Segmed and Verily Partner on Imaging Data
Segmed has partnered with Verily to expand researcher and developer access to real-world imaging data. The alliance aims to accelerate advances in clinical AI, digital pathology, and diagnostic analytics by providing enhanced datasets.
- The de-identified imaging data will be accessible through Verily's Pre Exchange and Workbench environment, a secure platform designed to accelerate precision health research. - The initial dataset provided by Segmed will concentrate on breast cancer, featuring a longitudinal group of 558 patients who have undergone at least one digital breast tomosynthesis (DBT) exam, including 271 patients with biopsy-confirmed malignancies. - Segmed's broader global imaging platform, which will be drawn upon, includes approximately 150 million de-identified imaging exams from roughly 30 million patients across five continents. - This collaboration is part of Verily's larger strategy of partnering with health organizations to advance AI-powered precision health; other recent collaborators include UCHealth and the University of Colorado Anschutz for AI model development in areas like oncology and neuroscience. - Verily, an Alphabet company, has previously partnered with digital pathology specialist Lumea to develop and validate AI algorithms for diagnosing and grading prostate cancer. - Segmed, a Stanford-founded startup, recently raised $10.4 million in a Series A round led by iGan Partners and Advocate Health to expand its footprint in healthcare AI and life sciences. - A key goal of the partnership is to enable the creation of more comprehensive, multimodal datasets by linking imaging studies with other real-world sources like electronic health records (EHR) and claims data. - The collaboration reflects a broader industry trend where the focus is shifting from the sheer quantity of real-world data to the quality and usability of integrated data platforms for developing and validating AI tools.
