Sentiment analysis in medical settings: New opportunities and challenges
The researchers explored current state of the art techniques of sentiment analysis in the medical domain and potential benefits of sentiment analysis in medicine, especially with regard to clinical documents. Clinical documents can reflect a physician’s opinion about the patient’s health status. Identifying the opinion and the target of the opinion (i.e. the aspect of the patient’s health status the physician is talking about) in clinical documents can be important to access clinical data, monitor a patient’s health status or provide automated decision support for physicians
Notions of sentiment in medicine and sentiment analysis in a medical context
The researchers identified five areas of medicine regarding clinical documents that sentiment is expressed: health status, medical condition, diagnosis, effect of a medical event, medical procedure and medication. According to the article, sentiment analysis in the medical domain can be categorized based on the source of the document, the task, the method and the level. With regards to the document source it can be carried out on medical web content, biomedical literature, clinical note . In terms of the task; polarity analysis and outcome classification. Two methods exist for carrying out sentiment analysis (rule-based and machine learning), while sentiment analysis can be carried out at the word level or sentence level.
Applications areas of sentiment analysis to clinical documents
- Health status aggregation
- Quality Assessment
- Outcome research
- Mining and retrieving personal health information and opinions
- Analyzing emotions and studying emotional effects
- Adverse drug events detection
- Analyzing emotions and the study of emotional affects
Medical sentiment analysis is an adequate way of determining and measuring the level of satisfaction the patient has towards a hospital or clinic after a procedure or health outcome.
- Denecke, K., & Deng, Y. (2015). Sentiment analysis in medical settings: New opportunities and challenges. Artificial Intelligence in Medicine. Retrieved from http://www.sciencedirect.com/science/article/pii/S0933365715000299