Gait Event Detection and Travel Distance Using Waist-Worn Accelerometers across a Range of Speeds: Automated Approach

Sensors, Journal Year: 2024, Volume and Issue: 24(4), P. 1155 - 1155

Published: Feb. 9, 2024

Estimation of temporospatial clinical features gait (CFs), such as step count and length, duration, frequency, speed, distance traveled, is an important component community-based mobility evaluation using wearable accelerometers. However, accurate unsupervised computerized measurement CFs individuals with Duchenne muscular dystrophy (DMD) who have progressive loss ambulatory difficult due to differences in patterns magnitudes acceleration across their range attainable velocities. This paper proposes a novel calibration method. It aims detect steps, estimate stride lengths, determine travel distance. The approach involves combination observation, machine-learning-based detection, regression-based length prediction. method demonstrates high accuracy children DMD typically developing controls (TDs) regardless the participant’s level ability. Fifteen fifteen TDs underwent supervised testing speeds 10 m or 25 run/walk (10 MRW, MRW), 100 (100 6-min walk (6 MWT), free-walk (FW) evaluations while wearing mobile-phone-based accelerometer at waist near body’s center mass. Following by trained evaluator, were extracted from data multi-step process results compared ground-truth observation data. Model predictions vs. observed values for counts, showed strong correlation (Pearson’s r = −0.9929 0.9986, p < 0.0001). estimates demonstrated mean (SD) percentage error 1.49% (7.04%) 1.18% (9.91%) 0.37% (7.52%) observations combined 6 MWT, FW tasks. Our study findings indicate that single waist-worn calibrated individual’s characteristics our methods accurately measures distances common both TD peers.

Language: Английский

---

Wearable sensors in paediatric neurology

Developmental Medicine & Child Neurology, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 31, 2025

Wearable sensors have the potential to transform diagnosis, monitoring, and management of children who neurological conditions. Traditional methods for assessing disorders rely on clinical scales subjective measures. The snapshot disease progression at a particular time point, lack cooperation by during assessments, susceptibility bias limit utility these sensors, which capture data continuously in natural settings, offer non-invasive objective alternative traditional methods. This review examines role wearable various paediatric conditions, including cerebral palsy, epilepsy, autism spectrum disorder, attention-deficit/hyperactivity as well Rett syndrome, Down Angelman Prader-Willi neuromuscular such Duchenne muscular dystrophy spinal atrophy, ataxia, Gaucher disease, headaches, sleep disorders. highlights their application tracking motor function, seizure activity, daily movement patterns gain insights into therapeutic response. Although challenges related population size, compliance, ethics, regulatory approval remain, technology promises improve trials outcomes patients neurology.

Language: Английский

---

The necessity for skeletal muscle contractile assays to assess treatment efficacy in DMD

Rare Disease and Orphan Drugs Journal, Journal Year: 2025, Volume and Issue: 4(1)

Published: March 3, 2025

Body movement relies on skeletal muscles generating power to move limbs effectively. Power is defined as force multiplied by velocity: a muscle produces at specific velocity (the speed of shortening) and this results in power. In diseases like Duchenne Muscular Dystrophy (DMD), the absence dystrophin weakens impairs their shortening velocity, leading decreased consequently, impaired movement. Additionally, diaphragm heart are also affected DMD, causing difficulty breathing cardiac function. Compromised cardiorespiratory function can ultimately lead death. Given complex etiology DMD essential role all muscles, it crucial assess potential treatments for effectiveness improving This review focuses fundamental physiological assays used evaluate muscles. Common include force-frequency, force-velocity, power, eccentric protocols, which conducted ex vivo , situ small rodents (such mice rats) larger intermediate animal models such Golden Retriever dog. Existing data support use contractile objective tools assessing efficacy treatments.

Language: Английский

---

A Machine-Learning Method of Predicting Vital Capacity Plateau Value for Ventilatory Pump Failure Based on Data Mining

Healthcare, Journal Year: 2021, Volume and Issue: 9(10), P. 1306 - 1306

Published: Sept. 30, 2021

Ventilatory pump failure is a common cause of death for patients with neuromuscular diseases. The vital capacity plateau value (VCPLAT) an important indicator to judge the status ventilatory congenital myopathy, Duchenne muscular dystrophy and spinal atrophy. Due complex relationship between VCPLAT patient’s own condition, it difficult predict pediatric disease from medical perspective. We established prediction model based on data mining machine learning. first performed correlation analysis recursive feature elimination cross-validation (RFECV) provide high-quality combinations. Based this, Light Gradient Boosting Machine (LightGBM) algorithm was establish powerful performance. Finally, we verified validity superiority proposed method via comparison other models in similar works. After 10-fold cross-validation, had best performance its explained variance score (EVS), mean absolute error (MAE), squared (MSE), root square (RMSE), median (MedAE) R2 were 0.949, 0.028, 0.002, 0.045, 0.015 0.948, respectively. It also well test datasets. Therefore, can accurately effectively VCPLAT, thereby determining severity condition auxiliary decision-making doctors clinical diagnosis treatment.

Language: Английский

---

Population longitudinal analysis of Gait Profile Score and North Star Ambulatory Assessment in children with Duchenne muscular dystrophy

CPT Pharmacometrics & Systems Pharmacology, Journal Year: 2024, Volume and Issue: 13(5), P. 891 - 903

Published: March 27, 2024

Abstract Duchenne muscular dystrophy (DMD) is a rare X‐linked recessive disorder characterized by loss‐of‐function mutations in the gene encoding dystrophin. These lead to progressive functional deterioration including muscle weakness, respiratory insufficiency, and musculoskeletal deformities. Three‐dimensional gait analysis (3DGA) has been used as tool analyze pathology through quantification of altered joint kinematics, kinetics, activity patterns. Among 3DGA indices, Gait Profile Score (GPS), sensitive overall measure detect clinically relevant changes patterns children with DMD. To enhance our understanding clinical translation 3DGA, we report here development population nonlinear mixed‐effect model that jointly describes disease progression index, GPS, endpoint, North Star Ambulatory Assessment (NSAA). The final consists quadratic structure for GPS linear GPS‐NSAA correlation. Our was able capture improvement function NSAA younger subjects, well decline older subjects. Furthermore, predicted (CFB) at 1 year reasonably DMD subjects ≤7 years old baseline. tended slightly underpredict after those >7 baseline, but prediction summary statistics were maintained within standard deviation observed data. Quantitative models such this may help answer questions facilitate novel therapies

Language: Английский

---

Human Pose Estimation for Clinical Analysis of Gait Pathologies

Bioinformatics and Biology Insights, Journal Year: 2024, Volume and Issue: 18

Published: Jan. 1, 2024

Gait analysis serves as a critical diagnostic tool for identifying neurologic and musculoskeletal damage. Traditional manual of motion data, however, is labor-intensive heavily reliant on the expertise judgment therapist. This study introduces binary classification method quantitative assessment gait impairments, specifically focusing Duchenne muscular dystrophy (DMD), prevalent fatal neuromuscular genetic disorder. The research compares spatiotemporal sagittal kinematic features derived from 2D 3D human pose estimation trajectories against concurrently recorded capture (MoCap) data healthy children. proposed model leverages novel benchmark dataset, collected YouTube publicly available datasets their typically developed peers, to extract time-distance variables (e.g. speed, step length, stride time, cadence) joint angles lower extremity hip, knee, knee flexion angles). Machine learning deep techniques are employed discern patterns that can identify children exhibiting DMD disturbances. While current capable distinguishing between subjects those with DMD, it does not differentiate patients other impairments. Experimental results validate efficacy our cost-effective method, which relies RGB video, in detecting abnormalities, achieving prediction accuracy 96.2% Support Vector (SVM) 97% network.

Language: Английский

---

Digital outcome measures in Duchene muscular dystrophy: Lessons learnt from clinical trials

Journal of Neuromuscular Diseases, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 8, 2024

Duchenne muscular dystrophy is a severe neuromuscular disorder characterized by progressive muscle degeneration resulting from mutations in the dystrophin gene. Digital outcome measures offer promising alternative to traditional used clinical trials. This review explores development and application of digital dystrophy, emphasizing feasibility, reliability, sensitivity, validity these measures. The stride velocity 95th centile has been validated as robust endpoint approved for use evaluation drugs treatment European Medicines Agency. Although have potential enhance efficiency accuracy trials, challenges such limited sample sizes patient compliance persist. integration artificial intelligence into data analysis progress, but further validation required before strategies can be incorporated future trial methodologies.

Language: Английский

---

DL-Enhanced GAIT Analysis for Rehabilitation: A Comprehensive Survey

International Journal of Electrical and Electronics Engineering, Journal Year: 2024, Volume and Issue: 11(11), P. 203 - 221

Published: Nov. 30, 2024

Integrating DL techniques has revolutionized gait analysis, enhancing the accuracy and efficiency of detecting characterizing abnormalities. This paper surveys recent studies employing Deep Learning algorithms (DL), such as Convolutional Neural Networks (CNNs) Recurrent (RNNs), to analyze patterns from diverse data sources with wearable sensors, video footage, motion capture systems. The advantages in handling complex, high-dimensional its potential uncover subtle indicative disease or recovery status are discussed. Furthermore, clinical applications DL-based emphasizing role personalized rehabilitation programs real-time monitoring, explored. also addresses challenges implementing settings, need for large, annotated datasets, computational resources, interdisciplinary collaboration. In conclusion, this survey highlights transformative methods analysis fracture Parkinson's patients. By providing a detailed overview current research identifying key trends challenges, work seems inform inspire further advancements field, ultimately outcomes quality life affected individuals.

Language: Английский

---

Automated Detection of Gait Events and Travel Distance Using Waist-worn Accelerometers Across a Typical Range of Walking and Running Speeds

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Estimation of temporospatial clinical features gait (CFs), such as step count and length, duration, frequency, speed, distance traveled, is an important component community-based mobility evaluation using wearable accelerometers. However, accurate unsupervised computerized measurement CFs individuals with Duchenne muscular dystrophy (DMD) who have progressive loss ambulatory difficult due to differences in patterns magnitudes acceleration across their range attainable velocities. This paper proposes a novel calibration method. It aims detect steps, estimate stride lengths, determine travel distance. The approach involves combination observation, machine-learning-based detection, regression-based length prediction. method demonstrates high accuracy children DMD typically developing controls (TDs) regardless the participant's level ability. Fifteen fifteen TDs underwent supervised testing speeds 10 m or 25 run/walk (10 MRW, MRW), 100 (100 6-min walk (6 MWT), free-walk (FW) evaluations while wearing mobile-phone-based accelerometer at waist near body's center mass. Following by trained evaluator, were extracted from data multi-step process results compared ground-truth observation data. Model predictions vs. observed values for counts, showed strong correlation. Our study findings indicate that single waist-worn calibrated individual's characteristics our methods accurately measures estimates distances common both TD peers.

Language: Английский

---