Automation bias is a form of cognitive bias occurring when humans overvalue information produced by an automated, usually computerized, system. Users of automated systems can fail to understand or ignore illogical or incorrect information produced by computer systems.
Computer programs may create erroneous information due to any number of problems ranging from hardware design to algorithmic bias. In fact in several cases, arguably the most famous of which are those of Therac-25 1 and therapy planning software from Multidata Systems International 2, faulty software led to patient deaths.
In diagnostic radiology, automation bias has been recognized as a problem, and some academic research has been designed to quantify it 3,4 as well as explore potentially mitigating factors 5,6. Automation bias is an increasing concern in radiology as the automation of much work in the field, especially that including creation of differential diagnoses through AI, evolves.
- 1. Fraass BA. Errors in radiotherapy: motivation for development of new radiotherapy quality assurance paradigms. (2008) International journal of radiation oncology, biology, physics. 71 (1 Suppl): S162-5.
- 2. Borrás C. Overexposure of radiation ther-apy patients in Panama: problem recognition and follow-up mea-sures. Rev Panam Salud Publica. 2006;20(2/3);173–87.
- 3. Alberdi E, Povykalo A, Strigini L, Ayton P. Effects of incorrect computer-aided detection (CAD) output on human decision-making in mammography. (2004) Academic radiology. 11 (8): 909-18.
- 4. Goddard K, Roudsari A, Wyatt JC. Automation bias: a systematic review of frequency, effect mediators, and mitigators. (2012) Journal of the American Medical Informatics Association : JAMIA. 19 (1): 121-7.
- 5. Jorritsma W, Cnossen F, van Ooijen PM. Improving the radiologist-CAD interaction: designing for appropriate trust. (2015) Clinical radiology. 70 (2): 115-22.
- 6. Drew T, Cunningham C, Wolfe JM. When and why might a computer-aided detection (CAD) system interfere with visual search? An eye-tracking study. (2012) Academic radiology. 19 (10): 1260-7.
Related Radiopaedia articles
- artificial intelligence (AI)
- imaging data sets
- computer-aided diagnosis (CAD)
- natural language processing
machine learning (overview)
- machine learning processes
- machine learning models
- visualizing and understanding neural networks
- common data preparation/preprocessing steps
- DICOM to bitmap conversion
- dimensionality reduction
- principal component analysis
- training, testing and validation datasets
- loss function
- optimization algorithms
- linear and quadratic
- batch normalization
- rule-based expert systems
- clinical trials
- descriptive studies
- Bayes' theorem
- sensitivity and specificity
- positive predictive value (PPV)
- negative predictive value (NPV)
- likelihood ratio (LR)
- normal distribution
- type I error
- type II error
- confidence interval
- ROC curve
- retrospective studies
- prospective studies
- analyzes of variance
- nonparametric statistics
- cognitive bias in image perception