Sai Rupesh Kagga
Multi-Modal AI for Early Detection of Dialysis Complications Using EHR, Machine, and Sensor Data.
Abstract:
Patients receiving dialysis experience frequent and often sudden clinical complications, including intradialytic hypotension, access failure, infection, and cardiovascular instability. Although each dialysis session generates large amounts of clinical information, these data are typically scattered across electronic health records, dialysis machines, laboratory systems, and monitoring devices. Because these sources are rarely analyzed together, early warning signs are often missed, and care remains largely reactive.
This presentation examines a practical approach to building multi-modal artificial intelligence systems that bring these data streams together to support earlier and more reliable detection of dialysis-related complications. It focuses on integrating longitudinal EHR data with real-time dialysis machine parameters, laboratory trends, and physiologic sensor signals to create a unified view of patient risk. Rather than emphasizing model accuracy alone, the talk addresses the full system lifecycle, including data alignment, handling incomplete and asynchronous inputs, and ensuring that predictions are clinically interpretable and actionable.
The session also discusses how such systems can be deployed in real dialysis environments using cloud-based architectures that meet security, privacy, and regulatory requirements. Real-world examples illustrate how multi-modal models can surface subtle risk patterns well before clinical deterioration becomes apparent, allowing care teams to intervene earlier and reduce adverse outcomes.
By shifting dialysis care from isolated data review to integrated, predictive intelligence, multi-modal AI offers a path toward safer, more personalized, and more proactive treatment. This talk provides a grounded, systems-level perspective on how software engineering, cloud infrastructure, and applied AI can be combined to deliver meaningful improvements in dialysis care.
Profile:
Sai Rupesh Kagga is a senior software developer specializing in large-scale healthcare systems and enterprise-grade application design. He has extensive experience in full-stack development, cloud-native architectures, and secure interoperability using standards such as FHIR and HL7. His work includes modernizing clinical platforms, improving diagnostic workflows, and developing reliable solutions that support real-time data exchange in healthcare settings. Sai has contributed to several digital health initiatives focused on strengthening system performance and clinical efficiency. He is known for his practical approach to solving complex engineering challenges and for guiding teams in building scalable, compliant, and resilient healthcare applications. His keynote reflects his continued commitment to advancing technology that supports clinicians and enhances patient care.