According to a New York Times article dated March 5, 2023, artificial intelligence, or A.I., may help in detecting cancer and in reducing human error.
The article concerned a program in Hungary, which has an aggressive breast cancer screening program. Breast cancer screening is done with mammograms, which are special x-rays (radiographs) of the breast. A mammogram screening is as good as the radiologist reading it. A poor reading will lead to a failure to diagnose and a delay in diagnosis that could be deadly.
In the Hungarian example, radiologists with decades of experience failed to detect cancer on a mammogram. However, with the assistance of A.I., another radiologist flagged areas that were at least potentially cancerous, and the patient will undergo biopsy. According to the article, the A.I. technology circles in red specific areas on the radiograph that may be problematic.
Furthermore, in a Hungarian study of some 275,000 mammograms, A.I. interpretations matched those of human radiologists. Indeed, cancer detection increased by 13 percent because A.I. saw more malignancies.
Clinics in the U.S. are starting to study the use of A.I. by radiologists. The technology is far from being used commonly but holds out hope of reducing misinterpretations.
While concern has been expressed about A.I. replacing radiologists, the more likely outcome is that A.I. will be a useful tool to assist radiologists. The fact is, radiologists read hundreds of images a day, and missing small—treatable—cancers is a real potential.
There are no comments for this post. Be the first and Add your Comment below.
Leave a Comment