 David Sipos, Rahul Goyal, Tomas Zapata. Addressing burnout in the healthcare workforce. The Lancet Regional Health - Europe. 2024. DOI: 10.1016/j.lanepe.2024.100961

105

6 Kommentare

🔍 Integrating LLM as a Judge in Your RAG Workflow Building upon our exploration of Enterprise RAG architecture and design best practices from two weeks ago (https://lnkd.in/eSggTNyE), and expanding on our examination of evaluation-driven development from last week (https://lnkd.in/eAsiprjH), we'll continue to delving further into the concept of LLM as a Judge. In my earlier post this week, we explored the function of Large Language Models (LLMs) as evaluators and how your specialized teams can contribute to training the LLM Judge (https://lnkd.in/eVz2i_4n). In today's discussion, we'll focus on how to integrate the trained LLM as a Judge in your RAG workflow. We will continue to leverage the Payer specific domain examples to help explain core concepts. Here's a step-by-step guide to integrating an LLM judge: 1/ RAG Response Generation: Generate response from user query and context. 2/ Prepare Evaluation Input: Compile question, response, and context into structured format. 3/ Domain-specific LLM Judge prompt: Use appropriate prompt for evaluation inference. 4/ LLM Judge Evaluation: Submit prepared input with specific evaluation prompt. 5/ Interpret Judge's Output: Analyze assessment ("Correct", "Incorrect", "Unclear"). 6/ Action Based on Evaluation: Handle outputs:If "Correct": Deliver to user. If "Incorrect"/"Unclear": Trigger review or fallback. 7/ Feedback Loop: Store evaluations for continuous improvement of RAG and judge models. Let's walk through this process using a healthcare payer example: User question: "What is the copay for a specialist visit under the Gold Plan?" 1/ RAG response: "Under the Gold Plan, the copay for a specialist visit is $40." 2/ Evaluation input: QUESTION: "What is the copay for a specialist visit under the Gold Plan?" RESPONSE: "Under the Gold Plan, the copay for a specialist visit is $40." CONTEXT: "Gold Plan specialist visits have a $40 copay as of January 1, 2024." 3/ LLM judge prompt: "Given the QUESTION about health insurance, is the RESPONSE correct based on the CONTEXT? Return 'Correct' or 'Incorrect'." 4/ LLM judge evaluation: Make inference call. 5/ Judge's Output: "Correct". 6/ Action: Approve response. 7/ Feedback: If "Incorrect" or "Unclear", trigger human review or use fallback response. By integrating an LLM judge into your RAG workflow, you create a powerful system that combines the efficiency of AI with the reliability of expert-guided evaluation. 💬 How are you planning to integrate AI-driven evaluation in your RAG systems? ♻️ Subscribe to my newsletter & repost if you find value in these insights: https://lnkd.in/g3bdneR7 #enterpriserag #evaluationtechniques #aievaluation #genai #datascience #machinelearning #aistrategy #cto #cdo #aicouncil #aws #enterpriseai #aiadoption #digitaltransformation #healthcarepayers #healthcareai #insurtech

19

2 Kommentare

About 5 or 6 years ago, Beth Friedmann and I had a long conversation about developing #digitaltwins of patients. It was her firm belief that one day, it would be a useful tool for caring for patients. Now that we have the technology to support digital twins, we have the opportunity to: - Develop new ways to deliver precision medicine - Accelerate drug discovery - Build hyper-personalized care pathways #personalizedmedicine #patientexperience #patientengagement #patientcare #providers #AI #AIinHealthcare #HealthcareonLinkedin

63

4 Kommentare

DataRobot demonstrates #ai #predictivemodeling #value across Germany's Universitätsmedizin Mannheim #healthcare hospital, by leading #patientcare #innovation for the 250,000 patients, triaged across 21 clinics annually. This is leading to an improved: Quality of Life (QoL) improvement for 30% of patients. https://lnkd.in/gZ_qYeEb #timeseries #visualai #regulation #governance

16

𝗛𝗼𝘄 𝗦𝗵𝗼𝘂𝗹𝗱 𝗪𝗲 𝗥𝗲𝗴𝘂𝗹𝗮𝘁𝗲 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗶𝗻 𝗛𝗲𝗮𝗹𝘁𝗵𝗰𝗮𝗿𝗲? 🏥💡 As generative AI technologies like #ChatGPT and others rapidly advance, regulators face a unique challenge in adapting traditional frameworks for healthcare. Current FDA regulations, designed for specific drugs and devices, fall short when applied to AI models with vast and evolving capabilities. A novel approach is needed — one that treats these AI systems more like human clinicians, requiring rigorous training, supervision, and periodic reassessment. While challenges such as AI "hallucinations" and transparency in training data remain, a thoughtful regulatory framework can ensure these technologies safely enhance clinical practice. I was recently invited to deliver three keynotes during the 82nd International Pharmaceutical Federation (FIP) World Congress of Pharmacy and Pharmaceutical Sciences 2024 in Cape Town, South Africa from 1-4 September. During these sessions, I emphasized that we need a completely new approach and framework to regulate AI in healthcare — fully aligned with these insights. The future of #Healthcare and #HealthTech depends on it. 🚀 Special thanks to my friend Prinesh Reddy for sharing that article with me 🙏 #AIinHealthcare #GenerativeAI #DigitalHealth #Innovation #FIP2024 #FIPCongress Lars-Åke Söderlund (Exec MBA) Vice President Harvey Castro, MD, MBA. Professor Shafi Ahmed Dylan Attard MD, MRCSI, MEnt. Alex Diaconu Dr. Martha Boeckenfeld Dr. Khulood Alsayegh Dr. Mariam Mohamed Fatima Matar MD Phd Dr. Meshari F. Alwashmi Fady Hannah-Shmouni, MD FRCPC Xiaochen Z. Ricardo Moraes Virgilio Gibbon Ana Araujo Hassan Chaudhury HonFAPM Vijay K. Luthra Pilar Fernandez Hermida Jijo James Rémy LEVASTRE

26

6 Kommentare

What are the unique challenges of building digital platforms in healthcare and how can ecosystem parties collaborate toward a common good? I was fortunate to explore this question with Frank Konopka, Principal Key Expert for Digital Business Model Operations at Siemens Healthineers in the latest episode of #TheDigitalPodcast. Frank offers a deep dive into the competitive tensions between upstream and downstream platforms, the challenges of IP rights protection in relation to raw and annotated data, the role of regulation and the need for democratization through the use of open source software. In the short clip below (https://lnkd.in/dJUxczsq), Frank explains that while upstream platforms provide improved services (e.g. AI workflow systems) to existing medical devices such as MRI scanners, downstream platforms provide direct services to patients. Med tech companies are careful not to enter the downstream arena as this creates value capture tensions between their customers (e.g. the hospitals) and their customers' customers (e.g. the patients). Big tech companies are however more aggressive in entering the downstream arena as this is where there are strong economies of scale. 👉 Listen to the full episode on YouTube: https://lnkd.in/d7ri6Ja7 👉 and on Spotify: https://lnkd.in/dS2rjZQE

50

1 Kommentar

Transcribing conversations between doctors and patients is the biggest application for AI at Kaiser Permanente, the world’s leading hospital, according to Daniel Yang, and that hallucinations are a problem. He doesn’t mention radiology or other applications that were supposed to be completely automated by now. “One of the most exciting areas for AI and #technology is in supporting our clinicians to manage some of the administrative burden that they face on a day-to-day basis. Kaiser Permanente, to the best of my knowledge, is deploying the largest implementation of the clinical AI scribe technology in the country. The purpose is to generate a first draft of the clinical note from a recording of the patient’s encounter.” The WSJ asks if hallucinations are a problem? “we make sure that our clinicians are reviewing every note.” “We were hearing things like, it would write in the note, come back and see me in two weeks. But the doctor said, I never said that. It’s not an unreasonable thing, but it’s not something that we talked about in the visit.” The WSJ asks: “What has been the biggest failure with AI? You’ve got all these eager AI developers that get their hands on a data set and ask themselves, what AI tool can I develop from this data set, instead of starting with the problem and then finding the data and developing the tool.” “And two is, I think a lot of people fail to realize that developing the #algorithm is really the easy part. The part that really takes work, and the part that adds value, is redesigning the workflow to accommodate the AI tool. There are a lot of great solutions on paper, but the #healthcare systems may not have the expertise or the interest to really redesign the workflow to maximize the benefit from that AI tool.” Many of us have been saying this for years and they are heavily emphasized in my book, Unicorns, #Hype and Bubbles. The biggest healthcare success stories are in admin work and #AI people must focus on the workflow, or what I call the process, and not on individual productivity. It doesn’t matter how many reports, emails, and software code you generate. The WSJ asked Yang: “What are you most afraid of with AI in healthcare? “I don’t think that the enthusiasm around developing tools has been met with the same level of enthusiasm around testing, validating and demonstrating the safety and effectiveness of these tools. There’s a tremendous amount of effort in the development space, and I feel like the infrastructure to support responsible AI has not yet been able to catch up.” The WSJ asks: “What should AI never do? Is there a red line? We view them as augmenting, so I wouldn’t feel comfortable about AI automating clinical decision-making, in diagnosis or treatment.” Many people, including myself, have been saying this for years. #innovation #startups #artificialintelligence #ethics #doctors https://lnkd.in/gYTZgcE2

220

79 Kommentare