ADVANCES IN THE DIAGNOSIS OF HEMATOLOGICAL DISEASES WITH ARTIFICIAL INTELLIGENCE

Resumo

Advances in the diagnosis of blood diseases through the application of Artificial Intelligence (AI) are increasingly being explored in the healthcare field. AI enables rapid and accurate analysis of laboratory tests and images, identifying patterns that are often not noticeable to human observation. The objective of this study is to present the benefits and applications of AI in the diagnosis of hematological pathologies, such as leukemia and sickle cell anemia, highlighting its potential to enable earlier and more accessible diagnoses. The study is a literature review... bibliographic survey of recent studies on the use of machine learning techniques and convolutional neural networks in the analysis of blood counts and microscopic images, including the use of explainable Artificial Intelligence. The results indicate that AI can achieve accuracy greater than 95% in hematological diagnoses, with emphasis on models that achieved 98% accuracy in detecting acute lymphoblastic leukemia and 96% in identifying sickle cell anemia. In addition, its implementation in low-infrastructure contexts, through mobile devices, demonstrates potential for expanding access to diagnosis. It is also observed that the incorporation of AI in primary care increases resolution and improves clinical screening. It is concluded that Artificial Intelligence represents a significant advance for hematology, providing faster, more economical, and more reliable diagnoses. However, its full adoption requires clinical validation, professional training, and ethical regulation, ensuring that the technology acts as a support to medical work and contributes to greater equity in health.