<![CDATA[Pain detection is an important aspect in healthcare services, especially for patients who have difficulty communicating verbally, such as infants, critically ill patients, and individuals with neurological disorders. Advances in Natural Language Processing (NLP) have enabled automated analysis of medical text to support pain assessment more objectively and efficiently. This study aims to analyze the development of NLP methods used in pain detection through a Systematic Literature Review (SLR) approach. The study adopted the PRISMA guideline to identify, screen, and evaluate relevant articles obtained from Google Scholar and Scopus databases. The reviewed studies were published between 2017 and 2025. Based on the PRISMA selection process, 173 final articles met the inclusion criteria and were included in the final analysis. The results indicate that transformer-based models such as BERT, BioBERT, and ClinicalBERT achieved better performance compared to traditional machine learning and conventional deep learning methods. In addition, Electronic Health Records (EHR), clinical notes, and patient reports were identified as the most frequently used datasets in pain detection research. However, several challenges remain, including limited dataset availability, lack of evaluation standardization, and high computational requirements. This study is expected to provide a comprehensive overview of NLP-based pain detection and support future research in intelligent healthcare systems.]]>
Copyrights © 2026
