Figure 2
From: Chronic back pain sub-grouped via psychosocial, brain and physical factors using machine learning
(a) Heat Map and (b) Scatter plot of the distribution of sub-groups of individuals with CBP based on symptoms of depression and loneliness/social isolation. Data are at discrete intervals due to outcomes being questionnaire based. Data is presented on (a) normal item range and (b) normalised scale of 0–1. Higher scores indicate greater symptoms of depression or loneliness/social isolation. Based on the centroids of fuzzy c-means clustering, classes and colours are (1; blue) low social isolation and loneliness and moderate depressive symptoms (n = 776; 18.7%), (2; red) low loneliness/social isolation and depressive symptoms (n = 2296; 55.3%), (3; yellow) high loneliness/social isolation and moderate depressive symptoms (n = 185; 4.5%), (4; green) moderate loneliness/social isolation and high depressive symptoms (n = 297; 7.2%) and (5; pink) moderate loneliness/social isolation and low depressive symptoms (n = 602; 14.5%). Black squares on the heat map indicate no class was available at those values. The X value on the scatter plot indicates the centroid of that cluster.
