MKMahir H. Khan

Mahir H. Khan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Mahir H. Khan has a total of 128 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    MAESTRO: A Public, Generalizable Model for Stroke Lesion Segmentation from T1 MRI Across the Recovery Continuum

    by Mahir H Khan, Octavio Marin-Pardo, Stuti Chakraborty, Kaitlin Lee, Sarah Lee, Nandita Raman, Juan Eugenio Iglesias, Sook-Lei Liew

    Accurate stroke lesion segmentation is essential for large-scale neuroimaging studies, yet manual delineation remains labor-intensive, and existing automated methods often struggle to generalize across imaging protocols and stages of recovery. We developed MAESTRO, a deep learning framework for automated lesion segmentation across the stroke recovery continuum using T1-weighted (T1) MRI alone. We hypothesized that combining a transformer-based architecture with an image augmentation strategy would improve segmentation accuracy and robustness under heterogeneous imaging conditions. T1 MRI scans and expert-traced lesion masks from 955 stroke participants across 33 international cohorts were used to train and evaluate MAESTRO within the open-source nnU-Net framework. Performance was evaluated on a held-out test set using spatial and volumetric agreement metrics. An exploratory human-in-the-loop (HITL) evaluation compared correction of MAESTRO-generated segmentations with manual tracing from scratch. MAESTRO achieved the strongest performance across several evaluated model configurations, providing the most accurate lesion localization and lesion volume estimates (median Dice = 0.686; Pearson r = 0.861; ICC = 0.792). Segmentation performance was sensitive to lesion size and stroke chronicity but remained robust across diverse imaging conditions. Additionally, using a HITL workflow to correct MAESTRO segmentations reduced annotation time by 47.4% compared to manual tracing while improving accuracy relative to both automated and manual workflows. MAESTRO is publicly available to enable robust, automated stroke lesion segmentation from T1 MRI. When combined with human review and correction, MAESTRO offers a practical approach for generating standardized, high-quality lesion annotations, helping reduce a major barrier to large-scale stroke imaging studies.

  • Open Access

    A composite measure of cerebral small vessel disease predicts cognitive change after stroke

    by Mahir H. Khan, Stuti Chakraborty, Octavio Marin-Pardo, Giuseppe Barisano, Michael R. Borich, James H. Cole, Steven C. Cramer, Emily E. Fokas, Niko H. Fullmer, Leticia Hayes, et al.

    Post-stroke cognitive recovery is difficult to predict using focal lesion characteristics alone. The brain’s capacity to maintain cognitive function depends also on structural integrity of the whole brain. One way to measure brain health is through the severity of cerebral small vessel disease (CSVD) markers, which reflect aging-related pathologies that erode structural integrity. Here, we propose a composite measure of CSVD (cCSVD) integrating three independently validated biomarkers automatically quantified using T1-weighted MRIs: white matter hyperintensity volume (WMH; representing vascular injury), perivascular space count (PVS; putative glymphatic clearance), and brain-predicted age difference (brain-PAD; structural atrophy). We hypothesize that cCSVD, which captures the shared variance across these CSVD biomarkers, will be a robust indicator of whole-brain structural integrity and predict cognitive changes 3 months after stroke. We analyzed 65 early subacute stroke survivors with assessments within 21 days (baseline) and at 90 days (follow-up) post-stroke. WMH volume, PVS count, and brain-PAD were quantified from baseline T1-weighted MRIs, and then residualized for age, sex, days since stroke, and intracranial volume. Principal component analysis (PCA) of the residualized biomarkers was used to derive cCSVD. Beta regression with stability selection using LASSO was used to model three outcomes: baseline Montreal Cognitive Assessment (MoCA) scores, follow-up MoCA scores, and longitudinal change (follow-up score adjusted for baseline score). Logistic regression was used to test if baseline cCSVD predicted improvement in those with baseline cognitive impairment (MoCA < 26). The PCA revealed that the first principal component (PC1) explained 43.1% of the total variance among WMH volume, PVS count, and brain-PAD. The three biomarkers contributed nearly equally to PC1, which was subsequently used as the baseline cCSVD score. Lower baseline cCSVD was significantly associated with better MoCA scores at follow-up (β = −0.19, p = 0.009), even after adjusting for baseline MoCA (β = −0.12, p = 0.042), and, importantly, outperformed all individual biomarkers. Furthermore, lower cCSVD at baseline significantly increased the likelihood of improving to cognitively unimpaired status at three months (OR = 0.34, p = 0.036), independent of age and education. The composite CSVD captures the additive impact of vascular injury, glymphatic dysfunction, and structural atrophy on recovery in a way that individual measures do not. cCSVD accounts for shared variance across these domains, reflecting a patient’s latent capacity for cognitive recovery, where relative integrity in one CSVD domain may mitigate effects of another. This automated, T1-based framework offers a scalable tool for predicting post-stroke recovery.

  • Open Access

    Estimating white matter hyperintensities volume in individuals with stroke using T1-weighted images

    by Mahir H. Khan, Stuti Chakraborty, Jennifer K. Ferris, Lara A. Boyd, Mohamed Salah Khlif, Amy Brodtmann, Michael R. Borich, James H. Cole, Steven C. Cramer, Niko H. Fullmer, et al.


  • Open Access

    Associations between contralesional neuroplasticity and motor impairment through deep learning-derived MRI regional brain age in chronic stroke (ENIGMA): a multicohort, retrospective, observational study

    by Gilsoon Park, Mahir H Khan, Justin W Andrushko, Nerisa Banaj, Michael R Borich, Lara A Boyd, Amy Brodtmann, Truman R Brown, Cathrin M Buetefisch, Adriana B Conforto, et al.


  • Open Access

    Estimating white matter hyperintensities volume in individuals with stroke using T1-weighted images

    by Mahir H. Khan, Stuti Chakraborty, Jennifer K. Ferris, Lara A. Boyd, Mohamed Salah Khlif, Amy Brodtmann, Michael R. Borich, James H. Cole, Steven C. Cramer, Niko H. Fullmer, et al.

    Stroke recovery outcomes vary across individuals, motivating the search for biomarkers that can improve prediction. White matter hyperintensities (WMH) volume is a leading biomarker candidate, with FLAIR MRI typically used for WMH segmentation; however, T1-weighted (T1) MRI is often more available. Therefore, we evaluated the performance of two automated WMH segmentation methods (WMH-SynthSeg and SAMSEG) to determine whether WMH volume can be reliably estimated using T1 alone. We analyzed imaging data from 227 stroke patients across three datasets spanning early subacute to chronic recovery, each with gold-standard WMH masks and stroke lesion masks manually traced on available T1 and FLAIR scans. WMH was segmented using T1 only as input to WMH-SynthSeg and SAMSEG, as well as using both T1 and FLAIR as input to SAMSEG, as previously implemented in stroke recovery research. Automated WMH segmentations were compared to the gold-standard WMH mask: accuracy was assessed using Dice similarity index (SI) and cluster-level false negative ratio, while agreement was assessed using intraclass correlation, Pearson’s correlation, and volume ratio. We used linear mixed-effects models to evaluate whether SI was influenced by factors such as WMH volume, stroke lesion volume, WMH contrast, age, sex, and days since stroke, with dataset as a random effect. WMH-SynthSeg using T1-only input produced more accurate and reliable WMH segmentations compared to SAMSEG with T1-only input and performed comparably to SAMSEG using both T1 and FLAIR input. WMH-SynthSeg using T1-only input may be used for WMH volume estimation in stroke recovery research in the absence of multimodal imaging. Highlights – WMH volume, often assessed via multimodal imaging, predicts post-stroke outcomes – T1-only methods would facilitate WMH analysis if multimodal MRI is unavailable – It is unclear how T1-only methods perform in brains with stroke lesions – We show T1-based WMH-SynthSeg estimates strongly agree with gold standard methods – Accuracy was stable across stroke lesion sizes but varied with WMH volume/contrast

  • Open Access

    Brain Age Is Longitudinally Associated With Sensorimotor Impairment and Mild Cognitive Impairment in Subacute Stroke

    by Octavio Marin‐Pardo, Mahir H. Khan, Stuti Chakraborty, Michael R. Borich, Mayerly Castillo, James H. Cole, Steven C. Cramer, Miranda R. Donnelly, Emily E. Fokas, Niko H. Fullmer, et al.


  • Open Access

    Larger perivascular space volume fraction is associated with worse post-stroke sensorimotor outcomes: An ENIGMA analysis

    by Stuti Chakraborty, Jeiran Choupan, Octavio Marin-Pardo, Mahir H. Khan, Giuseppe Barisano, Bethany P. Tavenner, Miranda R. Donnelly, Aisha Abdullah, Justin W. Andrushko, Nerisa Banaj, et al.

    BACKGROUND: Perivascular Spaces (PVS) are a marker of cerebral small vessel disease (CSVD) that are visible on brain imaging. Larger PVS has been associated with poor quality of life and cognitive impairment post-stroke. However, the association between PVS and post-stroke sensorimotor outcomes has not been investigated. METHODS: 602 individuals with a history of stroke across 24 research cohorts from the ENIGMA Stroke Recovery Working Group were included. PVS volume fractions were obtained using a validated, automated segmentation pipeline from the basal ganglia (BG) and white matter centrum semiovale (CSO), separately. Robust mixed effects regressions were used to a) examine the cross-sectional association between PVS volume fraction and post-stroke sensorimotor outcomes and b) to examine whether PVS volume fraction was associated with other measures of CSVD and overall brain health (e.g., white matter hyperintensities [WMHs], brain age [measured by predicted age difference, brain-PAD]). RESULTS: Larger PVS volume fraction in the CSO, but not BG, was associated with worse post-stroke sensorimotor outcomes (b = -0.06, p = 0.047). Higher burden of deep WMH (b = 0.25, p <0.001), periventricular WMH (b = 0.16, p <0.001) and higher brain-PAD (b = 0.09, p <0.001) were associated with larger PVS volume fraction in the CSO. CONCLUSIONS: Our data show that PVS volume fraction in the CSO is cross-sectionally associated with sensorimotor outcomes after stroke, above and beyond standard lesion metrics. PVS may provide insight into how the overall vascular health of the brain impacts inter-individual differences in post-stroke sensorimotor outcomes.

  • Open Access

    Brain age is longitudinally associated with sensorimotor impairment and mild cognitive impairment in subacute stroke

    by Octavio Marin-Pardo, Mahir H. Khan, Stuti Chakraborty, Michael R. Borich, Mayerly Castillo, James H. Cole, Steven C. Cramer, Miranda R. Donnelly, Emily E. Fokas, Niko H. Fullmer, et al.

    BACKGROUND: Brain age, a proxy of overall brain health estimated from structural neuroimaging, has been associated with sensorimotor performance in chronic stroke. Similarly, post-stroke cognitive outcomes have been associated with accelerated brain aging. However, the relationships between brain age, sensorimotor, and cognitive outcomes in early subacute stroke (<3 months after onset) are less understood. METHODS: In this work, we investigated associations between stroke survivors’ brain-predicted age difference (brain-PAD, quantified as a person’s brain age minus their chronological age) and longitudinal measurements of motor impairment (Fugl-Meyer Upper Extremity Assessment [FMUE]) and cognitive impairment (Montreal Cognitive Assessment [MoCA]) in subacute stroke. We used high-resolution T1-weighted MRIs from 44 participants at baseline and three months after stroke onset to investigate associations between brain-PAD, MoCA, and FMUE scores with robust linear mixed-effects regression models and mediation analyses. RESULTS: We found negative associations between baseline brain-PAD and FMUE at baseline (β=-0.87, p=0.029) and three months (β=-0.87, p=0.011). Baseline brain-PAD was also negatively correlated with MoCA at three months (β=-0.13, p=0.015) but not at baseline (β=-0.11, p=0.141). Baseline brain-PAD was not associated with changes in FMUE (β=-0.01, p=0.930) or MoCA (β=-0.03, p=0.579). Finally, MoCA was not associated with FMUE at either time point, nor did it mediate the relationship between brain-PAD and FMUE. CONCLUSION: Overall, we show that baseline brain age predicts both motor and cognitive outcomes at three months. However, motor and cognitive outcomes are not directly associated with one other. This suggests that brain age is representative of changes in multiple, distinct neurological pathways post-stroke. Further research with longer time intervals is needed to examine whether brain age also predicts chronic stroke outcomes.

  • Open Access

    Greater lesion damage is bidirectionally related with accelerated brain aging after stroke

    by Mahir H. Khan, Octavio Marin-Pardo, Stuti Chakraborty, Michael R. Borich, Mayerly Castillo, James H. Cole, Steven C. Cramer, Miranda R. Donnelly, Emily E. Fokas, Niko H. Fullmer, et al.

    Regional neuron loss following stroke can result in remote brain changes due to diaschisis and secondary brain atrophy. Whole brain changes post-stroke can be captured by the predicted brain age difference (brain-PAD), a neuroimaging-derived biomarker of global brain health previously associated with poorer chronic stroke outcomes. We hypothesized that greater lesion damage would be longitudinally associated with worsening brain-PAD during subacute stroke, and conversely, that poorer baseline brain-PAD would be associated with enlarged lesion damage. We prospectively collected MRIs from 47 stroke patients across three sites within 3 weeks (baseline) and at 3 months (follow-up) post-stroke. Predicted brain age was estimated via a pretrained ridge regression model using 77 morphological features. Brain-PAD was calculated as predicted age minus chronological age. Robust linear mixed effects regression models were used to examine relationships between infarct volume and brain-PAD, adjusting for age, sex, time, and intracranial volume at baseline. Larger baseline infarct volume was associated with accelerated brain aging at 3 months (β=0.87, p=0.023). Conversely, larger baseline brain-PAD predicted larger increase in infarct volume at 3 months (β=0.02, p=0.009). These findings reveal a bidirectional relationship between focal stroke damage and global brain health during the subacute period, underscoring the importance of assessing both.


  • Open Access

    Association of daily physical activity with brain volumes and cervical spinal cord areas in multiple sclerosis

    by Valerie J Block, Shuiting Cheng, Jeremy Juwono, Richard Cuneo, Gina Kirkish, Amber M Alexander, Mahir Khan, Amit Akula, Eduardo Caverzasi, Nico Papinutto, et al.

    Background: Remote activity monitoring has the potential to evaluate real-world, motor function, and disability at home. The relationships of daily physical activity with spinal cord white matter and gray matter (GM) areas, multiple sclerosis (MS) disability and leg function, are unknown. Objective: Evaluate the association of structural central nervous system pathology with ambulatory disability. Methods: Fifty adults with progressive or relapsing MS with motor disability who could walk >2 minutes were assessed using clinician-evaluated, patient-reported outcomes, and quantitative brain and spinal cord magnetic resonance imaging (MRI) measures. Fitbit Flex2, worn on the non-dominant wrist, remotely assessed activity over 30 days. Univariate and multivariate analyses were performed to assess correlations between physical activity and other disability metrics. Results: Mean age was 53.3 years and median Expanded Disability Status Scale (EDSS) was 4.0. Average daily step counts (STEPS) were highly correlated with EDSS and walking measures. Greater STEPS were significantly correlated with greater C2-C3 spinal cord GM areas (ρ = 0.39, p = 0.04), total cord area (TCA; ρ = 0.35, p = 0.04), and cortical GM volume (ρ = 0.32, p = 0.04). Conclusion: These results provide preliminary evidence that spinal cord GM area is a neuroanatomical substrate associated with STEPS. STEPS could serve as a proxy to alert clinicians and researchers to possible changes in structural nervous system pathology.