Comparing shallow, deep, and transfer learning in predicting joint moments in running

Liew, Bernard X. W.; Rügamer, David; Zhai, Xiaojun; Wang, Yucheng; Morris, Susan; Netto, Kevin

doi:10.1016/j.jbiomech.2021.110820

Cited by 10 publications

(21 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Discussioncontrasting

confidence: 85%

“…Interestingly, our finding that transfer learning resulted in the best prediction performance was not supported by another study, albeit conducted in running ( Liew et al, 2021 ). In a previous study, the multivariate time-series kinematic predictors were transformed into static images using cubic spline interpolation ( Liew et al, 2021 ). The purpose of the interpolation was so that the predictor dimension fitted the input dimensions of the VGG16 image model used for transfer learning ( Liew et al, 2021 ).…”

Section: Discussioncontrasting

confidence: 85%

“…In a previous study, the multivariate time-series kinematic predictors were transformed into static images using cubic spline interpolation ( Liew et al, 2021 ). The purpose of the interpolation was so that the predictor dimension fitted the input dimensions of the VGG16 image model used for transfer learning ( Liew et al, 2021 ). For example, from an original array of 490 [observations] × 101 [gait cycle] × 9 [variables] × 3 [axes], the data was transformed into a 490 × 150 × 150 × 3 shape ( Liew et al, 2021 ).…”

Section: Discussionmentioning

confidence: 99%

“…Transfer learning achieved a Pearson correlation of 0.94–0.97 ( Johnson et al, 2019 ), which was similar to the correlation of 0.96 when using a custom CNN model ( Wang et al, 2020 ) for predicting knee abduction/adduction moments during walking. Interestingly, a direct benchmark study in running reported that transfer learning resulted in poorer correlation compared to a custom CNN model ( Liew et al, 2021 ). However, the previous study converted the time-series to images for modelling which could have introduced “noise” to the original data ( Liew et al, 2021 ).…”

Section: Introductionmentioning

confidence: 99%

“…Interestingly, a direct benchmark study in running reported that transfer learning resulted in poorer correlation compared to a custom CNN model ( Liew et al, 2021 ). However, the previous study converted the time-series to images for modelling which could have introduced “noise” to the original data ( Liew et al, 2021 ). To our knowledge, no studies have benchmarked the performance of different state-of-art time-series and image-based network architectures, and transfer learning for predicting knee joint moments in walking.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

Boukhennoufa

Altai

Zhai

et al. 2022

Front. Bioeng. Biotechnol.

Self Cite

View full text Add to dashboard Cite

Knee joint moments are commonly calculated to provide an indirect measure of knee joint loads. A shortcoming of inverse dynamics approaches is that the process of collecting and processing human motion data can be time-consuming. This study aimed to benchmark five different deep learning methods in using walking segment kinematics for predicting internal knee abduction impulse during walking. Three-dimensional kinematic and kinetic data used for the present analyses came from a publicly available dataset on walking (participants n = 33). The outcome for prediction was the internal knee abduction impulse over the stance phase. Three-dimensional (3D) angular and linear displacement, velocity, and acceleration of the seven lower body segment’s center of mass (COM), relative to a fixed global coordinate system were derived and formed the predictor space (126 time-series predictors). The total number of observations in the dataset was 6,737. The datasets were split into training (75%, n = 5,052) and testing (25%, n = 1685) datasets. Five deep learning models were benchmarked against inverse dynamics in quantifying knee abduction impulse. A baseline 2D convolutional network model achieved a mean absolute percentage error (MAPE) of 10.80%. Transfer learning with InceptionTime was the best performing model, achieving the best MAPE of 8.28%. Encoding the time-series as images then using a 2D convolutional model performed worse than the baseline model with a MAPE of 16.17%. Time-series based deep learning models were superior to an image-based method when predicting knee abduction moment impulse during walking. Future studies looking to develop wearable technologies will benefit from knowing the optimal network architecture, and the benefit of transfer learning for predicting joint moments.

show abstract

Section: Discussioncontrasting

confidence: 85%

Section: Discussioncontrasting

confidence: 85%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

Boukhennoufa

Altai

Zhai

et al. 2022

Front. Bioeng. Biotechnol.

Self Cite

View full text Add to dashboard Cite

show abstract

Functional Data Analysis: An Introduction and Recent Developments

Gertheiss,

Rügamer,

Liew

et al. 2024

Biometrical J

View full text Add to dashboard Cite

Functional data analysis (FDA) is a statistical framework that allows for the analysis of curves, images, or functions on higher dimensional domains. The goals of FDA, such as descriptive analyses, classification, and regression, are generally the same as for statistical analyses of scalar‐valued or multivariate data, but FDA brings additional challenges due to the high‐ and infinite dimensionality of observations and parameters, respectively. This paper provides an introduction to FDA, including a description of the most common statistical analysis techniques, their respective software implementations, and some recent developments in the field. The paper covers fundamental concepts such as descriptives and outliers, smoothing, amplitude and phase variation, and functional principal component analysis. It also discusses functional regression, statistical inference with functional data, functional classification and clustering, and machine learning approaches for functional data analysis. The methods discussed in this paper are widely applicable in fields such as medicine, biophysics, neuroscience, and chemistry and are increasingly relevant due to the widespread use of technologies that allow for the collection of functional data. Sparse functional data methods are also relevant for longitudinal data analysis. All presented methods are demonstrated using available software in R by analyzing a dataset on human motion and motor control. To facilitate the understanding of the methods, their implementation, and hands‐on application, the code for these practical examples is made available through a code and data supplement and on GitHub.

show abstract

Integrating an LSTM framework for predicting ankle joint biomechanics during gait using inertial sensors

Xiang,

Gu,

Gao

et al. 2024

Computers in Biology and Medicine

View full text Add to dashboard Cite

Comparing shallow, deep, and transfer learning in predicting joint moments in running

Cited by 10 publications

References 31 publications

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

Functional Data Analysis: An Introduction and Recent Developments

Integrating an LSTM framework for predicting ankle joint biomechanics during gait using inertial sensors

Contact Info

Product

Resources

About