AI-based image phenotyping of early adpkd - overall project summary: our mission is to facilitate translational and basic research on polycystic kidney disease and related disciplines by providing…

The National Institute of Diabetes and Digestive and Kidney Diseases awarded a Cooperative Agreement totaling $611,688 to Weill Medical College of Cornell University on September 12, 2025, under the Diabetes, Digestive, and Kidney Diseases Extramural Research program (CFDA 93.847) to develop an AI-based image phenotyping core for autosomal dominant polycystic kidney disease (ADPKD) research.

The recipient will establish a biomedical research core that comprehensively labels and annotates abdominal-pelvic MRI and CT images from ADPKD subjects, including annotations of organs, cysts, fat, and muscles. The core will create a secure, centralized repository storing research images from PKD-RRC Center subjects, 400 existing subjects in the Rogosin ADPKD Data Repository, CRISP Study images, and additional ADPKD subjects as they enroll or utilize the core. The repository will generate quantitative imaging biomarkers including renal and extrarenal organ volumes and cyst volumes to enable longitudinal assessment of disease progression and treatment response within individual subjects. The labeled image database will provide researchers with an essential toolbox for extracting imaging biomarkers and exploring analytical improvements in noninvasively characterizing and measuring anatomic features of ADPKD across all organs and tissues visible on abdominal-pelvic imaging.

Work is performed in New York, New York. The period of performance extends from September 12, 2025, through August 31, 2030.

AI-based image phenotyping of early adpkd - overall project summary: our mission is to facilitate translational and basic research on polycystic kidney disease and related disciplines by providing access to accurate, reproducible, comprehensive measurements of all features of PKD visible on human clinical imaging (abdominal pelvic MRI and CT). This proposal builds upon a long tradition of pioneering PKD imaging research by our investigators who have been at the forefront of PKD imaging. We propose a biomedical research core that will comprehensively label images of adpkd subjects enrolled in research or being considered for research projects. Adpkd subject image labeling will be optimized for highly accurate and reproducible image phenotyping by providing access to an essential "toolbox" for extracting imaging biomarkers from PKD subjects. We also provide imaging expertise tailored specifically to PKD research and access to adpkd image data. The PKD image phenotyping repository core, will fully label and store in one secure location, all research images on PKD-RRC core center subjects and 400 existing subjects in our rogosin adpkd data repository, crisp study images and more adpkd subjects as they enroll or utilize this core. This includes annotations of 0RGANS/cysts/FAT/muscles for providing reports to the contributors on these critical imaging biomarkers (e.g., renal and extrarenal organ volumes, cyst volumes). This database of annotated images will empower imaging and PKD scientists to explore analytical improvements in noninvasively characterizing and measuring the anatomic features of adpkd within all organs and tissues visible on abdominal-pelvic MRI and CT. These quantitative measurements will assess disease progression and responses to treatment within individual subjects. This proposal addresses a critical need in the PKD imaging research pipeline, from the basic science of image analysis to clinical investigation of this systemic disorder and assessing response to treatments. In addition, core leaders will pioneer solutions to the needs of PKD researchers. We expect this strategy will continue our proven track record of supporting high-impact science nationally. These activities will be organized by an administrative core that leverages the resources of cornell university. Our commitment to the principle of collaborative science includes streamlining procedures for collecting and sharing images, imaging biomarkers, and expertise, integrating core activities, and collaborating with other rtccs within the PKD research consortium.