Living in the era of technology, almost every one of us uses smart devices that make life easier for us. However, these technologies are not only useful in everyday life, but are increasingly invading the medical sector. Each of us now has the opportunity to monitor our health on a daily basis, and in case of a malfunction , to receive notifications to our smart devices and seek help in a timely manner. And when appearing in the doctor’s office, many would be surprised – this one uses even more advanced technology in his workplace. This is artificial intelligence, with the help of which, it turns out, the work of doctors becomes much more effective. But can we and are we willing to entrust technology to our most precious asset – health?
Artificial intelligence in the operating room
The technological development is truly impressive. Technologies that serve health monitoring save lives, but it is also no secret that various technologies not only help to independently monitor the state of health, its various changes and avoid danger in a timely manner, but also directly help to treat patients. One such technological tool is artificial intelligence (AI), which has infinite possibilities that help not only in everyday life, but also in medicine.
As with smart devices, as with AI, entrusting human health to these technologies still seems unsafe to some, but according to Dr. Donato Jocius, an interventional radiologist at VUL Santaros Clinics, AI inevitably penetrates the medical sector. Although AI has become more familiar to the general public until recently, and its functionality is being tested by an increasing number of employees, students and schoolchildren from different sectors, the radiologist explains that the predecessors of this tool are not very new in radiology.
“Diagnostic radiological systems have various protocols for automatic segmentation, vascular detection tools, and so on. Currently, there are AI systems that help detect tumors, evaluate their different parameters. AI is also entering the scope of patient treatment – there are various surgical robotic systems that facilitate difficult tasks, allow you to accurately perform specific movements, theoretically reduce the likelihood of mistakes, fatigue factor.
Interventional radiology is no exception – robotic or partially robotic systems have been developed to carry out the treatment of tumors, which help to accurately position the needle, predict the future field of treatment, evaluate the effect of the procedure performed here and now – in the operating room, and, if necessary, perform additional actions.
More broadly, AI can or is used in the development of medicines, in the monitoring of the health system, in the diagnosis of diseases, in the management and analysis of health data, in the making of personalized decisions, in the application of both surgical and medical treatments – this is very wide,” the doctor lists the unlimited possibilities of this tool.
When asked if AI can completely replace the therapeutic procedures currently being performed, the doctor says it’s hard to judge — time will tell. Nevertheless, according to him, AI can already increase the accuracy of the diagnostic and therapeutic procedures performed, reduce the time of the procedure, reduce the influence of human error and fatigue.
“AI can analyze and incorporate large amounts of complex complex data, recognize inclinations and nature of various alterations that can only be partially recognized by man. Another important aspect is that AI does not have a human inclination in decision-making, in other words, the human factor is eliminated. AI can increase efficiency, accuracy, and reduce workload, which would substantially increase the time spent on the patient, as well as the focus of doctors on difficult cases.
From an interventional radiology perspective, DI is useful in several ways. Firstly, in the planning of procedures, tools such as virtual reality allow pre-procedural planning, accurate assessment of complex anatomy or simulation of the procedure and its outcome, without any potential risk to the patient. And here, during the procedure, AI can serve in many ways. In the treatment of vascular stenoses (constrictions) or aneurysms (dilations), it can help to select the right vascular stents or stentgrafts, in complex cases it helps to assess the lumen of blood vessels, to predict the future effect of treatment.
When performing three-dimensional (computed tomography, magnetic resonance imaging or positron emission tomography) performed before the procedure, the automatic fusion with working intraoperative 2D or computed tomography images performed with the help of AI allows to increase the accuracy and precision of both diagnostic procedures, for example, biopsies, and various therapeutic procedures, for example, tumor ablations. In such cases, AI systems not only help to choose the optimal path, but at the same time are also able to assess changing physiological aspects, such as the reimbursement of movement when the patient breathes , which increases the accuracy of the procedure, shortens its time,” explains the doctor.
For doctors, AI is not a competitor, but rather an assistant
However, skeptics are partly right – this tool has not only advantages, but also disadvantages that can be questioned when discussing whether AI is capable of providing medical assistance by avoiding mistakes or completely replacing doctors. Therefore, this tool is still still only an opportunity to achieve higher results, and human hands and mind in medicine, at least for the time being, are an inevitable factor.
“In various literary sources, we can also find claims that AI will probably never reach the level of holistic thinking of a person. AI needs to see an object hundreds or thousands of times in order to receive constant repetitive feedback and subsequently be able to recognize it, and a person can do this simply by observing the object even once.
This is especially true when it comes to rare diseases or pathological changes that do not occur infrequently – such changes may not be detected by AI systems. In addition, AI ‘learns’ using certain data, so using an AI system to analyze a slightly different type of data can lead to errors,” he says.
The doctor also emphasizes that one of the main disadvantages of AI is the absence of a “life” experience that is needed to solve realistic tasks, assessing everything in an abstract and generalized way.
“The issue of AI liability is another pressing and complex problem – algorithms can make mistakes, inaccurately assess pathology or choose a treatment – in today’s medicine, this inevitably leads to legal and liability problems,” says Dr. Jocius.
Read more: Delfi.lt
