Apple hosted the Machine Learning for Health Workshop. The Workshop focused on using machine learning (ML) to improve the healthcare industry. Dr. Yindalon (Yin) Aphinyanaphongs, a workshop attendee, talked about a new cycle to quickly refine and reapply clinical data in healthcare settings. Another speaker, Dr. Ziad Obermeyer, discussed the difficulties in using machine learning to predict sudden cardiac death. The presentations, highlighed the benefits of using and accessible health data for developing AI algorithms.
Fairness and robustness are critical in ML for health. Traditional models are often not diverse enough across countries and demographic factors. ML models trained on these datasets learn biases and are not applicable to different populations. Professor Daniel Gatica-Perez worked with partners on developing a multicountry mobile-sensing dataset to mitigate these biases.
It’s important to address individual differences when designing ML models for health. Building AI systems tailored to individual patterns and objectives is crucial in healthcare. Apple is working on the Apple Women’s Health Study to collect data on menstrual cycles, health, and behavior in a digital cohort.
Lastly, privacy in ML for health is important. Protecting sensitive health data is a top priority when developing AI tools for healthcare. Dr. Shrikanth (Shri) Narayanan presented on how machine intelligence can be used to analyze speech and language development in children with autism. He emphasized the need for privacy and confidentiality in healthcare data.The Apple Machine Learning for Health Workshop was a major success in bringing together experts in machine learning and healthcare to discuss and improve the future of artificial intelligence in the healthcare industry.