IMU Sensor Fusion Algorithm for Monitoring Knee Kinematics in ACL Reconstructed Patients

Machine Learning-Based Gesture Recognition Glove: Design and Implementation

In the evolving field of human-computer interaction (HCI), gesture recognition has emerged as a critical focus, with smart gloves equipped with sensors playing one of the most important roles. Despite the significance of dynamic gesture recognition, most research on data gloves has concentrated on static gestures, with only a small percentage addressing dynamic gestures or both. This study explores the development of a low-cost smart glove prototype designed to capture and classify dynamic hand gestures for game control and presents a prototype of data gloves equipped with five flex sensors, five force sensors, and one inertial measurement unit (IMU) sensor. To classify dynamic gestures, we developed a neural network-based classifier, utilizing a convolutional neural network (CNN) with three two-dimensional convolutional layers and rectified linear unit (ReLU) activation where its accuracy was 90%. The developed glove effectively captures dynamic gestures for game control, achieving high classification accuracy, precision, and recall, as evidenced by the confusion matrix and training metrics. Despite limitations in the number of gestures and participants, the solution offers a cost-effective and accurate approach to gesture recognition, with potential applications in VR/AR environments.

Novel Tools to Approach and Measure Outcomes in Patients with Fractures

Upper extremity fractures are prevalent and pose a great burden to patients and society. In the US alone, the annual incidence of upper extremity fractures is 67.6 fractures per 10,000 persons. While the majority of patients with upper extremity fractures demonstrate satisfactory outcomes when treated appropriately (the details of which are discussed in prior articles), the importance of follow-up and outcome measurement cannot be understated. Outcome measurement allows for accountability and improvement in clinical outcomes and research. The purpose of this article is to describe recent advances in methods and tools for assessing clinical and research outcomes in hand and upper extremity care. Three specific advances that are broadly changing the landscape of follow-up care of our patients include: 1) telemedicine, 2) patient-reported outcome measurement, and 3) wearables/remote patient monitoring.

Evaluation of a Smart Knee Brace for Range of Motion and Velocity Monitoring during Rehabilitation Exercises and an Exergame

Anterior cruciate ligament (ACL) injuries often require a lengthy duration of rehabilitation for patients to return to their prior level of function. Adherence to rehabilitation during this prolonged period can be subpar due to the treatment duration and poor adherence to home exercises. This work evaluates whether a smart instrumented knee brace system is capable of monitoring knee range of motion and velocity during a series of common knee rehabilitation exercises and an exergame. A total of 15 healthy participants completed a series of common knee rehabilitation exercises and played an exergame while wearing a smart instrumented knee brace. The range of motion (ROM) and velocity of the knee recorded by the knee brace was compared to a reference optoelectronic system. The results show good agreement between the knee brace system and the reference system for all exercises performed. Participants were able to quickly learn how to play the exergame and scored well within the game. The system investigated in this study has the potential to allow rehabilitation to occur outside of the clinic with the use of remote monitoring, and improve adherence and outcomes through the use of an exergame.

Remote Patient Monitoring with Wearable Sensors Following Knee Arthroplasty

(Background) Inertial Measurement Units (IMUs) provide a low-cost, portable solution to obtain functional measures similar to those captured with three-dimensional gait analysis, including spatiotemporal gait characteristics. The primary aim of this study was to determine the feasibility of a remote patient monitoring (RPM) workflow using ankle-worn IMUs measuring impact load, limb impact load asymmetry and knee range of motion in combination with patient-reported outcome measures. (Methods) A pilot cohort of 14 patients undergoing primary knee arthroplasty for osteoarthritis was prospectively enrolled. RPM in the community was performed weekly from 2 up to 6 weeks post-operatively using wearable IMUs. The following data were collected using IMUs: mobility (Bone Stimulus and cumulative impact load), impact load asymmetry and maximum knee flexion angle. In addition, scores from the Oxford Knee Score (OKS), EuroQol Five-dimension (EQ-5D) with EuroQol visual analogue scale (EQ-VAS) and 6 Minute Walk Test were collected. (Results) On average, the Bone Stimulus and cumulative impact load improved 52% (p = 0.002) and 371% (p = 0.035), compared to Post-Op Week 2. The impact load asymmetry value trended (p = 0.372) towards equal impact loading between the operative and non-operative limb. The mean maximum flexion angle achieved was 99.25° at Post-Operative Week 6, but this was not significantly different from pre-operative measurements (p = 0.1563). There were significant improvements in the mean EQ-5D (0.20; p = 0.047) and OKS (10.86; p < 0.001) scores both by 6 weeks after surgery, compared to pre-operative scores. (Conclusions) This pilot study demonstrates the feasibility of a reliable and low-maintenance workflow system to remotely monitor post-operative progress in knee arthroplasty patients. Preliminary data indicate IMU outputs relating to mobility, impact load asymmetry and range of motion can be obtained using commercially available IMU sensors. Further studies are required to directly correlate the IMU sensor outputs with patient outcomes to establish clinical significance.

Monitoring work-related physical activity and estimating lower-limb loading: a proof-of-concept study

BACKGROUND

Physical activity (PA) is important to general health and knee osteoarthritis (OA). Excessive workplace PA is an established risk factor for knee OA however, appropriate methods of measurement are unclear. There is a need to examine and assess the utility of new methods of measuring workplace PA and estimating knee load prior to application to large-scale, knee OA cohorts. Our aims, therefore, were to monitor workplace PA and estimate lower-limb loading across different occupations in health participants.

METHODS

Twenty-four healthy adults, currently working full-time in a single occupation (≥ 35 h/week) and free of musculoskeletal disease, comorbidity and had no history of lower-limb injury/surgery (past 12-months) were recruited across New South Wales (Australia). A convenience sample was recruited with occupations assigned to levels of workload; sedentary, light manual and heavy manual. Metrics of workplace PA including tasks performed (i.e., sitting), step-count and lower-limb loading were monitored over 10 working days using a daily survey, smartwatch, and a smartphone.

RESULTS

Participants of light manual occupations had the greatest between-person variations in mean lower-limb load (from 2 to 59 kg*m/s3). Lower-limb load for most participants of the light manual group was similar to a single participant in heavy manual work (30 kg*m/s3) and was at least three times greater than the sedentary group (2 kg*m/s3). The trends of workplace PA over working hours were largely consistent, per individual, but rare events of extreme loads were observed across all participants (up to 760 kg*m/s3).

CONCLUSIONS

There are large interpersonal variations in metrics of workplace PA, particularly among light and heavy manual occupations. Our estimates of lower-limb loading were largely consistent with pre-conceived levels of physical demand. We present a new approach to monitoring PA and estimating lower-limb loading, which could be applied to future occupational studies of knee OA.

Integrating a Potentiometer into a Knee Brace Shows High Potential for Continuous Knee Motion Monitoring

Continuous monitoring of knee motion can provide deep insights into patients' rehabilitation status after knee injury and help to better identify their individual therapeutic needs. Potentiometers have been identified as one possible sensor type for continuous monitoring of knee motion. However, to verify their use in monitoring real-life environments, further research is needed. We aimed to validate a potentiometer-embedded knee brace to measure sagittal knee kinematics during various daily activities, as well as to assess its potential to continuously monitor knee motion. To this end, the sagittal knee motion of 32 healthy subjects was recorded simultaneously by an instrumented knee brace and an optoelectronic reference system during activities of daily living to assess the agreement between these two measurement systems. To evaluate the potentiometer's behavior during continuous monitoring, knee motion was continuously recorded in a subgroup (n = 9) who wore the knee brace over the course of a day. Our results show a strong agreement between the instrumented knee brace and reference system across all investigated activities as well as stable sensor behavior during continuous tracking. The presented potentiometer-based sensor system demonstrates strong potential as a device for measuring sagittal knee motion during daily activities as well as for continuous knee motion monitoring.

Wearable Hand Module and Real-Time Tracking Algorithms for Measuring Finger Joint Angles of Different Hand Sizes with High Accuracy using FBG Strain Sensor

This paper presents a wearable hand module which was made of five fiber Bragg grating (FBG) strain sensor and algorithms to achieve high accuracy even when worn on different hand sizes of users. For real-time calculation with high accuracy, FBG strain sensors move continuously according to the size of the hand and the bending of the joint. Representatively, four algorithms were proposed; point strain (PTS), area summation (AREA), proportional summation (PS), and PS/interference (PS/I or PS/I_α). For more accurate and efficient assessments, 3D printed hand replica with different finger sizes was adopted and quantitative evaluations were performed for index~little fingers (77 to 117 mm) and thumb (68~78 mm). For index~little fingers, the optimized algorithms were PS and PS/I_α. For thumb, the optimized algorithms were PS/I_α and AREA. The average error angle of the wearable hand module was observed to be 0.47 ± 2.51° and mean absolute error (MAE) was achieved at 1.63 ± 1.97°. These results showed that more accurate hand modules than other glove modules applied to different hand sizes can be manufactured using FBG strain sensors which move continuously and algorithms for tracking this movable FBG sensors.