Joint angle estimation with wavelet neural networks

A scoping review of applications of artificial intelligence in kinematics and kinetics of ankle sprains - current state-of-the-art and future prospects

BACKGROUND

Despite the existence of evidence-based rehabilitation strategies that address biomechanical deficits, the persistence of recurrent ankle problems in 70% of patients with acute ankle sprains highlights the unresolved nature of this issue. Artificial intelligence (AI) emerges as a promising tool to identify definitive predictors for ankle sprains. This paper aims to summarize the use of AI in investigating the ankle biomechanics of healthy and subjects with ankle sprains.

METHODS

Articles published between 2010 and 2023 were searched from five electronic databases. 59 papers were included for analysis with regards to: i). types of motion tested (functional vs. purposeful ankle movement); ii) types of biomechanical parameters measured (kinetic vs kinematic); iii) types of sensor systems used (lab-based vs field-based); and, iv) AI techniques used.

FINDINGS

Most studies (83.1%) examined biomechanics during functional motion. Single kinematic parameter, specifically ankle range of motion, could obtain accuracy up to 100% in identifying injury status. Wearable sensor exhibited high reliability for use in both laboratory and on-field/clinical settings. AI algorithms primarily utilized electromyography and joint angle information as input data. Support vector machine was the most used supervised learning algorithm (18.64%), while artificial neural network demonstrated the highest accuracy in eight studies.

INTERPRETATIONS

The potential for remote patient monitoring is evident with the adoption of field-based devices. Nevertheless, AI-based sensors are underutilized in detecting ankle motions at risk of sprain. We identify three key challenges: sensor designs, the controllability of AI models, and the integration of AI-sensor models, providing valuable insights for future research.

Lower-limb sagittal joint angles during gait can be predicted based on foot acceleration and angular velocity

Background and purpose

Continuous monitoring of lower-limb movement may help in the early detection and control/reduction of diseases (such as the progression of orthopedic diseases) by applying suitable interventions. Therefore, it is invaluable to calculate the lower-limb movement (sagittal joint angles) while walking daily for continuous evaluation of such risks. Although cameras in a motion capture system are necessary for calculating lower-limb sagittal joint angles during gait, the method is unrealistic considering the setting is difficult to achieve in daily life. Therefore, the estimation of lower-limb sagittal joint angles during walking based on variables, which can be measured using wearable sensors (e.g., foot acceleration and angular velocity), is important. This study estimates the lower-limb sagittal joint angles during gait from the norms of foot acceleration and angular velocity using machine learning and validates the accuracy of the estimated joint angles with those obtained using a motion capture system.

Methods

Healthy adults (n = 200) were asked to walk at a comfortable speed (10 trials), and their lower-limb sagittal joint angles, foot accelerations, and angular velocities were obtained. Using these variables, we established a feedforward neural network and estimated the lower-limb sagittal joint angles.

Results

The average root mean squared errors of the lower-limb sagittal joint angles during gait ranged between 2.5°-7.0° (hip: 7.0°; knee: 4.0°; and ankle: 2.5°).

Conclusion

These results show that we can estimate the lower-limb sagittal joint angles during gait using only the norms of foot acceleration and angular velocity, which can help calculate the lower-limb sagittal joint angles during daily walking.

Lower body kinematics estimation during walking using an accelerometer

Measuring and predicting accurate joint angles are important to developing analytical tools to gauge users' progress. Such measurement is usually performed in laboratory settings, which is difficult and expensive. So, the aim of this study was continuous estimation of lower limb joint angles during walking using an accelerometer and random forest (RF). Thus, 73 subjects (26 women and 47 men) voluntarily participated in this study. The subjects walked at the slow, moderate, and fast speeds on a walkway, which was covered with 10 Vicon camera. Acceleration was used as input for a RF to estimate ankle, knee, and hip angles (in transverse, frontal, and sagittal planes). Pearson correlation coefficient (r) and Mean Square Error (MSE) were computed between the experimental and estimated data. Paired statistical parametric mapping (SPM) t-test was used to compare the experimental and estimated data throughout gait cycle. The results of this study showed that the MSE of joint angles between the experimental and estimated data ranged from 0.04 to 24.29 and r > 0.91. Moreover, the findings of SPM indicated that there was no significant difference between the experimental and estimated data of ankle, knee, and hip angles in all three planes throughout gait cycle. The results of our research developed a more accessible, portable procedure to quantifying lower limb joint angles by an accelerometer and RF. So, such wearable-based joint angles have the potential to be used in outside-laboratory settings to measure walking kinematics.

Closing the Wearable Gap: Foot-ankle kinematic modeling via deep learning models based on a smart sock wearable

The development of wearable technology, which enables motion tracking analysis for human movement outside the laboratory, can improve awareness of personal health and performance. This study used a wearable smart sock prototype to track foot-ankle kinematics during gait movement. Multivariable linear regression and two deep learning models, including long short-term memory (LSTM) and convolutional neural networks, were trained to estimate the joint angles in sagittal and frontal planes measured by an optical motion capture system. Participant-specific models were established for ten healthy subjects walking on a treadmill. The prototype was tested at various walking speeds to assess its ability to track movements for multiple speeds and generalize models for estimating joint angles in sagittal and frontal planes. LSTM outperformed other models with lower mean absolute error (MAE), lower root mean squared error, and higher R-squared values. The average MAE score was less than 1.138° and 0.939° in sagittal and frontal planes, respectively, when training models for each speed and 2.15° and 1.14° when trained and evaluated for all speeds. These results indicate wearable smart socks to generalize foot-ankle kinematics over various walking speeds with relatively low error and could consequently be used to measure gait parameters without the need for a lab-constricted motion capture system.

Estimation of lower-limb sagittal joint moments during gait using vertical ground reaction force

Lower-limb sagittal joint moments during gait are important variables for evaluating the risk of disease progression, such as that of orthopedic diseases. Therefore, quantifying lower-limb sagittal joint moments during walking is important to continuously evaluate the risk of disease progression. A motion capture system and force plate are employed in the calculation of lower-limb sagittal joint moments during gait. However, they cannot be used during daily walking. Therefore, it is important to estimate these moments during walking from the vertical ground reaction force (vGRF), which can be measured using a wearable sensor, such as an insole device. Thus, this study aimed to estimate the lower-limb sagittal joint moments during gait using only the vGRF and confirmed its accuracy. This study included 188 healthy adults, and each participant walked at a comfortable speed (10 trials). We estimated the moments from the vGRF using a feedforward neural network. Our major findings are that our method can estimate lower-limb sagittal joint moments using the vGRF with accuracies of NRMSE¯ within 6.0-11.7% (NRMSEs¯ of the hip, knee, and ankle were 8.4, 11.7, and 6.0%, respectively). To the best of our knowledge, this study is the first to estimate lower-limb sagittal joint moments (including those of the hip, knee, and ankle joints) during gait using only the vGRF. Our method may be useful to estimate lower-limb sagittal joint moments during daily walking using only the vGRF, which can be measured by an insole device in the future.

The Effect of Law Students in Entrepreneurial Psychology Under the Artificial Intelligence Technology

With the increasingly serious employment situation in China, the government and schools encourage college students to start businesses to alleviate employment pressure. College student's successful entrepreneurship depends on national preferential policies, social support, and, most importantly, their healthy and solid psychological quality and entrepreneurial psychological quality. The purpose is to understand the entrepreneurial psychology of college students and study the entrepreneurial psychological effect. Firstly, the four aspects of entrepreneurial psychology are summarized, including entrepreneurial awareness, entrepreneurial volition, entrepreneurial ability, and entrepreneurial personality. Secondly, the research status of college students' entrepreneurial psychology is reviewed, and the existing problems are pointed out. Thirdly, the combined model of wavelet transform and Neural Network (NN) is proposed, and the feasibility of the proposed model is evaluated through the analysis of college students' entrepreneurial psychology. The wavelet NN is used in experimental design to predict college students' entrepreneurial psychology, and the predicted results are compared with the actual value. From the perspective of the prediction results of entrepreneurial psychology, the combination of wavelet algorithm and neural network is more accurate for entrepreneurial psychology prediction and evaluation results of law students. Overall, the difference between the predicted value and the actual value is within 0.3 points, which is relatively stable. According to the analysis of single-factor results, the scores of students of different majors in the four dimensions of entrepreneurial psychology are all higher than 3.5, but there is no significant difference among the four dimensions (P > 0.05), indicating that the major has no significant impact on entrepreneurial psychology; law students with different educational backgrounds have significant differences in entrepreneurial psychology (P < 0.05), among which students with a master's degree have the strongest entrepreneurial will, while doctoral students have the lowest entrepreneurial will; in terms of entrepreneurial psychological capital, men's self-efficacy is higher than women's, and the difference is significant (P < 0.05). The difference between males and females in the scores of entrepreneurial psychological factors' four aspects is not very obvious. In terms of entrepreneurial psychological capital, males' self-efficacy is significantly higher than females' (P < 0.05). Artificial Intelligence (AI) technology has great application prospects in the prediction and evaluation of college students' entrepreneurial psychology, and college students' entrepreneurial psychology is highly correlated with gender and education.