Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Stepwise enhanced photoelectrochemical activity of BiOCl by Fe(III)-doping and aptamer immobilization for ultrasensitive detection of tetracycline.

Food chemistry·2026

Same author

Reverse-engineered ratiometric sensor assisted by dual redox-active DNA intercalators for ultrasensitive and reliable detection of mercury(II).

Analytica chimica acta·2026

Same author

Waste-to-waste synergistic regenerative combustion of ultra-low concentration methane and coal gangue in fluidization: Multi-parameter coupling effects.

Environmental research·2026

Same author

Hemin-Engineered UiO-66(Ce) as a Highly Active Nanozyme for Electrochemical Sensing of Hydrogen Peroxide and Hydrogen in Beverages.

Journal of agricultural and food chemistry·2026

Same author

Fluorine-Driven Polychromatic Colorimetric Screening of PFAS via Iodide-Mediated Gold Nanorod Etching on a Dual-Ligand Cu-MOF: A "No Aggregation, Low Inhibition" Sensing Paradigm.

Analytical chemistry·2026

Same author

Lithium-induced reconstruction of Li<sub>4</sub>Ti<sub>5</sub>O<sub>12</sub> for acidic H<sub>2</sub>O<sub>2</sub> production <i>via</i> two-electron water oxidation.

Chemical communications (Cambridge, England)·2026

Search research articles

Related Experiment Video

Updated: Nov 8, 2025

An Objective and Child-friendly Assessment of Arm Function by Using a 3-D Sensor

An Objective and Child-friendly Assessment of Arm Function by Using a 3-D Sensor

Published on: February 12, 2018

Simple action for depression detection: using kinect-recorded human kinematic skeletal data.

Wentao Li¹, Qingxiang Wang², Xin Liu³

¹School of Computer Science and Technology, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China.

|April 23, 2021

Summary

Machine learning models using Kinect skeletal data can help identify depression. Gradient Boosting achieved 76.92% accuracy, with better results in older individuals and females.

Keywords:

Depression detection Human skeleton joints Kinect sensor Machine learning

More Related Videos

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

Measuring the Kinematics of Daily Living Movements with Motion Capture Systems in Virtual Reality

Measuring the Kinematics of Daily Living Movements with Motion Capture Systems in Virtual Reality

Published on: April 5, 2018

Related Experiment Videos

Last Updated: Nov 8, 2025

An Objective and Child-friendly Assessment of Arm Function by Using a 3-D Sensor

An Objective and Child-friendly Assessment of Arm Function by Using a 3-D Sensor

Published on: February 12, 2018

Design and Analysis for Fall Detection System Simplification

Design and Analysis for Fall Detection System Simplification

Published on: April 6, 2020

Measuring the Kinematics of Daily Living Movements with Motion Capture Systems in Virtual Reality

Measuring the Kinematics of Daily Living Movements with Motion Capture Systems in Virtual Reality

Published on: April 5, 2018

Area of Science:

Computational psychiatry
Machine learning in healthcare
Biomedical signal processing

Background:

Depression is a global mental health challenge with significant societal burden.
While physical signs of depression are recognized, few studies use machine learning with whole-body kinematic data for recognition.
This research explores kinematic cues for aiding depression classification.

Purpose of the Study:

To construct a machine learning model for automatic depression classification using kinematic skeleton data.
To evaluate the effectiveness of different machine learning algorithms in recognizing depression from body movement.
To investigate the influence of gender and age on depression recognition accuracy.

Main Methods:

Kinect V2 device used to collect kinematic skeleton data from body joints.
Spatial and low-level features extracted from 3D coordinates.
Four machine learning models (SVM, Logistic Regression, Random Forest, Gradient Boosting) were trained and evaluated.
Performance metrics included precision, recall, sensitivity, specificity, ROC curve, and confusion matrix.

Main Results:

Gradient Boosting achieved the highest prediction accuracy of 76.92% for depression classification.
Gender-based analysis showed 66.67% accuracy for males and 71.73% for females.
Age-based classification yielded 76.92% accuracy in older adults (age >40) and 53.85% in younger adults (age <=40).

Conclusions:

Computational models utilizing Kinect-captured skeletal data can effectively classify individuals with and without depression.
Gradient Boosting demonstrated superior performance among the tested machine learning tools.
Kinematic skeletal data shows promise as an effective tool for assisting in depression analysis, particularly in older populations.