[In the Laboratory of Systems Neuroscience of the Institute of Experimental Medicine, our main objective is the better understanding of the brain mechanisms of cognitive functions. We investigate the nervous system basis of learning, memory, attention and decision-making not only in normal conditions but also in pathological ones, especially concerning the neurodegenerative dementias such as Alzheimer’s and Parkinson’s dis­ease. 



Our main profile is conducting animal studies, but we carry out hu­man research too in collaboration with clinical partners. We analyse multichannel, i.e. multidimensional electrophysiological signals using both classical and modern statistical methods and new and powerful machine learning algorithms. 



This study presents the activity of our team, while focusing on artificial intelligence app­lications.

]

The aim of the study was to investigate the question: Can MRI radiomics analysis of the periaqueductal gray region elucidate the pathophysiological mechanisms underlying various migraine subtypes, and can a machine learning model using these radiomics features accurately differentiate between migraine patients and healthy individuals, as well as between migraine subtypes, including atypical cases with overlapping symptoms?

[Indication for recanalization therapy of acute ischemic stroke is based on imaging procedures. In order to minimize the time loss passing by recog­nizing the condition and the transfer of images to other facility, we established a stroke imaging network (eStroke network) supported by Artificial Intelligence (AI) in Hungary. Our study aims to present this system. 

Organized by the National Institute of Mental Health, Neurology, and Neurosurgery (NIMNN), we included a total of 28 stroke centers, among them 4 thrombectomy centers. An earlier network of the University of Pécs and the widened network parallel with the NIMNN project cover 10 centers thus the service is now available in 38 stroke centers of this country. Stroke CT scans are automatically transmitted via the central teleradiology server to a central image processing server which analyzes the size of the ischemic area (ASPECT score), detects large vessel occlusion and it’s localization, analyzes the quality of collateral circula­tion and standard CT perfusion parameters using an AI based software (eStroke, Brainomix Ltd.). Results and processed images are sent automatically back to the PACS system of the sending institution and that of the concerning thrombectomy center and become available in anonymous form via cloud by desktop computers or mobile application. 

During the first year of opera­tion, the system has processed 38,060 scans of 16,276 patients. In NIMNN experience by samples of 65 and 152 cases, for drip and ship patients the time from the first alerting of the ambulance service, until arrival at the thrombectomy center was reduced by 38 minutes from 4:18 to 3:40 minutes. 

Building an AI based central stroke imaging network for improving of stroke care’s results is technically feasible. Operation of the eStroke system is capable of reducing patient transportation times, however, further optimization is needed. ]

[The paper discusses latest developments in the field of artificial intelligence (AI) and its applications for the future of nursing care and services. Authors emphasize that AI should not be treated as a threat to nurses but as a potential support that may relieve nurses from performing routine tasks which can be automated by AI and robotics. The article cites examples from international experiments with AI and nursing robots that make the case for the controlled and intelligent use of technology in support of a nursing workforce that experiences pressures and stress because of increasing workloads and decreasing number of nursing staff. Authors call for the recognition that nurses fall behind in participating in the development of future nurse specific technologies, also due to the fact that nurses receive no education in the field of AI and robotics. Authors conclude that for nurses to stay in the driving seat of future nurse technologies, graduate and/or postgraduate level education for nurses in AI and robotics is a must.]

[The progress of molecular information based on personalized precision medicine has reached a new milestone. Actually, about 6 million mutations of 600 genes may be related to the development of cancer, and on average, 3-4 of these “driver” mutations are present in each patient. Due to the progress in molecular diagnostics, we can now routinely identify the molecular profile of tumors in clinical settings. By clinical translation, there are actually available more than 125 targeted pharmaceuticals and hundreds of such therapies are under clinical trial. As a result, we have many first-line and licenced treatment options to be elected by molecular information as the optimal one for every patient. There is an increasing need for complex informatics solutions by medical software. Geneticists, molecular biologists, molecular pathologists, molecular pharmacologists are already using bioinformatics and interpretation software on their daily work. Today, online digital tools of artificial intelligence are also available for physicians for assisted treatment planning. Telemedicine, videoconferencing provide solutions for interdisciplinary virtual molecular tumor boards, which democratizes the access to precision oncology for all doctors and patients. ]