Diagnostic Reading #20: Five “Must Read” Articles on HIT and Radiology
Reading Time: 3 minutes read
Top radiology CPT codes, and how to incorporate values-based practice into radiography are in the news.
This week’s articles in Diagnostic Reading include: top radiology procedures at imaging centers; why radiologists shouldn’t fear deep learning; a new guide explains how to incorporate values-based practice in radiography; a new imaging technique may help mental health patients; and deep learning outperforms dermatologists.
Radiologists use medical imaging technology—such as X-rays, CT scans, MRIs, ultrasounds and others—to diagnose and treat injuries and illnesses in patients. This article spotlights the top radiology procedures at imaging centers by total charges and by volume using claims analytics.
3 reasons radiologists shouldn’t sweat deep learning – Health Imaging
The human-level success of deep learning has made some in medicine question whether automation may eventually take over many tasks performed by radiologists. In a recent editorial published in the Journal of the American College of Radiology, an author and radiologist from Stanford University School of Medicine provided three examples of why radiologists should not fear deep neural networks (DNNs). Read a similar opinion in this blog by Dr. Eliot Siegel.
A guide to values-based practice in radiography – Everything Rad
Every patient is different. Therefore, it is important not to make assumptions about how a patient is feeling or what might be important to them. In this blog, one of the editors of the handbook on Values-based Practice in Diagnostic & Therapeutic Radiography explains how to incorporate VBP into radiography.
Researchers have found that low-intensity ultrasound can change decision-making processes in the brain, according to a report in Nature Neuroscience. The lead author said this imaging technique, called ultrasound neurostimulation, could “improve the lives of millions of patients with mental health conditions by stimulating brain tissues with millimeter accuracy.”
Deep learning outdoes even deeply experienced dermatologists – AI in Healthcare
A deep-learning algorithm trained entirely on open-source images has outperformed 136 of 157 dermatologists at classifying melanoma, according to a study in the European Journal of Cancer. The team found the algorithm scored higher than most of the doctors on both average specificity and average sensitivity. The research team acknowledged several limitations in their study design. The work follows similar research with similar findings published last fall in Annals of Oncology.