Advancing Healthcare of Vulnerable Populations Through AI and Wearable Devices

Explore a 44-minute conference talk on leveraging AI and wearable devices to improve healthcare for vulnerable populations. Delve into multidisciplinary collaborations addressing clinical challenges, focusing on AI-based neonatal pain assessment and sickle cell disease progression prediction. Examine the development of datasets, generative and multimodal approaches, and automated predictor quantification. Discover experimental results, ongoing work, and the potential of real-time temporal analysis and explainable AI in healthcare. Gain insights into the challenges and opportunities presented by automated clinical data analysis and risk stratification for vulnerable patient groups.

Intro
Outline
Vulnerable Population
Multidisciplinary Collaboration
Clinical Problem - Current Practice
AI-based Neonatal Pain Assessment
USF Multimodal Pain Dataset
Generative & Multimodal Approach
Experimental Results
Summary & Ongoing Work
Predictors of SCD Progression
Automated Predictor Quantification
Automated Imaging Predictor Quantification
Real-time Temporal Analysis
Explainability
Datasets for EchoQuan Development
Results: EchoQuan vs. Expert
Automated Clinical Data
Dataset for SCD Risk Stratification
Big Potential, Big Challenges