跳到正文
原文
Medical Xpress·· 2026-10-21AI 评分53

Duke团队在npj Parkinson's Disease发表TULIP三维无标记动作量化评估帕金森病研究

Measuring Parkinson's motor symptoms more objectively with 3D movement tracking

AI 导读

Duke University团队在npj Parkinson's Disease发表研究,介绍TULIP三维无标记动作量化系统。该研究使用同步摄像头记录手部、前臂及步态运动,通过机器学习重建三维姿态,将运动转化为角度、距离等精确量化指标。结果显示该量化测量可自动区分帕金森病参与者与无病者,并用于估计疾病严重度与病程,团队称其可为运动评估与临床试验提供更敏感的数字生物标志物。

正文
3D movement tracking system objectively evaluates Parkinson's disease progression
Two researchers in the Dunn lab demonstrate how TULIP can track movement, such as the opening and closing of a fist. Credit: Duke University

When clinicians evaluate someone with Parkinson's disease, they typically ask patients to perform a series of movements, like making a fist, rotating their forearms or walking across a room. They'll then assign a score from 0, indicating normal movement, to 4, indicating severe impairment.

This scoring system, called the Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS), is simple, allowing clinicians around the world to assess a patient's motor symptoms. But because it places a broad range of movement characteristics into a small number of categories, it can be difficult to capture subtle changes in movement or precisely monitor how Parkinson's disease progresses over time.

"One clinician's definition of score 1 (written as 'minor impairment' in the score guidelines) can vary a lot from others, so while the guidelines standardize scoring, in the end there is still some degree of subjectivity," said Kyungdo Kim, a Ph.D. student in Duke University's Department of Biomedical Engineering. "If we want to more precisely track how this disease progresses, we need an explainable, objective system for quantifying movement."

Measuring movement in three dimensions

A new approach, developed by Kim, Dr. Kyle Mitchell, an associate professor of neurology, Timothy Dunn, an assistant professor of biomedical engineering, and others in the Dunn lab could provide just that. The research, published in npj Parkinson's Disease, introduces a system called TULIP that captures and quantifies human movement in three dimensions.

Short for Three-dimensional Understanding and Learning of Impairments in Parkinson's disease, TULIP uses synchronized cameras to record a person from multiple angles without any markers. Because the patient doesn't need to wear any sensors or special equipment, the system can capture the participant's natural, unencumbered movement. Machine learning algorithms then reconstruct the person's movements as a three-dimensional pose.

The resulting 3D data allow the researchers to convert these movements into precise, quantitative measurements, such as degrees of rotation or centimeters of movement.

"Using our system, instead of saying 'Oh, your gait score is 1,' instead we can say, 'because your leg angle changes from 45 to 60 degrees, your gait has improved and your medication may be helping,'" said Kim. "It gives us a way to measure the disease's progression using the movement itself."

Three examinations reveal disease patterns

Working with researchers at the Duke Neurology Clinic, the team used TULIP to analyze participants performing three motor examinations commonly used to assess Parkinson's disease. For hand movements, the system measured characteristics of repeated opening and closing. It captured the forearm's inner and outer rotation. During walking, it analyzed patterns of whole-body movement and gait.

Combining measurements across the three examinations provided a more comprehensive picture than relying on any single movement task. Because each examination captures a different aspect of motor function, together they provide multiple windows into how Parkinson's disease affects movement.

These measurements allowed the researchers to automatically distinguish people with Parkinson's disease from those without it and estimate disease severity and duration. Changes in movement amplitude and speed, for example, became more pronounced with increasing disease duration.

"There isn't a widely accepted lab or imaging biomarker for Parkinson's disease, and TULIP gives us a way to move beyond a subjective clinical assessment and actually measure movement in a way that is quantitative and interpretable," said Dunn. "While Parkinson's disease is our current focus, this could be useful for other disorders that affect movement, like Alzheimer's disease, stroke and other neurological injuries."

Tracking treatment effects over time

The team envisions TULIP as a potential complement to the existing clinical assessments. More precise measurements could help physicians monitor changes in an individual patient over time and give clinical researchers a more sensitive tool for evaluating whether a treatment is improving motor function. This could be especially valuable in clinical trials, where detecting small changes can make it easier to determine if a treatment is having an effect.

Although the team will continue to refine its system by testing on larger and more diverse patient populations, it remains optimistic about the clinical potential of TULIP.

"This work provides a quantifiable and explainable framework for characterizing mobility in Parkinson's disease, and establishes a foundation for digital biomarkers that can support long-term monitoring of neurological disorders in both clinical and research settings," said Kim.

Publication details

Kyungdo Kim et al., Explainable 3D multi-scale movement quantification automates motor assessments and estimates Parkinson's disease duration, npj Parkinson's Disease (2026). DOI: 10.1038/s41531-026-01512-7

Journal information: npj Parkinson's Disease

Key medical concepts

Parkinson's Disease

Clinical categories

Neurology

Who's behind this story?

Lisa Lock

Lisa Lock

BA art history, MA material culture. Former museum editor, paramedic, and transplant coordinator. Editing for Science X since 2021. Full profile →

Robert Egan

Robert Egan

Bachelor's in mathematical biology, Master's in creative writing. Well-traveled with unique perspectives on science and language. Full profile →

Citation: Measuring Parkinson's motor symptoms more objectively with 3D movement tracking (2026, October 20) retrieved 20 October 2026 from https://medicalxpress.com/news/2026-10-parkinson-motor-symptoms-3d-movement.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

来源:Medical Xpress · medicalxpress.com