MRI-based machine learning analysis of perivascular spaces and their link to sleep disturbances, dementia, and mental distress in young adults with long-time mobile phone use
ObjectiveLong-term mobile phone use (LTMPU) has been linked to sleep disorders, mood disorders, and cognitive impairment, with MRI-detected enlarged perivascular spaces (EPVSs) as potential imaging markers. This study investigated computational MRI-visible EPVSs and their association with sleep dist...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-04-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnins.2025.1555054/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | ObjectiveLong-term mobile phone use (LTMPU) has been linked to sleep disorders, mood disorders, and cognitive impairment, with MRI-detected enlarged perivascular spaces (EPVSs) as potential imaging markers. This study investigated computational MRI-visible EPVSs and their association with sleep disturbance, dementia, and mental distress in young adults with LTMPU.MethodsThis retrospective study included 82 LTMPU patients who underwent MRI scans and assessments using six clinical scales: Montreal Cognitive Assessment (MoCA), Pittsburgh Sleep Quality Index (PSQI), Insomnia Severity Index (ISI), Epworth Sleepiness Scale (ESS), Hamilton Anxiety (HAM-A), and Hamilton Depression (HAM-D). Deep learning algorithms segmented EPVSs lesions, extracting quantitative metrics (count, volume, mean length, and mean curvature) across 17 brain subregions. Correlation analyses explored relationships between EPVSs indicators and clinical measurements. The BrainNet Viewer tool highlighted significant brain subregions and EPVSs traits linked to dementia, sleep disturbance, and mental distress.ResultsCorrelation analyses identified 23 significant indicator pairs (FDR-adjusted p < 0.05), including associations between nine EPVSs characteristics and MoCA scores: four with the PSQI, one with the ISI, three with the ESS, four with the HAM-A, and two with the HAM-D. Regression analyses revealed seven significant EPVSs features, with three linked to cognitive impairment: mean EPVSs length in the left basal ganglia and mean length/curvature in the left temporal lobe. Also, the mean EPVSs length in the left frontal lobe could indicate insomnia, sleepiness, and anxiety.ConclusionComputational EPVSs metrics offer insights into the EPVSs pathophysiology and its links to mood disorders, sleep disturbances, and cognitive impairment in LTMPU patients. These findings also highlight potential connections between EPVSs, excessive daytime sleepiness, and anxiety, contributing to a comprehensive understanding of these multifaceted conditions. |
|---|---|
| ISSN: | 1662-453X |