A refined approach for AI-assisted cancer diagnosis was developed by setting standards in the interpretation of scoring guidelines with a "PD-L1 Cell Atlas".
In a study investigating the efficiency gain in clinical routine with our AI Unilabs reported saving up to 90% diagnostic time compared to manual scoring. In addition, several patients needed to be re-classified leading to the identification of more patients eligible for treatment in the prospective study setting.
Our leading ‘0 Click’ AI algorithms are now available through Proscia’s Concentriq© platform used by Clinicians and Pharma R&D alike!
Our #ASCO2023 poster on HER2-low was selected for oral presentation by session chair Prof. David Rimm (Yale University School of Medicine). In his discussion he highlighted AI as an essential means to score HER2 in the future.
In a first of its kind comparative study, Mindpeak's CE-marked HER2 AI for breast cancer diagnosis was compared to three other commercial AIs (e.g. Visiopharm). Mindpeak's AI showed the highest accuracy of all AIs in terms of balanced clinical scores (88.1%) for HER2-low, higher than the score of the next best AI (67.3%) and higher than the concordance of pathologists with the reference score (85.5%).
Ziekenhuis Netwerk Antwerpen (ZNA) and GZA Hospitals chose Mindpeak’s PD-L1 AI to become the first lab in Belgium to use artificial intelligence in the fight against cancer.