PRISM

Chandra Prasetyo Utomo, Kohei Ichikawa, Nashuha Insani, Kundjanasith Thonglek, Kang Xingyuan, Chaerita Maulani, Ummi Azizah Rachmawati

PRISM is an advanced clinical decision-support tool engineered to improve outcomes for critically ill patients in the Intensive Care Unit (ICU). Sepsis is a life-threatening condition requiring rapid, highly tailored interventions, yet optimal treatment strategies often vary significantly among individuals. PRISM leverages sophisticated machine learning algorithms to continuously analyze high-frequency physiological data and comprehensive electronic health records in real-time. By tracking a patient’s evolving clinical state, the system dynamically models disease progression and predicts individual responses to various medical interventions, such as fluid resuscitation and vasopressor administration. PRISM then generates actionable, personalized treatment recommendations designed to stabilize the patient and reduce mortality risks. This project bridges the gap between complex artificial intelligence and bedside clinical care, providing ICU physicians with a powerful, data-driven assistant that enhances decision-making, optimizes resource utilization, and ultimately strives to save lives in high-stakes medical environments.

---