Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

The effect of time normalization and biomechanical signal processing techniques of ground reaction force curves on deep-learning model performance

Time-series data are common in biomechanical studies. These data often undergo pre-processing steps such as time normalization or filtering prior to use in further analyses, including deep-learning classification. In this context, it remains unclear how these preprocessing steps affect deep-learning model performance. Thus, the aim of this study is to assess the effect of time-normalization and filtering on the performance of deep-learning classification models. We also investigated the effect of amplitude scaling. Using a public dataset (Gutenburg Gait Database, a ground reaction force database of level overground walking at self-selected walking speed involving 350 healthy individuals), we trained convolutional neural network (CNN) and long short-term memory (LSTM) models to predict binary sex (male, female) using three-dimensional ground-reaction forces to which we applied different processing approaches: zero padding, interpolation to 100% of signal, filtering, and scaling (min-max, body mass). The results show that transformations resulted in differences in model performances. Highest performance was obtained using unfiltered data, zero-padding, and min-max amplitude scaling (F1-score of 91 and 87% for CNN and LSTM, respectively). Not filtering data and using min-max scaling generally improve performance for both model architectures. For interpolation, results are not consistent across model architectures. This study suggests that processing steps must be considered in applications where deep-learning classification performance is relevant.

Correlation between Harris hip score and gait analysis through artificial intelligence pose estimation in patients after total hip arthroplasty

BACKGROUND

Recently, open pose estimation using artificial intelligence (AI) has enabled the analysis of time series of human movements through digital video inputs. Analyzing a person's actual movement as a digitized image would give objectivity in evaluating a person's physical function. In the present study, we investigated the relationship of AI camera-based open pose estimation with Harris Hip Score (HHS) developed for patient-reported outcome (PRO) of hip joint function.

METHOD

HHS evaluation and pose estimation using AI camera were performed for a total of 56 patients after total hip arthroplasty in Gyeongsang National University Hospital. Joint angles and gait parameters were analyzed by extracting joint points from time-series data of the patient's movements. A total of 65 parameters were from raw data of the lower extremity. Principal component analysis (PCA) was used to find main parameters. K-means cluster, X-squared test, Random forest, and mean decrease Gini (MDG) graph were also applied.

RESULTS

The train model showed 75% prediction accuracy and the test model showed 81.8% reality prediction accuracy in Random forest. "Anklerang_max", "kneeankle_diff", and "anklerang_rl" showed the top 3 Gini importance score in the Mean Decrease Gini (MDG) graph.

CONCLUSION

The present study shows that pose estimation data using AI camera is related to HHS by presenting associated gait parameters. In addition, our results suggest that ankle angle associated parameters could be key factors of gait analysis in patients who undergo total hip arthroplasty.

Identifying underlying individuality across running, walking, and handwriting patterns with conditional cycle-consistent generative adversarial networks

In recent years, the analysis of movement patterns has increasingly focused on the individuality of movements. After long speculations about weak individuality, strong individuality is now accepted, and the first situation-dependent fine structures within it are already identified. Methodologically, however, only signals of the same movements have been compared so far. The goal of this work is to detect cross-movement commonalities of individual walking, running, and handwriting patterns using data augmentation. A total of 17 healthy adults (35.8 ± 11.1 years, eight women and nine men) each performed 627.9 ± 129.0 walking strides, 962.9 ± 182.0 running strides, and 59.25 ± 1.8 handwritings. Using the conditional cycle-consistent generative adversarial network (CycleGAN), conditioned on the participant's class, a pairwise transformation between the vertical ground reaction force during walking and running and the vertical pen pressure during handwriting was learned in the first step. In the second step, the original data of the respective movements were used to artificially generate the other movement data. In the third step, whether the artificially generated data could be correctly assigned to a person via classification using a support vector machine trained with original data of the movement was tested. The classification F1-score ranged from 46.8% for handwriting data generated from walking data to 98.9% for walking data generated from running data. Thus, cross-movement individual patterns could be identified. Therefore, the methodology presented in this study may help to enable cross-movement analysis and the artificial generation of larger amounts of data.

Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data

Walking gait data acquired with force platforms may be used for person re-identification (re-ID) in various authentication, surveillance, and forensics applications. Current force platform-based re-ID systems classify a fixed set of identities (IDs), which presents a problem when IDs are added or removed from the database. We formulated force platform-based re-ID as a deep metric learning (DML) task, whereby a deep neural network learns a feature representation that can be compared between inputs using a distance metric. The force platform dataset used in this study is one of the largest and the most comprehensive of its kind, containing 193 IDs with significant variations in clothing, footwear, walking speed, and time between trials. Several DML model architectures were evaluated in a challenging setting where none of the IDs were seen during training (i.e., zero-shot re-ID) and there was only one prior sample per ID to compare with each query sample. The best architecture was 85% accurate in this setting, though an analysis of changes in walking speed and footwear between measurement instances revealed that accuracy was 28% higher on same-speed, same-footwear comparisons, compared to cross-speed, cross-footwear comparisons. These results demonstrate the potential of DML algorithms for zero-shot re-ID using force platform data, and highlight challenging cases.

Diagnosis and prognosis of COVID-19 employing analysis of patients' plasma and serum via LC-MS and machine learning

OBJECTIVE

To implement and evaluate machine learning (ML) algorithms for the prediction of COVID-19 diagnosis, severity, and fatality and to assess biomarkers potentially associated with these outcomes.

MATERIAL AND METHODS

Serum (n = 96) and plasma (n = 96) samples from patients with COVID-19 (acute, severe and fatal illness) from two independent hospitals in China were analyzed by LC-MS. Samples from healthy volunteers and from patients with pneumonia caused by other viruses (i.e. negative RT-PCR for COVID-19) were used as controls. Seven different ML-based models were built: PLS-DA, ANNDA, XGBoostDA, SIMCA, SVM, LREG and KNN.

RESULTS

The PLS-DA model presented the best performance for both datasets, with accuracy rates to predict the diagnosis, severity and fatality of COVID-19 of 93%, 94% and 97%, respectively. Low levels of the metabolites ribothymidine, 4-hydroxyphenylacetoylcarnitine and uridine were associated with COVID-19 positivity, whereas high levels of N-acetyl-glucosamine-1-phosphate, cysteinylglycine, methyl isobutyrate, l-ornithine and 5,6-dihydro-5-methyluracil were significantly related to greater severity and fatality from COVID-19.

CONCLUSION

The PLS-DA model can help to predict SARS-CoV-2 diagnosis, severity and fatality in daily practice. Some biomarkers typically increased in COVID-19 patients' serum or plasma (i.e. ribothymidine, N-acetyl-glucosamine-1-phosphate, l-ornithine, 5,6-dihydro-5-methyluracil) should be further evaluated as prognostic indicators of the disease.

Machine Learning Approach to Support the Detection of Parkinson's Disease in IMU-Based Gait Analysis

The aim of this study was to determine which supervised machine learning (ML) algorithm can most accurately classify people with Parkinson’s disease (pwPD) from speed-matched healthy subjects (HS) based on a selected minimum set of IMU-derived gait features. Twenty-two gait features were extrapolated from the trunk acceleration patterns of 81 pwPD and 80 HS, including spatiotemporal, pelvic kinematics, and acceleration-derived gait stability indexes. After a three-level feature selection procedure, seven gait features were considered for implementing five ML algorithms: support vector machine (SVM), artificial neural network, decision trees (DT), random forest (RF), and K-nearest neighbors. Accuracy, precision, recall, and F1 score were calculated. SVM, DT, and RF showed the best classification performances, with prediction accuracy higher than 80% on the test set. The conceptual model of approaching ML that we proposed could reduce the risk of overrepresenting multicollinear gait features in the model, reducing the risk of overfitting in the test performances while fostering the explainability of the results.

Predicting the Internal Knee Abduction Impulse During Walking Using Deep Learning

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.

A Survey of Human Gait-Based Artificial Intelligence Applications

We performed an electronic database search of published works from 2012 to mid-2021 that focus on human gait studies and apply machine learning techniques. We identified six key applications of machine learning using gait data: 1) Gait analysis where analyzing techniques and certain biomechanical analysis factors are improved by utilizing artificial intelligence algorithms, 2) Health and Wellness, with applications in gait monitoring for abnormal gait detection, recognition of human activities, fall detection and sports performance, 3) Human Pose Tracking using one-person or multi-person tracking and localization systems such as OpenPose, Simultaneous Localization and Mapping (SLAM), etc., 4) Gait-based biometrics with applications in person identification, authentication, and re-identification as well as gender and age recognition 5) “Smart gait” applications ranging from smart socks, shoes, and other wearables to smart homes and smart retail stores that incorporate continuous monitoring and control systems and 6) Animation that reconstructs human motion utilizing gait data, simulation and machine learning techniques. Our goal is to provide a single broad-based survey of the applications of machine learning technology in gait analysis and identify future areas of potential study and growth. We discuss the machine learning techniques that have been used with a focus on the tasks they perform, the problems they attempt to solve, and the trade-offs they navigate.

Gutenberg Gait Database, a ground reaction force database of level overground walking in healthy individuals

The Gutenberg Gait Database comprises data of 350 healthy individuals recorded in our laboratory over the past seven years. The database contains ground reaction force (GRF) and center of pressure (COP) data of two consecutive steps measured - by two force plates embedded in the ground - during level overground walking at self-selected walking speed. The database includes participants of varying ages, from 11 to 64 years. For each participant, up to eight gait analysis sessions were recorded, with each session comprising at least eight gait trials. The database provides unprocessed (raw) and processed (ready-to-use) data, including three-dimensional GRF and two-dimensional COP signals during the stance phase. These data records offer new possibilities for future studies on human gait, e.g., the application as a reference set for the analysis of pathological gait patterns, or for automatic classification using machine learning. In the future, the database will be expanded continuously to obtain an even larger and well-balanced database with respect to age, sex, and other gait-specific factors.

Effect of Lubricant Type on the Friction Behaviours and Surface Topography in Metal Forming of Ti-6Al-4V Titanium Alloy Sheets

The aim of the research described in this paper is to analyse the synergistic effect of types of synthetic oil and their density on the value of the coefficient of friction (COF) of Ti-6Al-4V titanium alloy sheets. Lubrication performance of commercial synthetic oils (machine, gear, engine and hydraulic) was tested in a strip draw friction test. The friction tests consisted of pulling a strip specimen between two cylindrical fixed countersamples. The countersamples were placed in the simulator base mounted on a uniaxial tensile test machine. Due to the complex synergistic effect of different strip drawing test parameters on the COF, artificial neural networks were used to find this relationship. In the case of both dry and lubricated conditions, a clear trend was found of a reduction of the coefficient of friction with nominal pressure. Engine oil 10W-40 was found to be the least favourable lubricant in reducing the coefficient of friction of Grade 5 titanium sheets. The two main tribological mechanisms, i.e., galling and ploughing, played the most important role in the friction process on the test sheets. In the range of nominal pressures considered, and with the synthetic oils tested, the most favourable lubrication conditions can be obtained by using a type of oil with a low viscosity index and a high kinematic viscosity.

Machine learning-based prediction of survival prognosis in cervical cancer

BACKGROUND

Accurately forecasting the prognosis could improve cervical cancer management, however, the currently used clinical features are difficult to provide enough information. The aim of this study is to improve forecasting capability by developing a miRNAs-based machine learning survival prediction model.

RESULTS

The expression characteristics of miRNAs were chosen as features for model development. The cervical cancer miRNA expression data was obtained from The Cancer Genome Atlas database. Preprocessing, including unquantified data removal, missing value imputation, samples normalization, log transformation, and feature scaling, was performed. In total, 42 survival-related miRNAs were identified by Cox Proportional-Hazards analysis. The patients were optimally clustered into four groups with three different 5-years survival outcome (≥ 90%, ≈ 65%, ≤ 40%) by K-means clustering algorithm base on top 10 survival-related miRNAs. According to the K-means clustering result, a prediction model with high performance was established. The pathways analysis indicated that the miRNAs used play roles involved in the regulation of cancer stem cells.

CONCLUSION

A miRNAs-based machine learning cervical cancer survival prediction model was developed that robustly stratifies cervical cancer patients into high survival rate (5-years survival rate ≥ 90%), moderate survival rate (5-years survival rate ≈ 65%), and low survival rate (5-years survival rate ≤ 40%).

Comparison of different predicting models to assist the diagnosis of spinal lesions

In neurosurgical or orthopedic clinics, the differential diagnosis of lower back pain is often time-consuming and costly. This is especially true when there are several candidate diagnoses with similar symptoms that might confuse clinic physicians. Therefore, methods for the efficient differential diagnosis can help physicians to implement the most appropriate treatment and achieve the goal of pain reduction for their patients.In this study, we applied data-mining techniques from artificial intelligence technologies, in order to implement a computer-aided auxiliary differential diagnosis for a herniated intervertebral disc, spondylolithesis, and spinal stenosis. We collected questionnaires from 361 patients and analyzed the resulting data by using a linear discriminant analysis, clustering, and artificial neural network techniques to construct a related classification model and to compare the accuracy and implementation efficiency of the different methods.Our results indicate that a linear discriminant analysis has obvious advantages for classification and diagnosis, in terms of accuracy.We concluded that the judgment results from artificial intelligence can be used as a reference for medical personnel in their clinical diagnoses. Our method is expected to facilitate the early detection of symptoms and early treatment, so as to reduce the social resource costs and the huge burden of medical expenses, and to increase the quality of medical care.

Application of Artificial Intelligence in the Establishment of an Association Model between Metabolic Syndrome, TCM Constitution, and the Guidance of Medicated Diet Care

Background

This study conducted exploratory research using artificial intelligence methods. The main purpose of this study is to establish an association model between metabolic syndrome and the TCM (traditional Chinese medicine) constitution using the characteristics of individual physical examination data and to provide guidance for medicated diet care.

Methods

Basic demographic and laboratory data were collected from a regional hospital health examination database in northern Taiwan, and artificial intelligence algorithms, such as logistic regression, Bayesian network, and decision tree, were used to analyze and construct the association model between metabolic syndrome and the TCM constitution. Findings. It was found that the phlegm-dampness constitution (90.6%) accounts for the majority of TCM constitution classifications with a high risk of metabolic syndrome, and high cholesterol, blood glucose, and waist circumference were statistically significantly correlated with the phlegm-dampness constitution. This study also found that the age of patients with metabolic syndrome has been advanced, and shift work is one of the risk indicators. Therefore, based on the association model between metabolic syndrome and TCM constitution, in the future, metabolic syndrome can be predicted through the syndrome differentiation of the TCM constitution, and relevant medicated diet care schemes can be recommended for improvement.

Conclusion

In order to increase the public's knowledge and methods for mitigating metabolic syndrome, in the future, nursing staff can provide nonprescription medicated diet-related nursing guidance information via the prediction and assessment of the TCM constitution.

Using Accelerometer Data to Tune the Parameters of an Extended Kalman Filter for Optical Motion Capture: Preliminary Application to Gait Analysis

Optical motion capture is currently the most popular method for acquiring motion data in biomechanical applications. However, it presents a number of problems that make the process difficult and inefficient, such as marker occlusions and unwanted reflections. In addition, the obtained trajectories must be numerically differentiated twice in time in order to get the accelerations. Since the trajectories are normally noisy, they need to be filtered first, and the selection of the optimal amount of filtering is not trivial. In this work, an extended Kalman filter (EKF) that manages marker occlusions and undesired reflections in a robust way is presented. A preliminary test with inertial measurement units (IMUs) is carried out to determine their local reference frames. Then, the gait analysis of a healthy subject is performed using optical markers and IMUs simultaneously. The filtering parameters used in the optical motion capture process are tuned in order to achieve good correlation between the obtained accelerations and those measured by the IMUs. The results show that the EKF provides a robust and efficient method for optical system-based motion analysis, and that the availability of accelerations measured by inertial sensors can be very helpful for the adjustment of the filters.

Chronic Pain Diagnosis Using Machine Learning, Questionnaires, and QST: A Sensitivity Experiment

In the last decade, machine learning has been widely used in different fields, especially because of its capacity to work with complex data. With the support of machine learning techniques, different studies have been using data-driven approaches to better understand some syndromes like mild cognitive impairment, Alzheimer’s disease, schizophrenia, and chronic pain. Chronic pain is a complex disease that can recurrently be misdiagnosed due to its comorbidities with other syndromes with which it shares symptoms. Within that context, several studies have been suggesting different machine learning algorithms to classify or predict chronic pain conditions. Those algorithms were fed with a diversity of data types, from self-report data based on questionnaires to the most advanced brain imaging techniques. In this study, we assessed the sensitivity of different algorithms and datasets classifying chronic pain syndromes. Together with this assessment, we highlighted important methodological steps that should be taken into account when an experiment using machine learning is conducted. The best results were obtained by ensemble-based algorithms and the dataset containing the greatest diversity of information, resulting in area under the receiver operating curve (AUC) values of around 0.85. In addition, the performance of the algorithms is strongly related to the hyper-parameters. Thus, a good strategy for hyper-parameter optimization should be used to extract the most from the algorithm. These findings support the notion that machine learning can be a powerful tool to better understand chronic pain conditions.