PocketHealth unveils image reader for patients

TORONTO – PocketHealth, a connected care company, has launched Image Reader, the first feature to apply AI-powered anatomical visualization directly into patients’ imaging results. By providing visual context within a scan, Image Reader helps patients better understand what they are seeing in their images, leading to more informed discussions with providers and fewer misunderstandings about their results.

Image Reader leverages multiple segmentation models including MedSAM (Medical Segmentation Anywhere Model) that have been refined to support a patient-friendly application of medical imaging segmentation. Developed by AI scientist Dr. Bo Wang (pictured) and his team, MedSAM is a model specifically designed for universal medical image segmentation across diverse modalities.

Building on these capabilities, PocketHealth Image Reader identifies and transforms anatomy mentioned from radiology reports onto segmented anatomical structures in CT and X-ray scans, ultimately improving patient understanding of their medical images.

“Reviewing medical imaging results can be complex and intimidating for patients, and radiology reports often lack visual references, making it harder for patients to connect findings to their own anatomy,” said Dr. Ram Chadalavada, chief medical officer, PocketHealth and vice-chair of radiology informatics at UC Health. “PocketHealth has improved patient literacy around medical reports with Report Reader – now, Image Reader builds on that by automatically identifying organs and bones in scans, offering clarity that helps patients feel more confident, informed, and engaged in their care.”

Image Reader is PocketHealth’s newest feature designed to help patients interpret their imaging with greater clarity and confidence expanding PocketHealth’s patient-centered toolkit, complementing Report Reader and MyCare Navigator.

PocketHealth is the first to bring AI-powered anatomical visualization directly to patients within their medical imaging records, providing a more interactive experience. Fine-tuned for generalizability, PocketHealth’s model accurately identifies anatomy across a range of imaging sources, scanner types and clinical settings by automatically detecting and labeling organs and bones within medical images.

“Medical imaging AI has primarily focused on clinical applications, but there’s an equally important opportunity to improve patient understanding,” said Dr. Bo Wang, chief AI scientist at the University Health Network (UHN), who led the development of MedSAM. “PocketHealth has taken an innovative approach by refining our segmentation model for real-world patient use. Using this technology to directly benefit patients is a meaningful step toward making medical imaging more accessible and insightful.”

ImageReader is currently optimized for a variety of CT and X-ray exams with plans to expand support for additional imaging modalities.

About PocketHealth
PocketHealth is a connected care company that empowers patients to take control of their health by providing seamless access to their diagnostic imaging records. With enterprise image exchange and intelligent patient engagement solutions, PocketHealth enables healthcare organizations to seamlessly share records and surface clinical insights, while eliminating network image-sharing dependencies to create efficiencies, reduce costs, and keep patients at the center of their healthcare journey. Learn more at pockethealth.com.