PolyU Develops AI Virtual Patient System to Predict Cancer Treatment Outcomes
On.cc · 1 SOURCESabout 2 hours ago2 MIN
Summary
Hong Kong Polytechnic University (PolyU) researchers have developed a groundbreaking artificial intelligence (AI) virtual patient simulation system designed to predict the effectiveness of cancer treatments by dynamically integrating patients' genetic, medical imaging, and clinical data. The technology, which enables real-time tracking of patient conditions and simulation of various treatment outcomes, was showcased at the 2026 World Mobile Communications Conference in Barcelona, Spain, where it was nominated for the finals of the Best Mobile Internet Health and Wellbeing Innovation Award at the 2026 Global Mobile Awards . Led by Professor Chan Wing-chi, Associate Professor in PolyU's Department of Health Technology and Informatics, the system is now being gradually implemented in clinical environments with support from various funding programs .
Key Points
- PolyU's AI virtual patient simulation system integrates multimodal data including genetic information, medical imaging, and clinical records to monitor patient conditions in real time
- The system achieved 82.55% accuracy in predicting immunotherapy response for non-small cell lung cancer patients using its proprietary Vision-Global Relationship Fusion Network (ViGNet) framework
- Unlike conventional medical AI tools that rely on single data sources, this platform provides comprehensive predictive analysis by considering dynamic changes in patient conditions
- The technology includes both a healthcare provider platform for medical staff and a mobile application for patients to upload medical records and track daily symptoms
- Patient data security is ensured through encrypted "deep feature QR codes" that enable safe transmission of medical records between clinics, hospitals, and devices
Why It Matters
This AI system addresses a critical gap in cancer treatment by enabling personalized medicine approaches that consider each patient's unique biological characteristics and disease progression. For Hong Kong's healthcare system, which faces growing demand for cancer care amid an aging population, the technology could significantly reduce diagnostic and treatment evaluation times while optimizing healthcare resources. The 82.55% prediction accuracy for immunotherapy response specifically helps oncologists avoid ineffective treatments, potentially reducing both costs and unnecessary side effects for patients.
This AI system addresses a critical gap in cancer treatment by enabling personalized medicine approaches that consider each patient's unique biological characteristics and disease progression. For Hong Kong's healthcare system, which faces growing demand for cancer care amid an aging population, the technology could significantly reduce diagnostic and treatment evaluation times while optimizing healthcare resources. The 82.55% prediction accuracy for immunotherapy response specifically helps oncologists avoid ineffective treatments, potentially reducing both costs and unnecessary side effects for patients.