Gender Fairness of Machine Learning Algorithms for Pain Detection

TL;DR

This study evaluates the gender fairness of machine learning and deep learning algorithms for automated pain detection from facial expressions, revealing persistent biases and highlighting the need for fairness-aware methods.

Contribution

It provides a comparative analysis of traditional ML and DL models on pain detection fairness across genders using the UNBC dataset.

Findings

ViT achieved highest accuracy and fairness metrics

All models showed gender-based biases

Trade-off observed between accuracy and fairness

Abstract

Automated pain detection through machine learning (ML) and deep learning (DL) algorithms holds significant potential in healthcare, particularly for patients unable to self-report pain levels. However, the accuracy and fairness of these algorithms across different demographic groups (e.g., gender) remain under-researched. This paper investigates the gender fairness of ML and DL models trained on the UNBC-McMaster Shoulder Pain Expression Archive Database, evaluating the performance of various models in detecting pain based solely on the visual modality of participants' facial expressions. We compare traditional ML algorithms, Linear Support Vector Machine (L SVM) and Radial Basis Function SVM (RBF SVM), with DL methods, Convolutional Neural Network (CNN) and Vision Transformer (ViT), using a range of performance and fairness metrics. While ViT achieved the highest accuracy and a selection of fairness metrics, all models exhibited gender-based biases. These findings highlight the persistent trade-off between accuracy and fairness, emphasising the need for fairness-aware techniques to mitigate biases in automated healthcare systems.