Spectral Clustering Reveals Different Profiles of Central Sensitization in Women with Carpal Tunnel Syndrome

Identification of subgroups patients with chronic pain provides meaningful insights into the characteristics a specific population, helping to identify individuals at risk chronification and determine appropriate therapeutic strategies. This paper proposes use spectral clustering (SC) distinguish (clusters) carpal tunnel syndrome (CTS), making obtained patient profiling argue about potential management implications. SC is powerful algorithm that builds similarity graph among data points (the patients), tries find subsets are strongly connected themselves, but weakly others. It was chosen due its advantages respect other simpler techniques, such as k-means, fact it has been successfully applied similar problems. Clinical (age, duration symptoms, intensity, function, symptom severity), psycho-physical (pressure thresholds—PPTs—over three main nerve trunks upper extremity, cervical spine, tunnel, tibialis anterior), psychological (depressive levels), motor (pinch tip grip force) variables were collected in 208 women clinical/electromyographic diagnosis CTS, whose symptoms usually started unilaterally eventually evolved bilateral symmetry. used clusters without any previous assumptions, yielding clusters. Patients cluster 1 exhibited worse clinical features, higher widespread pressure hyperalgesia, depressive levels, lower pinch force than two. 2 showed generalized thermal hyperalgesia Cluster 0 less hypersensitivity pain, severe more normal output (tip force). The presence different altered nociceptive processing (one group being sensitive another pain) could lead programs.