C. Matthew Hawkins, MD, discusses challenges of natural language processing and how more standardization can reduce error rates of speech recognition software.
[[{"type":"media","view_mode":"media_crop","fid":"11967","attributes":{"alt":"","class":"media-image media-image-right","id":"media_crop_7562901937715","media_crop_h":"0","media_crop_image_style":"-1","media_crop_instance":"207","media_crop_rotate":"0","media_crop_scale_h":"0","media_crop_scale_w":"0","media_crop_w":"0","media_crop_x":"0","media_crop_y":"0","style":"margin: 5px; float: right; border-width: 0px; border-style: solid;","title":" ","typeof":"foaf:Image"}}]]As many as 22 percent of radiology reports contain errors associated with natural language processing, said C. Matthew Hawkins, MD, a pediatric radiology fellow at Cincinnati Children’s Hospital Medical Center. Hawkins is speaking on the topic at the SIIM 2012 conference in Orlando this week.
In this podcast, Hawkins discusses how these errors can change the meaning of radiology reports and how augmenting standardized content associated with various types of speech recognition software can decrease mistakes.
Considering Breast- and Lesion-Level Assessments with Mammography AI: What New Research Reveals
June 27th 2025While there was a decline of AUC for mammography AI software from breast-level assessments to lesion-level evaluation, the authors of a new study, involving 1,200 women, found that AI offered over a seven percent higher AUC for lesion-level interpretation in comparison to unassisted expert readers.