Natural language processing (NLP) is one of the most significant areas of research in artificial intelligence. In this paper, I have overviewed the processing and analysis of clinical texts that contain a large amount of useful information to support decision-making in medicine. But text analysis is connected with a number of complications, such as text markup, preprocessing, and various types of analysis. Moreover, clinical texts are particularly difficult to analyze for many reasons: frequent abbreviations, typos, and a large number of synonymous terms. To solve this problem various approaches are to be applied, including expert-created rules, machine learning, and deep learning as a more recent approach. A more formal description of the task mentioned above is named entity recognition (NER) in clinical records. To solve this problem I have developed methods to extract information on diseases and health conditions based on manually created rules, statistical machine learning and deep learning. Comparative experimental studies of the developed methods are conducted on a marked-up corpus of clinical records. Based on them, conclusions are made about the effectiveness of the developed methods.
Download PDF from the RUDN University website (in Russian): https://events.rudn.ru/event/107/papers/509/files/879-ittmm_submission_Kireev.pdf
Download all conference proceedings (PDF) at eLibrary (in Russian, registration required): https://elibrary.ru/item.asp?id=46236354
Kireev D. A. Intelligent analysis of clinical texts // Information and Telecommunication Technologies and Mathematical Modeling of High-Tech Systems: Proceedings of the All-Russian Conference with International Participation, Moscow, RUDN University, 19–23 April 2021. Moscow: RUDN, 2021, pp. 209–213.