It increasingly helps doctors analyze results, it is fast and precise, and in diagnostics it cannot be overestimated. Its advantages are especially appreciated by radiologists.
This is happening now – research effectiveness close to 100 percent, results within a few days, and all thanks to artificial intelligence.
– We simply collect data in a faster, perhaps a little rudimentary, way, and artificial intelligence, thanks to its intelligence, allows us to prepare the images that we then see faster and simply collect the material faster – says Professor Piotr Grzelak, head of the Department of Imaging Diagnostics at the Institute of the Polish Mother's Health Center in Łódź.
The widest application for AI has been found by radiologists. Pixel Technology from Łódź has developed a system that analyzes images, for example those obtained as a result of an X-ray or tomography examination.
The most important thing in this case is speed and accuracy, because in many cases the waiting time for descriptions is very long.
Artificial intelligence helps doctors in about 10 hospitals across the country
– We have created algorithms that can speed up the description of some tests by up to 74 percent, i.e. they automatically find all pathological changes in given organs. In some models, the best ones we have, we currently have 96-97 percent effectiveness. Super Doctor has 94 percent, while ordinary doctors have from 72 to 84 percent effectiveness – says Jakub Musiałek from Pixel Technology in Łódź.
The systems are already helping doctors in about 10 hospitals across the country. – Any system that will also speed up the work of a radiologist, that's great, that's very beneficial and could relieve us a little – says Tomasz Puzio, a radiologist from Pixel Technology.
Healing can become faster than ever
Radiology is not the only field of medicine in which AI helps doctors. In the Provincial Specialist Hospital of Biegański in Łódź, it is also used by cardiologists.
A special platform for hemodynamic measurements can help, for example, predict pressure drops in operated patients. – We are able to know what will happen in the future, i.e. prevent adverse events that we do not want patients to experience. We can treat them before they occur, because without these algorithms we are able as doctors to treat something that is already happening – says Arkadiusz Ammer, MD, from the Provincial Specialist Hospital in Łódź.
Thanks to artificial intelligence, analyzing a patient's medical history and adapting treatment can become faster than ever.
– Considering the potential of artificial intelligence in processing huge amounts of various types of data, combining this type of information about the patient provides great potential for personalizing therapy – says Professor, PhD, Engineer Anna Fabijańska from the Institute of Applied Computer Science, Lodz University of Technology.
In the future, we can expect a scenario in which fewer tests will be needed and they will provide doctors with more information.
– We will be doing increasingly more detailed tests. We will learn more and more from a single test and doctors with such an artificial intelligence system will create a tandem that will quickly go through the standard thousands of examples of pathology already known to us – predicts Jakub Musiałek.
When using AI, doctors should nevertheless be guided by knowledge and experience.
– Playing with artificial intelligence in medicine has somewhat reached a dead end, where we are looking for very medical solutions, diagnoses that sometimes the doctor makes intuitively. Unfortunately, artificial intelligence usually means binary choices and there is no room for intuition here – Piotr Grzelak points out.
– Of course, the doctor makes the decisions and in fact the system can only suggest at most, or which test should be assessed first, because, for example, the system suspects that there is some change that requires urgent assessment – comments Tomasz Puzio.
Artificial intelligence algorithms can only speed up or facilitate some processes.
Main image source: TVN24