Applications based on Artificial Intelligence and Machine Learning could soon be encountered regularly in our medical practices. Patients themselves probably won't notice, but for doctors, AI will make a big difference in diagnostics.
AI and Machine Learning will intervene where doctors have the most work: in finding the right diagnosis. For many conditions, diagnostics may be limited to looking at a single lab value or image. But in more complex cases, all the available data must be combined to form a meaningful picture from which the right conclusions can be drawn.
This does not only take time but is also tedious. This is where AI applications come in, because unlike humans, they can work without pausing. AI recognizes patterns in data and can use them to infer certain results. In the process, of course, an AI does not make its own diagnoses, but merely suggests diagnoses to a human, which in turn must confirm them.
In our second dive on 13 October, 2022, we jumped into the depths of AI and looked at how it makes the step from development to practice. Once again, three experts supported us with their experience in development and application.
Dr. Jan Alexandersson
AI is experiencing a surge of interest, and the medical world is eager to see if AI can take it to the next level. However, with all the enthusiasm for AI, the topic should also be viewed critically. Dr. Jan Alexandersson, head of the AAL Competence Center at the German Research Center for Artificial Intelligence, warns that we should mind the correct use of AI: "We have to be careful: AI is like a knife you can cut bread with. But you can also cut a throat with it, so we have to be very, very careful how you use it."
Dr. Boris Stanzel, head of clinical research and the Saar Macular Center, Sulzbach Eye Clinic, is excited about the possibilities AI offers. According to him, AI could outperform practicing physicians. However, it is important to mention that all participants agreed that AI cannot and should not replace humans, rather it should act in a supportive way: "AI can recognize more accurately, and with greater reliability, than, for example, a retina specialist with 15 years of professional experience. He or she is even surpassed by AI."
The Marburger Bund, the largest association of physicians in Germany, states in its MB Monitor 2022 that 32% of all physicians estimate that four hours a day are spent on administrative tasks. This time expenditure is also a thorn in Dr. Stanzel's side: "The time for patient care is impaired by the administrative burden. AI could solve this problem, leaving more time for patients."
Dr. Boris Stanzel
The fact that AI is supposed to suggest diagnoses means that close attention must be paid to how it makes decisions. Alexandre Barral, International R&D and Growth Manager at Portal Telemedicina, considers AI transparency to be very important. "AI is not like Netflix. We do not need to know why Netflix is showing us this particular movie, but in the healthcare sector, doctors or patients need to know why AI is displaying these kinds of results and how it is doing it." Barral continues, "AI cannot be like a black box, we need transparency."