Predicting TUG Score from Gait Characteristics with Video Analysis and Machine Learning

Ma, Jian

doi:10.1007/978-981-99-6187-0_1

Cited by 3 publications

(1 citation statement)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the realm of fall risk prediction using computer vision techniques, there has been comparatively less research, but notable strides have been made. In [24], a computer visionbased method was employed to predict the Timed-Up-and-Go (TUG) score, an indicator of fall risk. This was achieved by utilizing regression models, including linear regression and SVM regressors, based on 3D poses derived from video recordings captured by 2D/3D cameras.…”

Section: B Fall Predictionmentioning

confidence: 99%

Computer Vision Based Transfer Learning-Aided Transformer Model for Fall Detection and Prediction

McCall,

Kolawole,

Naz

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Falls bring about significant risks to individuals' well-being and independence, prompting widespread public health concerns. Swift detection and even predicting the risk of falls are crucial for implementing effective measures to alleviate the adverse consequences associated with such incidents. This study presents a new framework for identifying and forecasting fall risks. Our approach utilizes a novel transformer model trained on 2D poses extracted through an off-the-shelf pose extractor, incorporating transfer learning techniques. Initially, the transformer is trained on a large dataset containing 2D poses of general actions. Subsequently, we freeze the majority of its layers and fine-tune only the last few layers using relatively smaller datasets for fall detection and prediction tasks. Experimental results indicate that our proposed method outperforms traditional machine learning (e.g., SVM, Decision Tree, etc.) and deep learning approaches (e.g., LSTM, CNN, ST-GCN, PoseC3D, etc.) in both fall detection and prediction tasks across various datasets.

show abstract

Section: B Fall Predictionmentioning

confidence: 99%

Computer Vision Based Transfer Learning-Aided Transformer Model for Fall Detection and Prediction

McCall,

Kolawole,

Naz

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

Automation of observational gait assessment through an optical 3D motion system and transformers

Carneros-Prado,

González-Velázquez,

Dobrescu

et al. 2025

Appl Intell

View full text Add to dashboard Cite

Establishing the Foundations of Emotional Intelligence in Care Companion Robots to Mitigate Agitation Among High-Risk Patients With Dementia: Protocol for an Empathetic Patient-Robot Interaction Study

Nyamathi,

Dutt,

Lee

et al. 2024

JMIR Res Protoc

View full text Add to dashboard Cite

Background: There are an estimated 6.7 million persons living with dementia in the U.S.; expected to double by 2060. Those at highest risk, persons experiencing moderate to severe dementia (P-MSD), are 4-5 times more likely to fall than those without dementia, as they often experience unpredictable agitation, leading to unsteady gait. Socially assistive robots fail to address the dynamically changing emotional states associated with agitation, and there is a lack of understanding how emotional states change, how they impact agitation and gait over time, and how social robots can best respond by showing empathy.Objective: Design and validate a foundational model of emotional intelligence for empathic patient-robot interaction that mitigates agitation among those at highest risk, P-MSD.Methods: A design science approach will be used to: 1) collect and store granular, personal, chronological data (Personicle), using real-time visual, audio and physiological sensing technologies in a simulation lab and Board & Care facilities; 2) develop statistical models to understand and forecast a P-MSD's emotional state, agitation level and gait in real-time using ML/AI and the Personicle; 3) design and test an empathy-focused conversation model, focused on storytelling; and 4) test and evaluate the empathy-focused conversation model for the Care Companion Robot (CCR) in the community.

show abstract

Predicting TUG Score from Gait Characteristics with Video Analysis and Machine Learning

Cited by 3 publications

References 29 publications

Computer Vision Based Transfer Learning-Aided Transformer Model for Fall Detection and Prediction

Computer Vision Based Transfer Learning-Aided Transformer Model for Fall Detection and Prediction

Automation of observational gait assessment through an optical 3D motion system and transformers

Establishing the Foundations of Emotional Intelligence in Care Companion Robots to Mitigate Agitation Among High-Risk Patients With Dementia: Protocol for an Empathetic Patient-Robot Interaction Study

Contact Info

Product

Resources

About