Articles | Open Access | https://doi.org/10.55640/

ARTIFICIAL INTELLEGENCE IN MEDICINE

Askarova Komila Ergashevna, Savronova Samira Dilshod kizi , Student, Gulistan State university, English philology

Published Date 2025-12-03

Pages 24-27

Abstract

Generative artificial intelligence (GAI) is increasingly capable of automating a wide range of biomedical tasks, from clinical decision-making to the design and analysis of research studies. Using machine learning and transformer-based architectures, GAI can produce relevant text, images, and audio data in response to user inputs. Unlike earlier biomedical deep-learning systems that relied heavily on large, general-purpose labeled datasets, emerging evidence indicates that GAI can achieve strong performance with smaller, domain-specific data. Additionally, AI training methods have evolved from fully supervised learning to more data-efficient approaches, including weakly supervised, unsupervised, and reinforcement learning techniques. The newest forms of GAI—such as agent-based systems, mixture-of-experts models, and reasoning models—have expanded these tools’ ability to handle complex, multi-step processes. This review summarizes recent technical developments in GAI, examines how the latest models could enhance healthcare for clinicians and patients, and discusses validation strategies through specific examples that highlight both the current challenges and future opportunities in the field.