Deep Learning in Gait Parameter Prediction for OA and TKA Patients Wearing IMU Sensors

Renani, Mohsen Sharifi; Myers, Casey A.; Zandie, Rohola; Mahoor, Mohammad H.; Davidson, Bradley S.; Clary, Chadd W.

doi:10.3390/s20195553

Cited by 35 publications

(27 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With our search strategy, we did not find any paper on machine-learning algorithms for extracting gait features. However, we found a recent study [116] outside our search criteria, which implemented a deep learning approach to estimate the spatiotemporal gait features. The dataset for deep learning and the ground truth were collected from seven IMUs and motion capture combined with force plate systems, respectively, with multiple 5 m walk protocols at various speeds.…”

Section: Discussionmentioning

confidence: 99%

Recent Advances in Quantitative Gait Analysis Using Wearable Sensors: A Review

2021

View full text Add to dashboard Cite

The current gold standard for gait analysis involves performing the gait experiments in a laboratory environment with a constrained space. However, there is growing interest in using flexible, efficient, and inexpensive wearable sensors as tools to perform gait analysis. This review aimed to identify and summarize the current advances in wearable sensors for various aspects of gait analysis, such as the application of wearable gait analysis systems, sensor systems and their attachment locations, and the algorithms used for the analysis. The PRISMA guideline was adopted to find relevant studies from the period 2011 to 2020 from several scientific databases. A total of 76 articles were selected based on the inclusion and exclusion criteria. A wearable inertial measurement unit (IMU) attached to the lower limb region was found to be the most common approach for gait analysis. Temporal, spatial, and spatiotemporal features were the most common quantitative gait features extracted from the wearable sensors. The proposed frameworks showed varying performances, and an increased number of sensors did not necessarily improve the estimation performance metrics. A few studies have integrated various machine learning techniques for classification problems, correction algorithms, crosschecking functions, and scoring functions. Finally, this review paper discusses the challenges and future direction of the research on quantitative gait analysis.

show abstract

Section: Discussionmentioning

confidence: 99%

Recent Advances in Quantitative Gait Analysis Using Wearable Sensors: A Review

2021

View full text Add to dashboard Cite

show abstract

“…Selection of the appropriate neural network architecture is an important step in attaining the requisite accuracy for model predictions. In previous work, we systematically evaluated multiple neural network configurations for predicting spatiotemporal gait characteristics on this same dataset and found that convolutional neural networks yielded the highest accuracy predictions [ 28 ]. Mundt et al also compared LSTM and feedforward neural networks (FFNN) performance on time-normalized gait cycle input data and achieved better performance using the FFNN.…”

Section: Discussionmentioning

confidence: 99%

“…All the subjects who participated in this study had either end-stage osteoarthritis in the hip or knee or had recently recovered from a total joint arthroplasty. This patient population has been shown to exhibit gait adaptations, including a slower pace, shorter step length, reduced knee flexion, and increased levels of variability that may affect the generalization of the model to healthy individuals [ 28 , 48 , 49 , 50 ]. Gait measurements were taken in the laboratory environment, which may affect the subjects’ normal gait patterns.…”

Section: Discussionmentioning

confidence: 99%

“…Subjects were outfitted with 71 reflective markers on anatomical landmarks and 17 research-grade IMUs on various limb segments and the trunk. Only IMUs located on the pelvis, left thigh, left shank, and left foot were used in this analysis [ 28 ]. Thigh and shank IMUs were attached to rigid 4-marker clusters used to track the relative orientations of the IMUs, while markers placed directly on the IMUs were used to track IMU displacements.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

The Use of Synthetic IMU Signals in the Training of Deep Learning Models Significantly Improves the Accuracy of Joint Kinematic Predictions

Renani

Eustace

Myers

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

Gait analysis based on inertial sensors has become an effective method of quantifying movement mechanics, such as joint kinematics and kinetics. Machine learning techniques are used to reliably predict joint mechanics directly from streams of IMU signals for various activities. These data-driven models require comprehensive and representative training datasets to be generalizable across the movement variability seen in the population at large. Bottlenecks in model development frequently occur due to the lack of sufficient training data and the significant time and resources necessary to acquire these datasets. Reliable methods to generate synthetic biomechanical training data could streamline model development and potentially improve model performance. In this study, we developed a methodology to generate synthetic kinematics and the associated predicted IMU signals using open source musculoskeletal modeling software. These synthetic data were used to train neural networks to predict three degree-of-freedom joint rotations at the hip and knee during gait either in lieu of or along with previously measured experimental gait data. The accuracy of the models’ kinematic predictions was assessed using experimentally measured IMU signals and gait kinematics. Models trained using the synthetic data out-performed models using only the experimental data in five of the six rotational degrees of freedom at the hip and knee. On average, root mean square errors in joint angle predictions were improved by 38% at the hip (synthetic data RMSE: 2.3°, measured data RMSE: 4.5°) and 11% at the knee (synthetic data RMSE: 2.9°, measured data RMSE: 3.3°), when models trained solely on synthetic data were compared to measured data. When models were trained on both measured and synthetic data, root mean square errors were reduced by 54% at the hip (measured + synthetic data RMSE: 1.9°) and 45% at the knee (measured + synthetic data RMSE: 1.7°), compared to measured data alone. These findings enable future model development for different activities of clinical significance without the burden of generating large quantities of gait lab data for model training, streamlining model development, and ultimately improving model performance.

show abstract

“…[13][14][15][16][17] Recently, an increasing number of studies reported that AI can be used in the TKA procedure. [18][19][20] For example, Karnuta et al 18 used AI to identify arthroplasty implants from radiographs of the knee. Lind et al 21 also utilised AI to identify and classify fractures of the knee.…”

mentioning

confidence: 99%

Deep learning approach for guiding three‐dimensional computed tomography reconstruction of lower limbs for robotically‐assisted total knee arthroplasty

Zhang

Ding

et al. 2021

Robotics Computer Surgery

View full text Add to dashboard Cite

Background: Robotic-assisted total knee arthroplasty (TKA) was performed to promote the accuracy of bone resection and mechanical alignment. Among these TKA system procedures, 3D reconstruction of CT data of lower limbs consumes significant manpower. Artificial intelligence (AI) algorithms applying deep learning has been proved efficient in automated identification and visual processing.Methods: CT data of a total of 200 lower limbs scanning were used for AI-based 3D model construction and CT data of 20 lower limbs scanning were utilised for verification. Results:We showed that the performance of an AI-guided 3D reconstruction of CT data of lower limbs for robotic-assisted TKA was similar to that of the operatorbased approach. The time of 3D lower limb model construction using AI was 4.7 min. AI-based 3D models can be used for surgical planning. Conclusion:AI was used for the first time to guide the 3D reconstruction of CT data of lower limbs for facilitating robotic-assisted TKA. Incorporation of AI in 3D model reconstruction before TKA might reduce the workload of radiologists. K E Y W O R D Sartificial intelligence, HURWA robotic-assisted TKA system, robotically TKA assisted system, 3D CT model | INTRODUCTIONTotal knee arthroplasty (TKA) is an effective procedure for cases with end-stage rheumatoid arthritis or osteoarthritis of the knee. 1 The surgery was carried out in over 370,000 and 700,000 patients annually in China and USA, respectively. 2 Despite the development in implant materials and design, thromboembolic prophylaxis, rehabilitation, computer navigation and patient-specific implants, and the use of prophylactic antibiotics, over 20% of patients reported unsatisfaction following TKA. [3][4][5][6] Robotically assisted TKA was thus developed to promote mechanical alignment, accuracy of bone resection, soft tissue protection, and implant survivorship and implant stability. [7][8][9][10] In this connection, our previous studies showed that a novel HURWA robotic-assisted TKA system (BEIJING HURWA-ROBOT Technology Co. Ltd) could enhance the accuracy of bone resection in terms of resection angles and levels in the sawbone and the sheep models. 11,12 The procedures of robotic-assisted TKA system consists of the following steps: preoperative computed tomography (CT) scanning of lower limbs, three-dimensional (3D) reconstruction of CT data of lower limbs, CT-based planning, surgical planning depending on the 3D CT model, and the act of performing robotic-assisted TKA. Among

show abstract

Deep Learning in Gait Parameter Prediction for OA and TKA Patients Wearing IMU Sensors

Cited by 35 publications

References 45 publications

Recent Advances in Quantitative Gait Analysis Using Wearable Sensors: A Review

Recent Advances in Quantitative Gait Analysis Using Wearable Sensors: A Review

The Use of Synthetic IMU Signals in the Training of Deep Learning Models Significantly Improves the Accuracy of Joint Kinematic Predictions

Deep learning approach for guiding three‐dimensional computed tomography reconstruction of lower limbs for robotically‐assisted total knee arthroplasty

Contact Info

Product

Resources

About