Comparative evaluation of diffusion models for detecting gray matter microstructural alterations in post-stroke depression patients
- 1Department of Radiology, Huaibei People's Hospital , , ,
Abstract
Background: Post-stroke depression (PSD) is often underdiagnosed because its gray matter microstructural abnormalities can be subtle. This study compared four diffusion models (DTI, DKI, MAP, and NODDI) to identify objective imaging biomarkers for the early detection of PSD.
Methods: Sixty-four participants were enrolled, including 23 age- and sex-matched healthy controls (HC), 26 patients with acute cerebral infarction without PSD (non-PSD; HAMD-21 < 7), and 15 patients with PSD (HAMD-21 ≥ 7). Clinical assessments (HAMD-21, ADL, MoCA, and MMSE) and diffusion spectrum imaging were performed. Intergroup differences in gray matter microstructure, diagnostic performance based on the area under the curve (AUC), and partial correlations adjusted for age, sex, and years of education were analyzed. Infarct location was not included as a covariate.
Results: Significant differences among the three groups were observed in age, years of education, and all clinical scale scores ( p < 0.05). Altered gray matter microstructure was primarily identified in the middle frontal gyrus, posterior cingulate gyrus, paracingulate gyrus and amygdala. For group discrimination, the left anterior cingulate DKI MD_p10 demonstrated the best performance for distinguishing HC from non-PSD (AUC = 0.916). The left putamen NODDI ODI_p10 showed the highest AUC (0.916) for distinguishing HC from PSD, whereas the left putamen NODDI ODI_p50 yielded the highest AUC (0.856) for distinguishing non-PSD from PSD. After FDR correction, multiple diffusion metrics were significantly correlated with MMSE and MoCA scores, whereas no imaging parameter showed a significant correlation with HAMD-21.
Conclusion: Multimodal diffusion MRI detects microstructural abnormalities within emotion- and cognition-related circuits, particularly the amygdala and cingulate cortex, across the three groups. Temporolimbic NODDI parameters showed relatively higher AUC values in distinguishing PSD from HC, whereas anterior cingulate DKI metrics showed relatively higher AUC values in differentiating non-PSD stroke patients from HC. Combining multiple diffusion models may aid in the identification of PSD; however, these findings require further validation because infarct location was not adjusted for in the analyses.
Related
- Spectroscopic evidence of microstructural alterations in retrieved 1,000-cSt PDMS Silicone oil tamponades
- Psychological support for depression and anxiety across chronic illness contexts: a comparative evaluation of three general-purpose AI chatbots
- A Systems Biology Model Linking Peripheral NIMETOX Biomarkers to Multilevel Brain Structural and Functional Alterations Across Major Depression and Schizophrenia
- Longitudinal White Matter Diffusion Alterations in Medication Overuse Headache Before and After Medication Withdrawal: A Diffusion Tensor Imaging Study
- Along-tract white-matter microstructural differences in primary trigeminal neuralgia: a diffusion MRI tractometry study.
BibTeX
@article{zhang2026comparative,
title = {{Comparative evaluation of diffusion models for detecting gray matter microstructural alterations in post-stroke depression patients}},
author = {Fang Zhang and Jing Zhang and Lei Zhang and Hengjun Jin and Daqing Li and Wei Zhao},
journal = {Frontiers in neurology},
year = {2026},
month = sep,
doi = {10.3389/fneur.2026.1831557},
url = {https://europepmc.org/article/PMC/PMC13661199},
note = {CC BY},
}
RIS
TY - JOUR
TI - Comparative evaluation of diffusion models for detecting gray matter microstructural alterations in post-stroke depression patients
AU - Fang Zhang
AU - Jing Zhang
AU - Lei Zhang
AU - Hengjun Jin
AU - Daqing Li
AU - Wei Zhao
PY - 2026
DA - 2026/09/30
JO - Frontiers in neurology
DO - 10.3389/fneur.2026.1831557
UR - https://europepmc.org/article/PMC/PMC13661199
AB - Background: Post-stroke depression (PSD) is often underdiagnosed because its gray matter microstructural abnormalities can be subtle. This study compared four diffusion models (DTI, DKI, MAP, and NODDI) to identify objective imaging biomarkers for the early detection of PSD. Methods: Sixty-four participants were enrolled, including 23 age- and sex-matched healthy controls (HC), 26 patients with acute cerebral infarction without PSD (non-PSD; HAMD-21 < 7), and 15 patients with PSD (HAMD-21 ≥ 7). Clinical assessments (HAMD-21, ADL, MoCA, and MMSE) and diffusion spectrum imaging were performed. Intergroup differences in gray matter microstructure, diagnostic performance based on the area under the curve (AUC), and partial correlations adjusted for age, sex, and years of education were analyzed. Infarct location was not included as a covariate. Results: Significant differences among the three groups were observed in age, years of education, and all clinical scale scores ( p < 0.05). Altered gray matter microstructure was primarily identified in the middle frontal gyrus, posterior cingulate gyrus, paracingulate gyrus and amygdala. For group discrimination, the left anterior cingulate DKI MD_p10 demonstrated the best performance for distinguishing HC from non-PSD (AUC = 0.916). The left putamen NODDI ODI_p10 showed the highest AUC (0.916) for distinguishing HC from PSD, whereas the left putamen NODDI ODI_p50 yielded the highest AUC (0.856) for distinguishing non-PSD from PSD. After FDR correction, multiple diffusion metrics were significantly correlated with MMSE and MoCA scores, whereas no imaging parameter showed a significant correlation with HAMD-21. Conclusion: Multimodal diffusion MRI detects microstructural abnormalities within emotion- and cognition-related circuits, particularly the amygdala and cingulate cortex, across the three groups. Temporolimbic NODDI parameters showed relatively higher AUC values in distinguishing PSD from HC, whereas anterior cingulate DKI metrics showed relatively higher AUC values in differentiating non-PSD stroke patients from HC. Combining multiple diffusion models may aid in the identification of PSD; however, these findings require further validation because infarct location was not adjusted for in the analyses.
SN - 1664-2295
ER -