No abstract

PurposeThe breasts are mainly fatty and connective tissues with no muscles that directly support them, so wearing sports bras is one of the most effective means of alleviating the discomfort of breast movement and potential injury during vigorous physical exercise. However, the design and development processes of traditional sports bras are time-consuming and costly. Hence, a novel method of simulating the static contact pressure between a sports bra and women’s body based on the finite element (FE) and artificial neural network (ANN) models is developed in this study to contribute to the design considerations of sports bras.Design/methodology/approachThree-dimensional FE models of a female subject and sports bras with different fabric properties are developed to determine the amount of contact pressure exerted onto the body. The FE results are then verified by measuring the amount of pressure exerted by the sports bra on the skin with pressure sensors. The Taguchi technique is used to effectively reduce the number of trials from 625 to only 25 cases. These 25 results obtained through FE modelling are then used to provide the training set for the ANNs. Finally, a comparison between the FE and ANN results is carried out.FindingsA novel model of the static contact pressure between a sports bra and human subject based on the FE and ANN methods is presented in this paper. The root mean square error values show that there is only a small difference between the FE and ANN results.Originality/valueThe ANN function established in this study can be used to predict the mechanical behaviours of breasts and has a fundamental impact on the computer-aided design of functional garments in general.

A large number of studies have used electromyography (EMG) to measure the paraspinal muscle activity of adolescents with idiopathic scoliosis. However, investigations on the features of these muscles are very limited even though the information is useful for evaluating the effectiveness of various types of interventions, such as scoliosis-specific exercises. The aim of this cross-sectional study is to investigate the characteristics of participants with imbalanced muscle activity and the relationships among 13 features (physical features and EMG signal value). A total of 106 participants (69% with scoliosis; 78% female; 9–30 years old) are involved in this study. Their basic profile information is obtained, and the surface EMG signals of the upper trapezius, latissimus dorsi, and erector spinae (thoracic and erector spinae) lumbar muscles are tested in the static (sitting) and dynamic (prone extension position) conditions. Then, two machine learning approaches and an importance analysis are used. About 30% of the participants in this study find that balancing their paraspinal muscle activity during sitting is challenging. The most interesting finding is that the dynamic asymmetry of the erector spinae (lumbar) group of muscles is an important (third in importance) predictor of scoliosis aside from the angle of trunk rotation and height of the subject.

The spine is an important part of the human body. Thus, its curvature and shape are closely monitored, and treatment is required if abnormalities are detected. However, the current method of spinal examination mostly relies on two-dimensional static imaging, which does not provide real-time information on dynamic spinal behaviour. Therefore, this study explored an easier and more efficient method based on machine learning and sensors to determine the curvature of the spine. Fifteen participants were recruited and performed tests to generate data for training a neural network. This estimated the spinal curvature from the readings of three inertial measurement units and had an average absolute error of 0.261161 cm.

No abstract