Application of artificial intelligence in pathology

The development of digital technologies and computer vision algorithms extends the possibilities of Artificial Intelligence application in pathology. Neural networks based on deep learning are being successfully developed and used to perform tasks related to the diagnosis and classification of tumors, identification of immunohistochemical markers and morphometry. The use of Artificial Intelligence not only contributes to the objectification of the diagnostic process, but also reduces the burden on the pathologists, allowing them to concentrate on more complex cases. Despite this, there are limitations to the introduction of neural networks into routine pathology practice, including financial and legal difficulties, as well as a distrustful attitude towards automatic diagnosis among doctors and patients. The literature review provides information on Artificial Intelligence, machine learning and neural network architecture, as well as their integration into the practice of a pathologist. The software products used for quantitative morphological studies, diagnosis and prognosis of diseases are listed. The set of developed AI-based programs indicates a significant interest and relevance of their use in pathological and anatomical practice and opens new frontiers in personalized medicine.