Inteligenţa artificială în tehnicile de reproducere asistată

Abstract

Artificial intelligence (AI) represents a continuously expanding technique, recording real progress in multiple medical specialties. The improvement of calculation algorithms, the appearance of more and more powerful computers, the training of doctors to use this equipment, and data processing techniques have contributed to the increased performance of assisted reproduction technology. Currently, implementing AI algorithms in the clinical activity of assisted reproductive technology (ART) represents a challenge. There are a series of limitations of the deep learning and decision tree processes due to the risks regarding the validation of the results, the ethical aspects, the responsibility for the correctness of the information, the doctors’ reluctance to use AI, and the lack of empathy for patients. However, the continuous research on the AI algorithms is essential in clinical embryology and in assisted human reproduction techniques. The early diagnosis of infertility, the personalization of treatment schemes, and the prediction of ART success are just a few directions for promoting AI applications in reproductive medicine, without being able to replace the current clinical model. The possibility of making predictions regarding the quality of embryo transfer procedures and the normal pregnancy rate are factors that show the future direction of the development of in vitro fertilization techniques.