Unlocking Multimodal Clinical Data for AI | Baptist Health Case Study

Multimodal clinical data is everywhere in healthcare — but most of it is unusable for research, AI, and real-world evidence. This session shows how Baptist Health turned fragmented, sensitive patient data into structured, AI-ready patient journeys — without compromising privacy, governance, or HIPAA compliance. Nila Bhakuni (AVP Innovation, Baptist Health South Florida) and Linda Chen (COO, John Snow Labs) walk through what it actually takes to unlock secondary use of clinical data at scale — from raw EHR systems to production-ready AI infrastructure. At Baptist Health, managing 1.8M patients across 12 hospitals, the challenge was not access to data — but making it usable. Most insights were locked in unstructured notes, PDFs, imaging data, and fragmented systems. 📌 Applied Healthcare AI Summit 2026 — what actually works in real-world healthcare AI, from pilots to production systems. ⏱️ Key moments: 00:00 Why clinical data is hard to use for AI and research 02:20 Baptist Health scale and data complexity 04:40 Privacy, de-identification, and HIPAA constraints 07:00 From multimodal data to AI-ready patient journeys (OMOP) 10:10 From data to use: cohorts, trials, and patient-level insights #HealthcareAI #AIinHealthcare #ClinicalData #HIPAA #DataPrivacy #RealWorldEvidence #PatientData #AIGovernance #ClinicalResearch