Explainable Machine Learning based Medical Data Analysis
- Reference number
- SAB23-0012
- Project leader
- Chatterjee, Saikat
- Start and end dates
- 241201-251231
- Amount granted
- 1 300 000 SEK
- Administrative organization
- KTH - Royal Institute of Technology
- Research area
- Information, Communication and Systems Technology
Summary
Analysis of healthcare data with explanations provides a unique opportunity to develop novel technologies for accurate predictions and diagnostics of severe outcomes in hospitals. The main objective of this sabbatical grant is to explore this unique opportunity. At the Intervention Centre, Oslo University Hospital (OUH), Oslo, Norway, the applicant will spend one year time for the sabbatical. The applicant has two main objectives: (1) He will collaborate with four existing groups with his expertise in explainable machine learning (XML) for heath care problems based on his current experience with Karolinska University Hospital (KUH) and Karolinska Institute (KI). XML helps medical caregivers to understand algorithmic decisions. (2) He will learn the process of developing such an Intervention Center in the Hospital environment, and how to build such a unique facility at KUH, which currently does not exist. This is highly important to establish such an interdisciplinary centre inside hospital for direct access to medical data without data exchange, and direct cooperation with medical caregivers. Overall, the sabbatical will help to strengthen long-term cooperations between KTH, KUH, KI and OUH for medical data analysis. This will also establish connections between Region Stockholm and Oslo Region Authorities. The applicant will leverage his cooperations with Region Stockholm, KUH and KI, and several projects including SSF-supported mobility grant to spend time in KUH.
Popular science description
Artificial Intelligence (AI) can be used to detect severe medical events in life, help doctors for choice of medication, find details in medical images like X-ray images, etc. For example, let us consider infections due to virus and bacteria. AI can detect infections earlier than current clinical practice. Infections do not always show clear and specific symptoms, why doctors and nurses can miss the diagnosis. Once infection starts, the human immune system automatically starts to fight bacteria and viruses. As an effect, body parameters, such as heart rate, blood pressure, breathing pattern, body temperature, etc. start to change slowly. While humans can not recognise the slow subtle changes, AI can if data can be collected and analyzed. Doctors and nurses then have precious time to initiate life-saving interventions early. Each patient at modern hospitals generates a huge amount of data. Such data includes vital signs, electronic health records (EHR), text data, lab results, radiology imaging, ventilator states, fluid, medications, etc. All the data contain clues to the patient's medical state and can be used for early detection of deterioration, such as the infection buildup, development of diseases, cardiac arrest, neurological disorders, cancer onset, etc. Our long term vision is to use modern AI to analyse such data for early warning. Data analysis and early detection are integral part of a clinical decision support system. An autonomous decision support system will aid accurate medical decisions, saving lives. In all most all medical research arenas, a vision is to develop autonomous clinical decision support systems that will perform real time data acquisition, storage, analysis, and provide automatic monitoring and appropriate warning notifications. To develop such kind of autonomous clinical decision support systems that will assist medical caregivers, it is imperative that machine learning experts spend quality research time in hospitals for understanding a full scale system requirements and study several kind of medical cases. Further the experts need to clearly understand the needs of doctors and nurses, and how they currently analyze data and decide medical diagnostics, and where lies the scope of improvements. This is an inter-disciplinary research combining engineering with medical knowledge, helping to design AI-based medical data analysis systems and autonomous clinical decision support systems at Swedish hospitals in future.