Health IT firm Dicom Systems has anonymized 5.3 million musculoskeletal imaging exams at the Hospital for Special Surgery in New York City for the purposes of testing artificial intelligence (AI) algorithms.
Dicom System's proprietary deidentification tool allows for the removal of patient-identifying information from imaging exams without corrupting or truncating the image file -- thus making large datasets suitable for training AI algorithms, according to the company. The feature is fully customizable and available as an add-on to Dicom's Enterprise Imaging Unifier interface engine.
