When dealing with a stroke, acting swiftly is crucial. CT scans enable doctors to rapidly locate a blood clot within the brain to decide the best course of treatment. Such imaging is key in healthcare, not only for urgent cases but also in preparation for surgeries. MRIs are instrumental for surgeons to meticulously plan procedures in advance. These technologies share one thing: they produce extensive data that are integral for customized healthcare. Artificial intelligence, especially machine learning, is poised to harness this data to shape future diagnostics and treatments that fit individual patient needs. Yet, this technology comes with its own set of challenges and uncertainties. Now, a team of researchers is developing a system that automatically analyses and visualizes medical data, including their uncertainties.

For instance, uncertainties become apparent when assessing brain damage from a stroke. When a vessel blockage deprives brain regions of oxygen, these areas cease to function properly. It's often easier to identify the central part of the damage, but the outer edges are more ambiguous, sometimes leading to discrepancies among medical professionals on whether these areas constitute part of the lesion. Dealing with such ambiguities calls for significant medical expertise. A team of researchers from the University Kaiserslautern-Landau (Kaiserslautern, Germany) and University of Leipzig (Leipzig, Germany) is currently developing a uniform system for processing and evaluating medical image data as well as visualizing their uncertainties.

Named GUARDIAN, the system is designed to be user-friendly for clinical settings. Healthcare providers can upload their trained neural networks into the system and integrate them with existing processed data, like that from a stroke patient's scans. The system then autonomously analyzes the data and presents the findings, simplifying the process without requiring IT expertise. By highlighting areas of uncertainty, GUARDIAN supports medical professionals in re-examining the data and collaborating on the most suitable treatment plan for each patient. The team behind GUARDIAN will showcase their system at the “Medica” medical technology fair in Düsseldorf, taking place from the 13th to the 16th of November, at the Rhineland-Palatinate research booth. 

http://m.allmycbs.com/en/index.asp