Call us now:
@Thaddeus Dziekanowski, the U.S Federal Drug Administration (FDA) has issued draft guidance recommendations to industry on the submission AI enabled devices and software functions.It’s a robust read at 67 pages. Worth a scroll though as it sets the agenda for the Agencies evaluation of AI enabled medical devices. With a focus on the security of devices, product life cycle, and what product sponsors should consider when the submit applications to the FDA.One notes the focus on a risk based approach to the creation and use of model data management, integration on devices into existing infrastructure, data management, and cybersecurity. One likes the focus on the clarification of the reference standard or representative truth, being validated by evidence from current practice and reflect the clinical task. An important constituent of the AI enabled device that builds the Artificial Domain Intelligence for the specific device and application. In addition to the submission of validated, repeatable model outputs. With a view to minimising bias and hallucinations.The FDD focuses on cybersecurity as a means to preventing data poisoning, model inversion and evasion, data leakage, overfitting, model bias and performance drifting. All areas that any AI product or solution should be able to address in order to deliver a reliable and repeatable solution.Brian D. McCarthy Faisal Khan Rachel V. Rose, JD, MBA Ryan BucknerVeritas GRC#cybersecurity #cyberiskmanagement #cyberregulation #ai
View Original Post
https://www.linkedin.com/feed/update/urn%3Ali%3AugcPost%3A7282764436916908033