Radiology and Pathology Reports

Fillmore and his collaborators have developed natural language processing algorithms to automate the extraction of critical diagnostic data from unstructured radiology and pathology reports. His work demonstrates that these computational tools can accurately identify key clinical markers, such as Gleason scores in prostate cancer and evidence of venous thromboembolism, across large patient cohorts. These advancements reduce the reliance on manual chart review and facilitate the collection of standardized data for large-scale clinical research.