Sedgwick’s Post

'SOUP' can ruin your AI-Enabled medical device development? "The process of fine tuning foundation models for more specific medical purposes may introduce significant quantities of Software of Unknown Provenance (SOUP), as the manufacturer performing the fine tuning may only have limited documentation for the base foundation model.' The "Good Machine Learning Practice for Medical Device Development: Guiding Principles" document outlines essential guidelines for developing safe AI-enabled medical devices. Guiding Principles: 1- Understanding Device Purpose and Multidisciplinary Expertise An understanding of a device's intended use within clinical workflows and leveraging multidisciplinary expertise throughout the product lifecycle is crucial.(AKA involve your clinicians!) 2-Good Software Engineering and Security Practices: This includes robust software engineering, usability, data quality assurance, data management, and cybersecurity. 3- Representative Clinical Study Participants and Datasets: Prevent bias and promote generalizability by ensuring datasets and study participants reflect the intended patient population (eg demographics and clinical conditions). 4- Independence of Training and Test Datasets: To improve validity, the 'training' and 'test' datasets should be independent.

Practical, consensus guidance on how to get the best out of technology in healthcare is always welcome. User buy-in is critical for all devices, no matter how sophisticated, to be accepted and scaled to where they can help the most patients.

FDA, Health Canada, and UK’s MHRA have issued today new guiding principles for transparency in the deployment of machine learning in medical devices. See the link: https://lnkd.in/gm8-8G7M . These transparency guiding principles build on the June 2021 guiding principles issued by the same regulators for Good Machine Learning Practice for medical device development. The new guiding principles document seeks to ensure that appropriate information about medical devices employing machine learning is clearly communicated to relevant parties, so they have a useful understanding how the device makes decisions or takes action, and is organized to address the standard questions: who, why, what, where, when, and how. How are you incorporating these two guiding principles documents into your software development practices? Your labeling?

MHRA, FDA, and Health Canada collaborate to introduce five guiding principles for machine learning-enabled medical devices #AI #artificialintelligence #FDA #HealthCanada #Healthcare #llm #machinelearning #machinelearningenabledmedicaldevices #MHRA #Patientsafety #predeterminedchangecontrolplans #regulatoryadaptation #regulatoryprinciples

Regulators develop five principles for machine learning-enabled medical devices  : The new principles will remove the regulatory burden for MLMD manufacturers

Health professionals have expressed alarm at the perceived shortcomings of the U.S. regulatory system in ensuring the safety of AI medical devices. About 700 such devices have received FDA approval. Doctors find it difficult to get adequate information about how new AI devices were trained or validated; current FDA guidelines don’t require that such data be disclosed. Learn more in #TheBatch: https://hubs.la/Q02bx1JN0

Artificial Intelligence as a Medical Device AIaMD software manufacturers are you ready for the Medicines and Healthcare products Regulatory Agency AIRLOCK! https://lnkd.in/enZvfKJ6 We’ve acknowledged for sometime that the current conformity assessment processes associated with software as medical devices does not necessarily best-fit into the ‘physical’ engineered medical device framework therefore the MHRA and other collaborators such as Team AB (U.K. approved conformity assessment bodies) have established a pilot for a digital ‘airlock’ so that software developers can submit their products into a non-clinical environment and be soak tested to see for example of the algorithm ‘drifts’ as well as look for other hazards and hazardous situations. There is a webinar on the 6th June hosted by the MHRA on this topic here https://lnkd.in/esU-hhGp Bravo MHRA for dusting off that medical device safety leadership!!!!

MHRA, FDA, and Health Canada have collaborated to set guiding principles on transparency for machine-learning medical devices. These principles, in alignment with previous guidelines, provide essential transparency practices for all medical devices. Stay informed on these crucial standards: Published: 13 June 2024 #MHRA #FDA #HealthCanada #MedicalDevices #Medtech #AI #machinelearning

At the heart of Cytophenix's rapid test under development sits AI software automatically analysing millions of data points, yielding the results doctors use to guide antibiotic treatment. Recent provision by the US FDA of Guiding Principles for AI-enabled Medical Devices help smooth the regulatory pathway of AI devices on their journey to market. The guidance helps set regulator expectations for in vitro diagnostic devices utilising such technology, and establish clear and reasonable processes for regulatory consideration. https://lnkd.in/gHBN2t3J For more information see: https://lnkd.in/ejqWRmV3 #AntimicrobialSusceptibilityTests #ASTs #rapidASTs #RegulatoryMatters #IVD

Exciting news in the world of medical technology! The FDA has introduced new guidelines to promote transparency for machine learning-enabled medical devices. Device developers now have a clear roadmap on what to share and how to share it, fostering better transparency practices across the industry. Stay informed about the latest developments in medical technology. Read more here: #medicaldevices #artificialintelligence