Med AI Goes Multimodal, AI-Powered Ultrasounds Are Here and BSI Publishes Standards to Boost Trust in AI for Healthcare

Hello, and welcome to this edition of?The Med AI Capsule! ??????

I am thrilled to have you on board as we embark on a journey into the fascinating world of Medical AI. In this newsletter, I'll be your trusted source for the latest updates, breakthroughs, and thought-provoking insights in the ever-evolving field of artificial intelligence in healthcare.

Whether you're a healthcare professional, a tech enthusiast, or simply curious about the future of medicine,?'The Med AI Capsule'?is your compass to navigate the dynamic world of AI in healthcare.

Here are three of the latest developments in the field of AI in healthcare:

1.?Google Unveils Multimodal Generative AI Model 'Med-PaLM M' for Healthcare

Google's Med-PaLM M is a large multimodal generative model that can handle various types of medical data, including clinical language, medical images, and genomics.

Key takeaways:

• Med-PaLM M outperforms specialist models and even surpasses PaLM-E, proving the importance of adapting the model to biomedical data.
• Med-PaLM M can generalise to new medical tasks and concepts and perform multimodal reasoning without specific training.
• Preliminary evidence suggests that Med-PaLM M can accurately identify and describe medical conditions in images using only language-based instructions and prompts.

Counter arguments:

• Google acknowledges the complexity of personalised medical care and recognises that Med-PaLM 2’s results may not be generalisable to every medical question-answering setting and audience.
• Privacy concerns are likely to prevent many patients from sharing personal data, which could affect Med-PaLM 2’s efficiency.

2.?New AI ultrasound tech is first to land FDA approval to enhance prenatal care: ‘Better health outcomes'

• Sonio Detect, an AI-powered ultrasound scanning technology, has become the first product of its kind to land FDA approval.
• During testing, Sonio Detect scanned over 70,000 ultrasound images and identified more than 300 potential prenatal syndromes and anomalies with an accuracy rate of 95% or higher.

"Traditional fetal ultrasounds require manual labeling on every captured image," Sonio’s CEO and co-founder Cecile Brosset said in an interview with Fox News Digital.?"This manual process is prone to errors and forgetting important images of the fetus — especially when a standard second-trimester exam includes more than 80 images."

• Sonio Detect is compatible with any type of ultrasound technology, including GE, Samsung and Caron.
• Benefits: Sonio Detect helps healthcare professionals ensure that ultrasound exams are complete and that the images are of good quality. It also has the capability to automatically extract images from clips acquired by physicians.

3.?BSI publishes guidance to boost trust in AI for healthcare

• The British Standards Institution (BSI) has released high-level guidance to foster greater digital trust in AI products used for medical diagnoses and treatment.
• The guidance, titled ’Validation framework for the use of AI within healthcare – Specification (BS 30440),’ aims to bolster confidence among clinicians, healthcare professionals, and providers regarding the safe, effective, and ethical development of AI tools.

“The new guidance can help build digital trust in cutting-edge tools that represent enormous potential benefit to patients, and the professionals diagnosing and treating them. AI has the potential to shape our future in a positive way and we all need confidence in the tools being developed, especially in healthcare. This specification, which is auditable, can help guide everyone from doctors to healthcare leaders and patients to choose AI products that are safe, effective, and ethically produced.” - Scott Steedman, Director General for Standards at BSI

• The BS 30440 specification seeks to aid decision-making processes by providing criteria for evaluating healthcare AI products, including clinical benefit, performance standards, safe integration into clinical environments, ethical considerations, and equitable social outcomes.

Stay tuned for our upcoming editions where we'll uncover the latest breakthroughs and explore the potential of AI to shape a healthier tomorrow.

PS:?If you are a?medical professional who is interested in artificial intelligence but confused about where to start, feel free to check out my?CPD-accredited course on Bionabu?and my?Whatsapp Community. For?1:1 help or guidance, please visit?HERE.

Best regards,

Dr Avneesh