Rebecca L. Gillani
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Rebecca L. Gillani has a total of 66 co-authors. Here are the top co-authors by number of shared publications:
Brian J. Bacskai
Department of Neurology, MassGeneral Institute for Neurodegenerative Disease, Massachusetts General Hospital, Boston, MA, USA; Harvard Medical School, Boston, MA, USA
2 shared publicationsMarcelo Matiello
Department of Neurology Massachusetts General Brigham Boston Massachusetts USA
2 shared publicationsSara Whiteman
Department of Neurology, MassGeneral Institute for Neurodegenerative Disease, Massachusetts General Hospital, Boston, MA, USA
2 shared publications- Theodore J. Zwang
Massachusetts General Hospital, Department of Neurology, Boston, MA, USA
2 shared publications Aleksandar Videnovic
Department of Neurology Massachusetts General Brigham Boston Massachusetts USA
1 shared publication- Alev Erisir
Department of Psychology, University of Virginia, Charlottesville, USA
1 shared publication
Latest Publications
- Open Access
- 28 January 2026
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 13 (7) , 1412-1421
Feasibility and Tolerability of Performing Portable MRI for Neurological Disorders in an Outpatient Neurology Clinic: A Prospective Cohort
Background: Accessing brain magnetic resonance imaging (MRI) can be challenging, especially for underserved patients, which may lead to disparities in neurological diagnosis. Method: This mixed‐methods study enrolled adults with one of four neurological disorders: mild cognitive impairment or dementia of the Alzheimer type, multiple sclerosis (MS), Parkinson disease (PD), or stroke. Participants were enrolled at the Massachusetts General Hospital outpatient clinic in Boston (01/2021–08/2025) and underwent a point‐of‐care, low‐field portable MRI (pMRI) using a 0.064 T scanner (Hyperfine) and pre‐ and post‐scan surveys. For comparison, all participants had received a standard clinical brain MRI of at least 1.5T. Results: Of the 130 participants (54% male, 46% female), the mean age was 60.6 years (standard deviation (SD) = 17.5). 71% reported at least one barrier to MRI care. The median interval between pMRI and standard MRI brain imaging was 48.5 days (25th, 75th percentiles: 26, 112.5). Participants rated pMRI as highly tolerable, with 93% reporting comfort scores ≥ 7 (on a 10‐point scale), and 78% indicating willingness to undergo future pMRI. Mean comfort ratings were significantly higher for pMRI (8.4) compared to traditional MRI (7.5; p < 0.05), though the effect size was moderate ( η 2 = 0.096). No differences in tolerability or comfort were observed across disease, age, or sex groups. Common barriers to standard MRI access included transportation, cost, and scheduling, particularly among low‐income or unemployed participants. Qualitative feedback emphasized the need for better physical accommodations but broadly supported the acceptability and potential uses of pMRI. Conclusion: These findings support the feasibility, tolerability, and high patient acceptability of pMRI.
- 28 January 2026
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 13 (7) , 1412-1421
Open AccessFeasibility and Tolerability of Performing Portable MRI for Neurological Disorders in an Outpatient Neurology Clinic: A Prospective Cohort
Background: Accessing brain magnetic resonance imaging (MRI) can be challenging, especially for underserved patients, which may lead to disparities in neurological diagnosis. Method: This mixed‐methods study enrolled adults with one of four neurological disorders: mild cognitive impairment or dementia of the Alzheimer type, multiple sclerosis (MS), Parkinson disease (PD), or stroke. Participants were enrolled at the Massachusetts General Hospital outpatient clinic in Boston (01/2021–08/2025) and underwent a point‐of‐care, low‐field portable MRI (pMRI) using a 0.064 T scanner (Hyperfine) and pre‐ and post‐scan surveys. For comparison, all participants had received a standard clinical brain MRI of at least 1.5T. Results: Of the 130 participants (54% male, 46% female), the mean age was 60.6 years (standard deviation (SD) = 17.5). 71% reported at least one barrier to MRI care. The median interval between pMRI and standard MRI brain imaging was 48.5 days (25th, 75th percentiles: 26, 112.5). Participants rated pMRI as highly tolerable, with 93% reporting comfort scores ≥ 7 (on a 10‐point scale), and 78% indicating willingness to undergo future pMRI. Mean comfort ratings were significantly higher for pMRI (8.4) compared to traditional MRI (7.5; p < 0.05), though the effect size was moderate ( η 2 = 0.096). No differences in tolerability or comfort were observed across disease, age, or sex groups. Common barriers to standard MRI access included transportation, cost, and scheduling, particularly among low‐income or unemployed participants. Qualitative feedback emphasized the need for better physical accommodations but broadly supported the acceptability and potential uses of pMRI. Conclusion: These findings support the feasibility, tolerability, and high patient acceptability of pMRI.
- Open Access
- 29 August 2025
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 13 (1) , e63157
A Machine Learning Approach for Identifying People With Neuroinfectious Diseases in Electronic Health Records: Algorithm Development and Validation
Background: Identifying neuroinfectious disease (NID) cases using International Classification of Diseases billing codes is often imprecise, while manual chart reviews are labor-intensive. Machine learning models can leverage unstructured electronic health records to detect subtle NID indicators, process large data volumes efficiently, and reduce misclassification. While accurate NID classification is needed for research and clinical decision support, using unstructured notes for this purpose remains underexplored. Objective: The objective of this study is to develop and validate a machine learning model to identify NIDs from unstructured patient notes. Methods: Clinical notes from patients who had undergone lumbar puncture were obtained using the electronic health record of an academic hospital network (Mass General Brigham [MGB]), with half associated with NID-related diagnostic codes. Ground truth was established by chart review with 6 NID-expert physicians. NID keywords were generated with regular expressions, and extracted texts were converted into bag-of-words representations using n-grams (n=1, 2, 3). Notes were randomly split into training (80%), 2400 notes out of 3000, and hold-out testing (20%), 600 notes out of 3000, sets. Feature selection was performed using logistic regression with L1 regularization. An extreme gradient boosting (XGBoost) model classified NID cases, and performance was evaluated using the area under the receiver operating curve (AUROC) and the precision-recall curve (AUPRC). The performance of the natural language processing (NLP) model was contrasted with the Llama 3.2 auto-regressive model on the MGB test set. The NLP model was additionally validated on external data from an independent hospital (Beth Israel Deaconess Medical Center [BIDMC]). Results: This study included 3000 patient notes from MGB from January 22, 2010, to September 21, 2023. Of 1284 initial n-gram features, 342 were selected, with the most significant features being “meningitis,” “ventriculitis,” and “meningoencephalitis.” The XGBoost model achieved an AUROC of 0.98 (95% CI 0.96‐0.99) and AUPRC of 0.89 (95% CI 0.83‐0.94) on MGB test data. In comparison, NID identification using International Classification of Diseases billing codes showed high sensitivity (0.97) but poor specificity (0.59), overestimating NID cases. Llama 3.2 improved specificity (0.94) but had low sensitivity (0.64) and an AUROC of 0.80. In contrast, our NLP model balanced specificity (0.96) and sensitivity (0.84), outperforming both methods in accuracy and reliability on MGB data. When tested on external data from BIDMC, the NLP model maintained an AUROC of 0.98 (95% CI 0.96‐0.99), with an AUPRC of 0.78 (95% CI 0.66‐0.89). Conclusions: The NLP model accurately identifies NID cases from clinical notes. Validated across 2 independent hospital datasets, the model demonstrates feasibility for large-scale NID research and cohort generation. With further external validation, our results could be more generalizable to other institutions.
- 29 August 2025
- Published by JMIR Publications Inc. in JMIR Medical Informatics
- Vol. 13 (1) , e63157
Open AccessA Machine Learning Approach for Identifying People With Neuroinfectious Diseases in Electronic Health Records: Algorithm Development and Validation
Background: Identifying neuroinfectious disease (NID) cases using International Classification of Diseases billing codes is often imprecise, while manual chart reviews are labor-intensive. Machine learning models can leverage unstructured electronic health records to detect subtle NID indicators, process large data volumes efficiently, and reduce misclassification. While accurate NID classification is needed for research and clinical decision support, using unstructured notes for this purpose remains underexplored. Objective: The objective of this study is to develop and validate a machine learning model to identify NIDs from unstructured patient notes. Methods: Clinical notes from patients who had undergone lumbar puncture were obtained using the electronic health record of an academic hospital network (Mass General Brigham [MGB]), with half associated with NID-related diagnostic codes. Ground truth was established by chart review with 6 NID-expert physicians. NID keywords were generated with regular expressions, and extracted texts were converted into bag-of-words representations using n-grams (n=1, 2, 3). Notes were randomly split into training (80%), 2400 notes out of 3000, and hold-out testing (20%), 600 notes out of 3000, sets. Feature selection was performed using logistic regression with L1 regularization. An extreme gradient boosting (XGBoost) model classified NID cases, and performance was evaluated using the area under the receiver operating curve (AUROC) and the precision-recall curve (AUPRC). The performance of the natural language processing (NLP) model was contrasted with the Llama 3.2 auto-regressive model on the MGB test set. The NLP model was additionally validated on external data from an independent hospital (Beth Israel Deaconess Medical Center [BIDMC]). Results: This study included 3000 patient notes from MGB from January 22, 2010, to September 21, 2023. Of 1284 initial n-gram features, 342 were selected, with the most significant features being “meningitis,” “ventriculitis,” and “meningoencephalitis.” The XGBoost model achieved an AUROC of 0.98 (95% CI 0.96‐0.99) and AUPRC of 0.89 (95% CI 0.83‐0.94) on MGB test data. In comparison, NID identification using International Classification of Diseases billing codes showed high sensitivity (0.97) but poor specificity (0.59), overestimating NID cases. Llama 3.2 improved specificity (0.94) but had low sensitivity (0.64) and an AUROC of 0.80. In contrast, our NLP model balanced specificity (0.96) and sensitivity (0.84), outperforming both methods in accuracy and reliability on MGB data. When tested on external data from BIDMC, the NLP model maintained an AUROC of 0.98 (95% CI 0.96‐0.99), with an AUPRC of 0.78 (95% CI 0.66‐0.89). Conclusions: The NLP model accurately identifies NID cases from clinical notes. Validated across 2 independent hospital datasets, the model demonstrates feasibility for large-scale NID research and cohort generation. With further external validation, our results could be more generalizable to other institutions.
- Open Access
Transplantation of GABAergic Interneuron Progenitors Restores Cortical Circuit Function in an Alzheimer’s Disease Mouse Model
In addition to dementia, Alzheimer’s patients suffer from sleep impairments and aberrations in sleep-dependent brain rhythms. Deficits in inhibitory GABAergic interneuron function disrupt one of those rhythms, slow oscillation in particular, and actively contribute to Alzheimer’s progression. We tested the degree to which transplantation of healthy donor interneuron progenitors would restore slow oscillation rhythm in young APP mice. We harvested medial ganglionic eminence (MGE) progenitors from mouse embryos and transplanted them into host APP mutant cortices. 3D light-sheet and structured illumination microscopy revealed that transplanted MGE progenitors survived and matured into healthy interneurons. In vivo multiphoton calcium imaging and voltage-sensitive dye imaging showed functional integration and slow oscillation rescue in absence or presence of optogenetic stimulation. Our work provides proof-of-concept evidence that stem cell therapy may serve as a viable strategy to rescue functional impairments in cortical circuits of APP mice and potentially those of Alzheimer’s patients.
Open AccessTransplantation of GABAergic Interneuron Progenitors Restores Cortical Circuit Function in an Alzheimer’s Disease Mouse Model
In addition to dementia, Alzheimer’s patients suffer from sleep impairments and aberrations in sleep-dependent brain rhythms. Deficits in inhibitory GABAergic interneuron function disrupt one of those rhythms, slow oscillation in particular, and actively contribute to Alzheimer’s progression. We tested the degree to which transplantation of healthy donor interneuron progenitors would restore slow oscillation rhythm in young APP mice. We harvested medial ganglionic eminence (MGE) progenitors from mouse embryos and transplanted them into host APP mutant cortices. 3D light-sheet and structured illumination microscopy revealed that transplanted MGE progenitors survived and matured into healthy interneurons. In vivo multiphoton calcium imaging and voltage-sensitive dye imaging showed functional integration and slow oscillation rescue in absence or presence of optogenetic stimulation. Our work provides proof-of-concept evidence that stem cell therapy may serve as a viable strategy to rescue functional impairments in cortical circuits of APP mice and potentially those of Alzheimer’s patients.
- Open Access
Instability of excitatory synapses in experimental autoimmune encephalomyelitis and the outcome for excitatory circuit inputs to individual cortical neurons
Synapses are lost on a massive scale in the brain and spinal cord of people living with multiple sclerosis (PwMS), and this synaptic loss extends far beyond demyelinating lesions. Post-mortem studies show the long-term consequences of multiple sclerosis (MS) on synapses but do not inform on the early impacts of neuroinflammation on synapses that subsequently lead to synapse loss. How excitatory circuit inputs are altered across the dendritic tree of individual neurons under neuroinflammatory stress is not well understood. Here, we directly assessed the structural dynamics of labeled excitatory synapses in experimental autoimmune encephalomyelitis (EAE) as a model of immune-mediated cortical neuronal damage. We used in vivo two-photon imaging and a synthetic tissue-hydrogel super-resolution imaging technique to reveal the dynamics of excitatory synapses, map their location across the dendritic tree of individual neurons, and examine neurons at super-resolution for synaptic loss. We found that excitatory synapses are destabilized but not lost from dendritic spines in EAE, starting with the earliest imaging session before symptom onset. This led to dramatic changes in excitatory circuit inputs to individual cells. In EAE, stable synapses are replaced by synapses that appear or disappear across the imaging sessions or repeatedly change at the same location. These unstable excitatory inputs occur closer to one another in EAE than in healthy controls and are distributed across the dendritic tree. When imaged at super-resolution, we found that a small proportion of dendritic protrusions lost their presynapse and/or postsynapse. Our finding of diffuse destabilizing effects of neuroinflammation on excitatory synapses across cortical neurons may have significant functional consequences since normal dendritic spine dynamics and clustering are essential for learning and memory.
Open AccessInstability of excitatory synapses in experimental autoimmune encephalomyelitis and the outcome for excitatory circuit inputs to individual cortical neurons
Synapses are lost on a massive scale in the brain and spinal cord of people living with multiple sclerosis (PwMS), and this synaptic loss extends far beyond demyelinating lesions. Post-mortem studies show the long-term consequences of multiple sclerosis (MS) on synapses but do not inform on the early impacts of neuroinflammation on synapses that subsequently lead to synapse loss. How excitatory circuit inputs are altered across the dendritic tree of individual neurons under neuroinflammatory stress is not well understood. Here, we directly assessed the structural dynamics of labeled excitatory synapses in experimental autoimmune encephalomyelitis (EAE) as a model of immune-mediated cortical neuronal damage. We used in vivo two-photon imaging and a synthetic tissue-hydrogel super-resolution imaging technique to reveal the dynamics of excitatory synapses, map their location across the dendritic tree of individual neurons, and examine neurons at super-resolution for synaptic loss. We found that excitatory synapses are destabilized but not lost from dendritic spines in EAE, starting with the earliest imaging session before symptom onset. This led to dramatic changes in excitatory circuit inputs to individual cells. In EAE, stable synapses are replaced by synapses that appear or disappear across the imaging sessions or repeatedly change at the same location. These unstable excitatory inputs occur closer to one another in EAE than in healthy controls and are distributed across the dendritic tree. When imaged at super-resolution, we found that a small proportion of dendritic protrusions lost their presynapse and/or postsynapse. Our finding of diffuse destabilizing effects of neuroinflammation on excitatory synapses across cortical neurons may have significant functional consequences since normal dendritic spine dynamics and clustering are essential for learning and memory.
- 2 February 2023
- Published by Springer Nature in Nature Neuroscience
- Vol. 26 (3) , 470-480
Mapping thalamic innervation to individual L2/3 pyramidal neurons and modeling their ‘readout’ of visual input
The thalamus is the main gateway for sensory information from the periphery to the mammalian cerebral cortex. A major conundrum has been the discrepancy between the thalamus’s central role as the primary feedforward projection system into the neocortex and the sparseness of thalamocortical synapses. Here we use new methods, combining genetic tools and scalable tissue expansion microscopy for whole-cell synaptic mapping, revealing the number, density and size of thalamic versus cortical excitatory synapses onto individual layer 2/3 (L2/3) pyramidal cells (PCs) of the mouse primary visual cortex. We find that thalamic inputs are not only sparse, but remarkably heterogeneous in number and density across individual dendrites and neurons. Most surprising, despite their sparseness, thalamic synapses onto L2/3 PCs are smaller than their cortical counterparts. Incorporating these findings into fine-scale, anatomically faithful biophysical models of L2/3 PCs reveals how individual neurons with sparse and weak thalamocortical synapses, embedded in small heterogeneous neuronal ensembles, may reliably ‘read out’ visually driven thalamic input.
- 2 February 2023
- Published by Springer Nature in Nature Neuroscience
- Vol. 26 (3) , 470-480
Mapping thalamic innervation to individual L2/3 pyramidal neurons and modeling their ‘readout’ of visual input
The thalamus is the main gateway for sensory information from the periphery to the mammalian cerebral cortex. A major conundrum has been the discrepancy between the thalamus’s central role as the primary feedforward projection system into the neocortex and the sparseness of thalamocortical synapses. Here we use new methods, combining genetic tools and scalable tissue expansion microscopy for whole-cell synaptic mapping, revealing the number, density and size of thalamic versus cortical excitatory synapses onto individual layer 2/3 (L2/3) pyramidal cells (PCs) of the mouse primary visual cortex. We find that thalamic inputs are not only sparse, but remarkably heterogeneous in number and density across individual dendrites and neurons. Most surprising, despite their sparseness, thalamic synapses onto L2/3 PCs are smaller than their cortical counterparts. Incorporating these findings into fine-scale, anatomically faithful biophysical models of L2/3 PCs reveals how individual neurons with sparse and weak thalamocortical synapses, embedded in small heterogeneous neuronal ensembles, may reliably ‘read out’ visually driven thalamic input.
- 27 April 2022
- Published by SAGE Publications in Multiple Sclerosis Journal
- Vol. 28 (7) , 1146-1150
Anti-SARS-CoV-2 monoclonal antibodies for the treatment of active COVID-19 in multiple sclerosis: An observational study
Monoclonal antibodies (mAbs) against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) received emergency use authorization for the acute treatment of COVID-19. We are not aware of published data on their use in immunosuppressed people with multiple sclerosis (pwMS). We report 23 pwMS (mean age = 49 years, ocrelizumab ( n = 19), fingolimod ( n = 2), vaccinated with at least an initial series ( n = 19)) who received mAb for acute COVID-19. Following mAb receipt, approximately half recovered in <7 days (48%). There were no adverse events or deaths. Use of mAb for pwMS treated with fingolimod or ocrelizumab was not observed to be harmful and is likely helpful for treatment of acute COVID-19.
- 27 April 2022
- Published by SAGE Publications in Multiple Sclerosis Journal
- Vol. 28 (7) , 1146-1150
Anti-SARS-CoV-2 monoclonal antibodies for the treatment of active COVID-19 in multiple sclerosis: An observational study
Monoclonal antibodies (mAbs) against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) received emergency use authorization for the acute treatment of COVID-19. We are not aware of published data on their use in immunosuppressed people with multiple sclerosis (pwMS). We report 23 pwMS (mean age = 49 years, ocrelizumab ( n = 19), fingolimod ( n = 2), vaccinated with at least an initial series ( n = 19)) who received mAb for acute COVID-19. Following mAb receipt, approximately half recovered in <7 days (48%). There were no adverse events or deaths. Use of mAb for pwMS treated with fingolimod or ocrelizumab was not observed to be harmful and is likely helpful for treatment of acute COVID-19.