Fig. 5: Behavioral Tractography and Diffusion Analysis across three bundles describing population-specific behavioral dynamics across High-vs-Low Intensity and Predominant Positive-vs-Negative subgroups.

Figures have been derived using a dedicated open-source software (https://dev.mlnetwork-diplab.ch/) developed by the authors and available at (https://github.com/andreaimparato/Behavioral-Tractography-Toolbox/tree/main/MLNetwork). For each figure we provide a link to an online platform allowing 3D interactive network visualization. Comparison of population-specific trajectories of Behavioral-Tractography (BT) Bundles (7–9) in High-Symptom-Intensity vs Low-Symptom-Intensity subgroups defined according to SIPS-Dimension-1; and Predominantly-Positive-Symptoms vs Predominantly-Negative-Symptoms subgroups defined according to SIPS-Dimension-2. 1–3: Results of Behavioral-Tractography trajectory for Bundles 7-9 in High-Symptom-Intensity (A), Low-Symptom-Intensity (B), Predominantly-Positive-Symptoms (C) and Predominantly-Negative-Symptoms(D) subgroups; Green–yellow coding indicates path progression from TL1 to TL2 (Start-TL1 ➔ Exit-TL1: Dark Green; Exit-TL1 ➔ Entry-TL2: Light Green; Entry-TL2 ➔ End-TL2: Yellow). 1A–1D: Bundle 7: TL1- Lack-Cognitive-Wellbeing ➔TL2-Cognitive-Distress. 2A–2D: Bundle 8: TL1-Cognitive-Distress➔TL2-Lack-Affective-Wellbeing. 3A–3D: Bundle 9: TL1-Cognitive-Distress➔TL2- Lack-Cognitive-Wellbeing. 4A, 4B: Statistical comparison of average trajectories across nine BT-Bundles between High-Symptom-Intensity and Low-Symptom-Intensity (A)/Predominantly-Positive-Symptoms and Predominantly-Negative-Symptoms (B). For each bundle, we computed the ratio of between-group to within-group Euclidian distances between Bundle specific pathways across population. Significance was tested via 500-permutation null distribution; p values indicate the proportion of permutations exceeding observed ratios. Bundles with p < 0.05 differentiate populations while maintaining within-group consistency, reflecting distinct symptom-interaction dynamics. 5A–5D: Results of Behavioral-Diffusion-Analysis (BDA) for variables in LSI (5A), HSI (5B), PNS (5A) and PPS (5B). Warm colors identify variables that drive significant BT differences in Bundle-9 (TL1- Lack-Cognitive-Wellbeing ➔TL2-Cognitive-Distress). Cold colors identify variables that drive significant BT differences in Bundle-7 (TL1-Cognitive-Distress➔TL2- Lack-Cognitive-Wellbeing). BDA trajectories reflecting differential functional role of TL1-Variables connect TL1 to TL2, while BDA trajectories reflecting differential functional role of TL2-Variables connected TL2 to TL3. The individual PC paths that contribute to average BDA trajectories are represented in Black. Links to 3D interactive visualization: HSI: https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim1_cl7/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim1_cl8/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim1_cl9/. https://dev.mlnetwork-diplab.ch/3dvisualizer/qpa_dim1_hsi/. LSI: https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim2_cl7/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim2_cl8/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim1_dim2_cl9/. https://dev.mlnetwork-diplab.ch/3dvisualizer/qpa_dim1_lsi/. PNS: https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim1_cl7/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim1_cl8/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim1_cl9/. https://dev.mlnetwork-diplab.ch/3dvisualizer/qpa_dim2_pns/. PPS: https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim2_cl7/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim2_cl8/. https://dev.mlnetwork-diplab.ch/3dvisualizer/dim2_dim2_cl9/. https://dev.mlnetwork-diplab.ch/3dvisualizer/qpa_dim2_pps/.