Preliminary Investigation of Clustering Analysis for Understanding Individual Sensory Contributions in Human Postural...

In human postural control, visual, proprioceptive, and vestibular functions cooperatively work to maintain a stable posture, known as sensory reweighting. Considering that the training effect on sensory reweighting is suggested to vary depending on the original sensory weights, examining the variations in weights across individuals is crucial. This study investigated the individual characteristics of sensory contributions more comprehensively and quantitatively by clustering analysis. A total of 122 participants performed the modified Sensory Organization Test (mSOT), which consists of the four standing conditions, and the three relative sensory weights were calculated using the center of pressure (COP) sway in each condition. Each participant's sensory weights were classified with similar sensory properties using the X-means clustering method, showing that sensory characteristics could be divided into six clusters, suggesting the existence of diverse variations. Almost all of the clusters included both age groups, those under 50 years old and those 50 years and older, indicating that the use of sensory modalities varies among individuals regardless of age.Clinical relevance- These findings could provide foundational knowledge for developing sensory reweighting training programs tailored to specific sensory profiles, as well as for delivering clear and personalized feedback in health assessments.