AI may Save Hungarian Healthcare, Says Leading Doctor

Madácsy says the technology can help speed up the training of new doctors and take them to the level of an expert. The Endo-Kapszula Magánorvosi Centrum, where the doctor works, already uses AI for a plethora of tasks.

Despite initial fears from patients and professionals when AI was introduced to the healthcare industry in Hungary, its use has grown exponential within the past decade. The AI algorithms found in healthcare facilities allow machines to learn and perform tasks such as categorizing and grouping materials without human input.

Madácsy points out that the long waiting lists patients encounter are due to a lack of highly trained professionals. He says the only way this can be improved in the long term is through AI. He also claims that patients will receive more efficient and accurate treatment, emphasizing that personal plans and therapies can be developed faster with AI by providing details on risks and medical history.

No Other Choice

“There is indeed a real danger that in 30 to 50 years’ time, if we develop machines that work for us, people will no longer want to think. However, there is no other solution: it is simply impossible to do a lot of the day-to-day work in healthcare with human resources alone,” says Madácsy.

At the Endo-Kapszula Magánorvosi Centrum, AI is used to segment data in image assessment in capsule endoscopy. The machine detects pathological abnormalities by selecting the few hundred pictures (out of an average of 35,000) that contain abnormalities with 99% accuracy. This saves time as the human professionals only have to examine the flagged images. Madácsy states that AI can reduce the diagnostic time from one and a half hours to 10 minutes.

“This also helps to eliminate human errors such as possible misdiagnosis. After an eight or even 10-hour shift, the radiologist’s or endoscopist’s eyes and brain can get tired and may not recognize the images where the lesion is visible,” adds Madácsy.

Other key tasks for which AI is commonly used at the center are lung screenings and the detection of polyps. Tumors smaller than five millimeters can be detected by AI, while the inaccuracy rate for identifying polyps can be reduced by 50%.

Madácsy predicts that AI will soon also automatically diagnose CT, MR, and capsule scans. Beyond that, he believes AI will ultimately be able to perform surgeries and endoscopies while doctors merely oversee the process.

"In my opinion, in a few years it will not be a question of whether we use AI as doctors, but practically, those who don’t will be left behind,” says Madácsy.