Ergonomic design of mastectomy bra based on emotion measurements

Abstract

The mastectomy bra (M-bra) serves as an essential rehabilitative product for individuals who have undergone breast cancer surgery. This study aimed to develop an ergonomic design method for M-bras that addresses both the physiological and psychological needs of post-mastectomy breast cancer patients. Over 270 M-bras with diverse design options were collected from online markets, and more than 100 post-mastectomy patients participated in an online survey to rank their preferences regarding eight M-bra components and 19 design options. An orthogonal experimental design was applied to reduce the vast array of design combinations to 20 M-bra test samples. The Pleasure-Arousal-Dominance (PAD) emotion model and the Self-Assessment Manikin (SAM) scale were then used to collect PAD scores from 40 patients evaluating these test samples. Extreme difference analysis was conducted to evaluate the significance of the M-bra components and identify combinations that generated the highest and lowest PAD levels. Based on the PAD scores, 11 samples were categorized as "delighted" or "relaxed" M-bras due to their positive emotional polarity. Linear regression analysis showed strong correlations between the design options and PAD scores, with R² values of 0.962 for P, 0.819 for A, and 0.949 for D. A backpropagation neural network was developed to predict PAD values for various M-bra design options, achieving prediction errors of 8.06% for P, 6.15% for A, and 13.29% for D, demonstrating satisfactory performance given the subjective nature of the evaluations.