Algorithmic Bias in Clinical Populations—Evaluating and Improving Facial Analysis Technology in Older Adults With Dementia

Abstract

The need for the automated facial expression analysis arises in various clinical settings involving mental and physical health assessment of older adults. However, the effect of age (young versus old) and ability (healthy versus physical or cognitive impairment) on the performance of available methods has not yet been investigated. In this paper, we demonstrate a bias affecting the performance of common facial landmark detection and expression recognition algorithms on the faces of older adults with dementia. We also investigate the ways of mitigating this bias via the addition of representative training examples. Results show that landmark placement is less accurate when tested on the faces of individuals with dementia as compared to older adults who are cognitively healthy. Retraining or fine-tuning the methods with images of older adults’ faces improves the performance significantly, but the gap between older adults with versus without dementia persists. As the interest in using facial analysis methods in clinical applications grows, results of this study: 1) highlight the limitations of the existing models when applied to clinical populations and 2) shed light on methods of addressing these limitations as well as the need to develop algorithms designed to be fair with respect to variables such as age and ability.