The company reports this is the first-ever AI-based solution for cancer detection used in routine clinical practice – is already deployed in pathology labs with demonstrated success in alerting on missed cancer cases. With this CE-IVD Mark of Galen Prostate, Ibex is now partnering with leading institutes across Europe to implement AI and support pathologists in their diagnostic workflow.
Pathology is the medical specialty of diagnosing disease in patients and plays a pivotal role in cancer diagnosis and treatment decisions. To this day, cancer diagnosis is performed by pathologists looking at biopsies through a microscope. In recent years, a global increase in cancer incidence and in the number of screening tests has coincided with a decline in the number of pathologists, causing a crisis that affects labs worldwide and results in longer turnaround times and rising concerns about diagnostic accuracy.
Ibex uses AI to develop clinical-grade solutions that help pathologists detect and grade cancer in biopsies, helping them ensure diagnostic accuracy, integrate comprehensive quality control and enable a more efficient workflow. Galen Prostate is a vendor-neutral AI solution, evaluated across multiple scanning systems, staining platforms and pathology labs. The solution uses an algorithm that was trained on slides from a dataset of over 60,000 prostate slides from multiple institutions and representing a variety of diagnoses and clinical findings.
The CE-IVD Mark follows the outstanding results from a blinded multi-site prospective clinical study measuring, for the first time, performance against consensus diagnosis between multiple pathologists. The study included digitized slides from prostate core-needle biopsies analyzed by Galen Prostate. Performance was evaluated against the consensus diagnosis of multiple pathologists. Ibex’s AI-powered algorithm demonstrated the highest AUC (area under the ROC curve) reported in the field, 0.997, as well as very high sensitivity and specificity (98.6% and 95.6%, respectively). These results validate the robustness of Galen Prostate and support its adoption by pathology institutes aiming to improve their diagnostic accuracy by enhancing quality control with AI.
“We are impressed with the results of the study, particularly with the performance of Galen Prostate, which was in near perfect concordance with the consensus diagnosis between multiple pathologists. This solution can become an important tool for pathologists that look for ways to improve diagnostic quality and workflow efficiency,” said Delphine Raoux, MD, Head of Innovation Technologies at Medipath – the largest network of private pathology labs in France, who participated in the study. “Medipath believes that AI-powered solutions are the future of pathology and we are happy to participate in trials that demonstrate their accuracy and utility, and to collaborate with Ibex to help the evolution of their AI solutions.”
“We hit an important milestone with the CE-IVD Mark, facilitating a wider adoption of Galen Prostate in the market. These clinical-grade results confirm our robust and meticulous approach to AI, that corresponds to the pathologist’s own workflow,” said Joseph Mossel, Ibex Medical Analytics CEO. “Cancer incidence and testing for cancer keep growing, with an impact on pathologists’ turnaround times and misdiagnosed cases. Ibex is committed to providing practical solutions that help pathologists meet these challenges.”