Digital Reasoning announced results from its automated radiology report research. This is a 5th peer reviewed research published about our technology solutions – unmatched in our space. In a series of experiments on radiology reports from emergency departments, inpatient and outpatient healthcare systems, Digital Reasoning tested the feasibility of automating this process with natural language processing (NLP) and machine learning (ML), achieving a queue precision of 90.2%.
Reviewing incidental findings can be a labor-intensive process for Healthcare Systems. This approach promises to improve healthcare quality by increasing the rate of appropriate lung nodule incidental finding follow-up and treatment without excessive labor or risking overutilization.
Building on earlier pilots, Sarah Cannon, the Cancer Institute of HCA Healthcare, established a standard clinical workflow process to facilitate response to a lung nodule incidental finding, providing earlier diagnosis and improved outcomes. Incidental Findings of Lung Nodules were found in 9,949 reports with 6,884 representing unique patients with critical nodules. All patients with critical nodules were offered expedited clinical follow-up.