There are numerous imaging studies where radiologists will be supported by the aid of artificial intelligence (AI) software, but none has stirred more interest than AI's machine learning to the interpretations of screening mammograms. For several years, radiologists have benefited from advances in breast imaging technologies to assist in the identification of suspicious lesions and calcifications on mammograms, such as 3D digital breast tomosynthesis. The use of AI's machine learning technology is becoming widespread  - and not just in the United States.

A New York Times article's reports that radiologists in Hungary have made that Eastern European county a major testing ground for the use of AI to identify potential cancers on mammograms. According to the Times, radiologists in Hungary are utilizing AI technology on real patients at five hospitals and clinics that perform more than 35,000 breast screenings a year. 

Last year, after a test on more than 275,000 breast cancer cases in Hungary, AI software matched the performance of human radiologists when acting as the second reader of mammography scans. Using AI as a second reader cut down on radiologists’ workloads by at least 30 percent. The Times also reports at one Hungarian clinic last year, the use of AI increased the cancer detection rate by 13 percent as more malignancies were identified.

So the obvious question emerges as to whether AI will eventually replace "human radiologists." Not likely, say the Hungarian researchers who believe that the technology will be effective and trusted by patients only if it is used in partnership with trained doctors.

I agree with Hungarian software developer, Peter Kecskemethy, the co-founder of Kheiron Medical Technologies, "An AI-plus-doctor should replace doctor alone, but an AI should not replace the doctor."

Still to be determined is whether AI can produce accurate results on women of all ages, ethnicities and body types. And AI technology must prove it can recognize more complex forms of breast cancer while also reducing the number of false-positives that are not cancerous, leading to unnecessary biopsy procedures.

In light of these continued developments, who can doubt that questions about AI effectiveness will be resolved?