Related Experiment Video
Updated: Jan 17, 2026
Sit-to-stand-and-walk from 120% Knee Height: A Novel Approach to Assess Dynamic Postural Control Independent of Lead-limb
Published on: August 30, 2016
Biokinetic Profiles in Patellofemoral Pain Patients During a Step-Down Task: An Unsupervised Machine Learning
Leonardo Metsavaht1, Gustavo Leporace2, Felipe F Gonzalez2
1Departamento de Diagnóstico por Imagem, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, State of São Paulo, Brazil.
Background:
There is a lack of consensus regarding the optimal identification of motion-based subgroups (biokinetic profiles) of patients with patellofemoral pain (PFP) to guide effective management strategies.
Purpose:
To investigate different biokinetic profiles among patients with PFP during a step-down task and compare their clinical and physical characteristics.
Study Design:
Descriptive laboratory study.
Methods:
A total of 49 patients with PFP had their 3-dimensional kinematics assessed during a step-down task using an optoelectronic system. The variables analyzed were trunk and lower limb joint angles. Self-organizing maps and K-means clustering techniques were used to identify distinct biokinetic profiles. Clinical characteristics compared among profiles were hip/knee isometric strength and passive range of motion, descriptive characteristics, and patient-reported outcome measures.
Results:
Four biokinetic profiles were identified for the step-down task in patients with PFP. Profile 1 (Balanced Alignment Profile) exhibited a trunk-hip-knee aligned movement pattern, with increased knee flexion (P < .05). This profile presented the highest International Knee Documentation Committee (IKDC) Subjective Form and the lowest visual analog scale (VAS)-Pain scores (P < .05). Profile 2 (Trunk-hip-knee Compensation Profile) and Profile 3 (Pelvic-Hip Interactor Profile) exhibited increased dynamic knee valgus during the step-down task. However, Profile 2 presented limited trunk (P < .05) and knee flexion, while Profile 3 presented increased anterior pelvic tilt (P < .05) and trunk flexion. Profile 2 had excessive passive hip internal rotation (P < .05) and a majority of women (P < .05), while Profile 3 exhibited increased isometric hip and knee strength (P < .05) and lower levels of pain. Profile 4 (Protective Movement Profile) exhibited a possibly protective adaptation, showing decreased hip, knee, and pelvis peak angles and decreased ipsilateral trunk tilt (P < .05), as well as the lowest IKDC scores and the highest VAS-Pain scores (P < .05).
Conclusion:
This study identified 4 clinically relevant biokinetic profiles of patients with PFP based on their kinematics during a step-down task. The principal clinical value of this study is the development of a functional classification for patients with PFP. Each profile exhibits characteristics that may provide valuable insights for clinicians to implement targeted interventions and improve patient care. Acknowledging the variability in movement profiles and their implications in PFP underscores the importance of moving beyond one-size-fits-all treatment strategies.
Clinical Relevance:
This study uncovers previously unknown movement-based biokinetic profiles in patellofemoral pain and identifies several modifiable clinical characteristics associated with each profile, which could be addressed through targeted interventions.
