The relevance of breast motions and gaits in running exercises

Jie Zhou, Qian Mao, Jun Zhang, Newman Lau, Jianming ChenFashion 2022 · DOI

Introduction

Exercise-induced breast pain is a significant barrier to physical activity, affecting over half of all women. While sports bras are designed to provide support, evaluating their effectiveness typically requires expensive 3D motion capture systems and labor-intensive data processing. This study explores a more accessible alternative: predicting breast motion using gait parameters detectable by wearable sensors. By establishing the relationship between the "torso driving force" (gait) and breast surface deformation, researchers aim to streamline the design and evaluation of high-performance sports apparel.

Methodology

The study recruited 13 unmarried Chinese females (75B cup size) to ensure consistency in breast tissue morphology. Participants ran at a controlled speed of 7 km/h while a Vicon motion analysis system tracked 16 reflective markers—10 positioned on the breasts and 6 on the lower limbs.

Defining Breast Regions

To capture the complex surface extension and contraction of the breast, the researchers focused on the distance alteration (Δd) between specific regions rather than simple point displacement.

Fig. 1: The breast regions and the distance markers used to calculate surface deformation.

As shown in Figure 1, markers $R_0$ and $L_0$ served as central points, with distances $d_{01}$ through $d_{04}$ representing the lateral, superior, medial, and inferior regions respectively. Gait was simultaneously quantified through four parameters: the rotation angles of the hip, knee, and ankle joints, and the foot's height off the ground.

Predictive Modeling

The core of the research involved comparing three variations of Back-Propagation (BP) neural networks to find the most accurate predictive model:

  1. Standard BP: A baseline feed-forward network.
  2. GA-BP: Optimized using a Genetic Algorithm to refine initial weights.
  3. PSO-BP: Optimized via Particle Swarm Optimization, a technique mimicking social behavior in nature to find global optima.

A dataset of 21,827 data points was generated, with 21,500 used for training and 327 for testing the models' predictive capabilities.

Key Insights and Biomechanical Findings

The analysis revealed several critical aspects of breast biomechanics during running:

  • The "2-for-1" Oscillation: A striking discovery was that breast motion occurs at twice the frequency of the gait cycle. For every single step cycle of the legs, the breasts complete two full oscillations. This high-frequency movement is a primary driver of discomfort.
  • Regional Asymmetry: Significant motion diversity was confirmed both between the left and right breasts and across different regions of the same breast. Lateral (outer) regions exhibited the highest activity, while medial (inner) regions remained relatively stable.
  • Gait-Breast Correlation: Using Gray correlation analysis, the researchers found a "moderate" correlation (degrees between 0.6 and 0.75) between lower limb movements and breast deformation. This validates the use of gait data as a proxy for predicting breast motion.

Results

The PSO-BP model outperformed the other architectures across all metrics. It achieved a coefficient of determination ($R^2$) of 84.58% and a Mean Absolute Error (MAE) of 0.2108. This error rate was 11% lower than the standard BP network and 3% lower than the GA-BP model. Interestingly, the model performed slightly better on the left breast, likely due to inherent physiological differences in motion dispersion.

Conclusion and Future Work

This research demonstrates that machine learning can effectively bridge the gap between easily measurable gait data and complex breast biomechanics. While the current study focused on a specific cohort (75B cup size) in a nude state to establish a baseline, it paves the way for "application-centered" design tools. Future iterations will incorporate various bra styles and running speeds, potentially allowing manufacturers to use simple wearable sensors to optimize sports bra support in real-world environments.

Summary updated

Abstract

Abstract The control of breast motions is a critical indicator to evaluate the comfort and function of sports bras. If the breast motions can be predicted based on the gait parameters detected by wearable sensors, it will more economical and convenient to evaluate the bras. Thirteen unmarried Chinese females with a breast cup of 75B were recruited in this study to investigate the regularity of breast motions and the relevance between breast motions and gaits during running exercises. The breast motion indicator is the distance alteration of breast regions. The gaits were described by the rotation angles of the hip, knee, ankle joints, and the foot height off the ground. Firstly, the Mann-Whitney U test and the Kruskal-Wallis H test were utilized to analyze the motion diversity among the eight breast regions. Then, the gray correlation analysis was applied to explore the relevance between breast motions and gaits. Finally, the back-propagation neural network, the genetic algorithm, and the particle swarm optimization algorithm were utilized to construct the prediction models for breast motions based on gait parameters. The results demonstrate that the same breast regions on the bilateral breasts and the different breast regions on the ipsilateral breasts present a significant motion diversity. There is a moderate correlation between breast motions and gait parameters, and the back-propagation neural network optimized by the particle swarm optimization algorithm performs better in breast motion prediction, which has a coefficient of determination of 84.58% and a mean absolute error of 0.2108.
BibTeX
@inproceedings{zhou2022relevance,
  author = {Jie Zhou and Qian Mao and Jun Zhang and Newman Lau and Jianming Chen},
  title = {{The relevance of breast motions and gaits in running exercises}},
  booktitle = {Fashion},
  year = {2022},
  doi = {10.1186/s40691-021-00283-4},
}
Figure