A Comparative Study on Text Summarization in Healthcare Domain

As online text data is enormously increasing, finding the relevant data is becoming crucial. Automatic text summarization technology helps provide relevant data summaries quickly. Especially in the field of healthcare, medical records consist of various information such as observation records, diagnostic test reports, discharge summaries, prescriptions, etc., so doctors may not have time to carefully read the concise. The tailored medical text summarization aids in providing points of care to the patients. Several methods are available in literature, but no method is superior in performance. This study is focused on medical text summarization methods such as SEQ2SEQ, Bio-Bert, BERTSUMTEXT, BART, multi indicator text summarization system (MINTS), encoder-decoder with attention (EDA), and multi-objective evolutionary algorithm (MOEA). According to this review multi-objective evolutionary algorithm is identified with highly efficient model with their ROUGE values. Further we can improve the performance of medical document summarization using deep learning techniques in medical documents.