Citation, DOI, disclosures and article data
At the time the article was created Candace Makeda Moore had no recorded disclosures.View Candace Makeda Moore's current disclosures
At the time the article was last revised Andrew Murphy had no recorded disclosures.View Andrew Murphy's current disclosures
Selection bias is a type of bias created when the data sampled is not representative of the data of the population or group that a study or model aims to make a prediction about. Selection bias is the result of systematic errors in data selection and collection. Practically-speaking selection bias often occurs when the sample size of data is incorrect or the assignment of patients or data to groups is non-random. There are many types of selection bias described in the medical literature, for example survival bias and survivorship bias which can have overlapping definitions.
Selection bias is of particular concern in terms of the development of artificial intelligence technologies 1. In terms of machine learning, training data to create an algorithm (as opposed to the sampled data in a study) may have a different distribution of pathology than the population for which the algorithm is intended.
- 1: Tang A, Tam R, Cadrin-Chênevert A, Guest W, Chong J, Barfett J, Chepelev L, Cairns R, Mitchell JR, Cicero MD, Poudrette MG, Jaremko JL, Reinhold C, Gallix B, Gray B, Geis R. Canadian Association of Radiologists White Paper on Artificial Intelligence in Radiology. (2018) Canadian Association of Radiologists journal = Journal l'Association canadienne des radiologistes. 69 (2): 120-135. doi:10.1016/j.carj.2018.02.002 - Pubmed