Speaker Class C
Future prospects for artificial intelligence and machine learning in medicine
The widespread adoption of artificial intelligence (AI), in particular the deep-learning technique, made possible by the unrestricted use of databases, the utilization of big data, significantly increased processing power, and cloud storage. With its complex challenges and tasks, the medical sector is just beginning to realize an impact of AI on at multiple levels. For clinicians, AI supports rapid, accurate laboratory, genetic and image interpretation. Patients get access and follow their own data to enhance health, which benefits both patients and practitioners. Furthermore, it improves health systems by enhancing workflow and the ability to reduce administrative and other medical errors. At population level, AI will be extremely efficient in prediction of epidemics and pandemics. Additionally, by learning more about the information contained in the "black box" of several convolutional artificial neural networks might shed light on previously obscure biological pathways. Machine learning (ML) is a tool that can help with understanding radiological, endoscopic, and histological images, making diagnoses, forecasting the course of diseases, and even advising on treatment and surgical options. The internal layers of ML algorithms' interpretability should be improved for xAI (explained AI) medicine to be successfully implemented. Without being specifically programmed for each task (supervised or unsupervised), ML enables computers to learn. Continuous self correction of AI operation by backward propagation results in increasingly precise performance, accuracy, and speed. Vital signs and test findings are examples of early disease indicators that AI systems might assist detect. AI will speed up the drug discovery process by analyzing and forecasting billions of data points to generate new medications and efficacy of vaccines. For each patient, artificial intelligence redefines personalized therapies and reduces potential negative side effects, through the simultaneous combination of genetic, epigenetic and image data.
The robustness, usability, and utility of AI will undoubtedly transform many aspects of our existence, including clinical research and medicine.
Biography
Andras Falus (born 1947), PhD, DSc, med. habil. Professor Emeritus of Immunology and Genetics at Dept Genetics, Cell and Immunobiology, Semmelweis University, Budapest.
Member of Hungarian Academy of Sciences and Academia Europeae. Former President of Hungarian Society of Immunology. Founder of Hungarian Biobanking system. Member of Henry Kunkel Society of Rockefeller University, NY. Founder Editor of Immunome Research, earlier board member of Autoimmunity, Inflammation Research and Cellular Molecular Life Sciences.
Chief Editor of Hungarian Science (official periodical of Hungarian Academy of Sciences) between 2017-2023. Founder of EDUVITAL, a nonprofit Health Educational Society.
Interest: systems biology, epigenetics, immunogenomics, artificial intelligence in medical biology, microbiome, public health, peer education.
Fellowships: Odense (1980-81), Boston (1984-86), Osaka (1989), Bern (1991).
Peer reviewed journal articles: 435, books: 8, book chapters: 23.
Total citations: ~ 18.350, H-index : 61 (Google Scholar), 49 (Scopus)
Website: http://gsi.semmelweis.hu/index.php/en/about-us/staff/174-falusandras