Marcus C. Ng
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
Research Subjects
Research Interests
Marcus C. Ng has a total of 158 co-authors. Here are the top co-authors by number of shared publications:
Darion Toutant
Biomedical Engineering University of Manitoba Winnipeg Manitoba Canada
4 shared publications- Aline Herlopian
Department of Neurology, Comprehensive Epilepsy Center, Yale University School of Medicine, New Haven, CT, USA
3 shared publications - Fábio Augusto Nascimento
Washington University School of Medicine St. Louis Missouri USA
3 shared publications - Jonathan J. Halford
Medical University of South Carolina, Charlestown, SC, USA
3 shared publications - Michael Brandon Westover
Department of Neurology, Stanford University, Stanford, CA, United States of America
3 shared publications Milena K. Pavlova
Division of Epilepsy, EEG, and Sleep Neurology, Department of Neurology, Brigham and Women’s Faulkner Hospital, Harvard Medical School, Boston, MA, USA
3 shared publications
Latest Publications
- Open Access
- 28 October 2026
- Published by SAGE Publications in Journal of Biological Rhythms
Seizures, Sleep, and the Midnight Sun
The well-known bidirectional link between seizures and sleep may be most conspicuous in environments where the chief zeitgeber – sunlight – undergoes extreme but consistent changes over the course of a year. Routine circannual encounters with the midnight sun and midday darkness strategically position Earth’s polar regions as natural laboratories to study this effect. We reviewed 2337 studies that related status epilepticus (SE), light exposure, and sleep in people in the Arctic or Antarctic. Status epilepticus represented the most extreme – and arguably incontrovertible – expression of seizure. After applying filters and exclusion criteria, 5 studies were included for SE and 36 for sleep. Status epilepticus was correlated with increasing daylength going into summer (ie, brighter sunlight vector ) at high latitudes. Sunlight and chronotype demonstrated a linear relationship, with more sunlight going into summer (ie, brighter sunlight vector) advancing chronotype, less sunlight going into winter (ie, darker sunlight vector ) delaying chronotype, and extreme cases becoming free-running in either vector direction. U-shaped relationships were also surmised, with increased sleep fragmentation and decreased sleep duration for both brighter and darker sunlight vectors. A parsimonious synthesis is that a brighter sunlight vector advances chronotype to shorten and fragment sleep, thereby triggering more SE and seizures. Although these data are derived from Earth’s poles, they are not specific to polar regions; rather, they hold ramifications for everyone on planet Earth – 1 in 10 of whom will have a seizure in their lifetime.
- 28 October 2026
- Published by SAGE Publications in Journal of Biological Rhythms
Open AccessSeizures, Sleep, and the Midnight Sun
The well-known bidirectional link between seizures and sleep may be most conspicuous in environments where the chief zeitgeber – sunlight – undergoes extreme but consistent changes over the course of a year. Routine circannual encounters with the midnight sun and midday darkness strategically position Earth’s polar regions as natural laboratories to study this effect. We reviewed 2337 studies that related status epilepticus (SE), light exposure, and sleep in people in the Arctic or Antarctic. Status epilepticus represented the most extreme – and arguably incontrovertible – expression of seizure. After applying filters and exclusion criteria, 5 studies were included for SE and 36 for sleep. Status epilepticus was correlated with increasing daylength going into summer (ie, brighter sunlight vector ) at high latitudes. Sunlight and chronotype demonstrated a linear relationship, with more sunlight going into summer (ie, brighter sunlight vector) advancing chronotype, less sunlight going into winter (ie, darker sunlight vector ) delaying chronotype, and extreme cases becoming free-running in either vector direction. U-shaped relationships were also surmised, with increased sleep fragmentation and decreased sleep duration for both brighter and darker sunlight vectors. A parsimonious synthesis is that a brighter sunlight vector advances chronotype to shorten and fragment sleep, thereby triggering more SE and seizures. Although these data are derived from Earth’s poles, they are not specific to polar regions; rather, they hold ramifications for everyone on planet Earth – 1 in 10 of whom will have a seizure in their lifetime.
- Open Access
- 20 August 2026
- Published by Elsevier in SSRN Electronic Journal
Gabapentin Use During Pregnancy and Adverse Neonatal Outcomes: A Population Based Cohort Study
Background Over the last decade, gabapentin has become one of the most utilized antiseizure medications during pregnancy, but association between gabapentin use during pregnancy and adverse neonatal outcomes among pregnant people with epilepsy (PPWE), and pregnant people without epilepsy (PPWOE) is not clear. Methods We conducted a retrospective population-based cohort study among pregnant people in Manitoba, between 1998 to 2021 to examine the association between gabapentin exposure in-utero and the risk of low birth weight (LBW), preterm birth, NICU admissions, maternal length of hospital stays (LOS), infants LOS, infant mortality, neonatal mortality, neonatal readmissions, neonatal respiratory distress syndrome, and severe neonatal morbidity. We performed crude, adjusted and high dimensional propensity scores (HDPS) analysis using generalised estimating equations with exchangeable matrix. Results Among 1057 pregnant people exposed to gabapentin, we found a significant increased risk of LBW (aOR 1.95, 95% CI 1.58-2.41), preterm birth (aOR 1.68, 95% CI 1.39-2.03), NICU admissions (aOR 2.04, 95% CI 1.71-2.42), pregnant peoples LOS (aOR 1.33, 95% CI 1.15-1.54), infant LOS (aOR 2.09, 95% CI 1.81-2.41) and neonatal respiratory distress syndrome (aOR 3.36, 95% CI 1.16-9.71). In exposed PPWE and PPWOE we found a significant increased risk of infant mortality aOR 5.39 (95% CI 1.11-26.2). The results from HDPS models showed attenuated estimates and non-significant results, suggesting strong unmeasured confounding among adjusted regression models. Conclusion We observed potential risk for specific neonatal outcomes, however unmeasured confounding can not be dismissed. Future studies on the effect of gabapentin during pregnancy are warranted.
- 20 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessGabapentin Use During Pregnancy and Adverse Neonatal Outcomes: A Population Based Cohort Study
Background Over the last decade, gabapentin has become one of the most utilized antiseizure medications during pregnancy, but association between gabapentin use during pregnancy and adverse neonatal outcomes among pregnant people with epilepsy (PPWE), and pregnant people without epilepsy (PPWOE) is not clear. Methods We conducted a retrospective population-based cohort study among pregnant people in Manitoba, between 1998 to 2021 to examine the association between gabapentin exposure in-utero and the risk of low birth weight (LBW), preterm birth, NICU admissions, maternal length of hospital stays (LOS), infants LOS, infant mortality, neonatal mortality, neonatal readmissions, neonatal respiratory distress syndrome, and severe neonatal morbidity. We performed crude, adjusted and high dimensional propensity scores (HDPS) analysis using generalised estimating equations with exchangeable matrix. Results Among 1057 pregnant people exposed to gabapentin, we found a significant increased risk of LBW (aOR 1.95, 95% CI 1.58-2.41), preterm birth (aOR 1.68, 95% CI 1.39-2.03), NICU admissions (aOR 2.04, 95% CI 1.71-2.42), pregnant peoples LOS (aOR 1.33, 95% CI 1.15-1.54), infant LOS (aOR 2.09, 95% CI 1.81-2.41) and neonatal respiratory distress syndrome (aOR 3.36, 95% CI 1.16-9.71). In exposed PPWE and PPWOE we found a significant increased risk of infant mortality aOR 5.39 (95% CI 1.11-26.2). The results from HDPS models showed attenuated estimates and non-significant results, suggesting strong unmeasured confounding among adjusted regression models. Conclusion We observed potential risk for specific neonatal outcomes, however unmeasured confounding can not be dismissed. Future studies on the effect of gabapentin during pregnancy are warranted.
- Open Access
- 8 January 2026
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 13 (6) , 1179-1185
Prognostic Implications of Sleep Architecture for Patients Admitted to the Intensive Care Unit With Status Epilepticus
Objective: Status epilepticus (SE) is associated with significant mortality. Sleep architecture may reflect normal brain function. Impaired sleep architecture is associated with poorer outcomes in numerous conditions. Here we investigate the association of sleep architecture in continuous EEG (cEEG) with survival in SE. Methods: Retrospective single‐centre study of all adult cEEGs in 2015–2019 with electrographic SE. Patients were dichotomized by the presence of sleep architecture (REM or non‐REM) in any epoch per an epileptologist blinded to patient outcome. Categorical variables were assessed using Fisher's exact test. Two‐sample t ‐tests and Wilcoxon rank‐sum tests compared normally and non‐normally distributed variables, respectively. Survival analysis at 6 months used the Mantel–Cox log‐rank test. Multivariate analysis using Firth logistic regression assessed for confounders. Results: Twenty‐five adults (68% male) had electrographic SE (median cEEG duration 2 days, hospitalization 13 days). Mean number of intravenous sedatives was 1.44 and anti‐seizure medication 2.96. Thirteen adults had sleep architecture, with shorter cEEG delays (median 1 vs. 2 days, p = 0.0104) and younger age (median 56 vs. 72.5 years, p = 0.0386). Sleep architecture was associated with better outcomes at 6‐months ( p = 0.0073), higher survival rates at 6‐months (76.9% vs. 25%, p = 0.017), and longer survival durations at 6‐months (160.9 vs. 73.3 days, p = 0.0034). cEEG delay was not associated with survival ( p = 0.107). Significance: In addition to younger age—a known positive SE prognosticator—sleep architecture during cEEG in adults with electrographic SE was associated with increased survival, possibly due to the resilience of younger brains generating sleep architecture in the face of SE, acting as a marker for improved prognosis.
- 8 January 2026
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 13 (6) , 1179-1185
Open AccessPrognostic Implications of Sleep Architecture for Patients Admitted to the Intensive Care Unit With Status Epilepticus
Objective: Status epilepticus (SE) is associated with significant mortality. Sleep architecture may reflect normal brain function. Impaired sleep architecture is associated with poorer outcomes in numerous conditions. Here we investigate the association of sleep architecture in continuous EEG (cEEG) with survival in SE. Methods: Retrospective single‐centre study of all adult cEEGs in 2015–2019 with electrographic SE. Patients were dichotomized by the presence of sleep architecture (REM or non‐REM) in any epoch per an epileptologist blinded to patient outcome. Categorical variables were assessed using Fisher's exact test. Two‐sample t ‐tests and Wilcoxon rank‐sum tests compared normally and non‐normally distributed variables, respectively. Survival analysis at 6 months used the Mantel–Cox log‐rank test. Multivariate analysis using Firth logistic regression assessed for confounders. Results: Twenty‐five adults (68% male) had electrographic SE (median cEEG duration 2 days, hospitalization 13 days). Mean number of intravenous sedatives was 1.44 and anti‐seizure medication 2.96. Thirteen adults had sleep architecture, with shorter cEEG delays (median 1 vs. 2 days, p = 0.0104) and younger age (median 56 vs. 72.5 years, p = 0.0386). Sleep architecture was associated with better outcomes at 6‐months ( p = 0.0073), higher survival rates at 6‐months (76.9% vs. 25%, p = 0.017), and longer survival durations at 6‐months (160.9 vs. 73.3 days, p = 0.0034). cEEG delay was not associated with survival ( p = 0.107). Significance: In addition to younger age—a known positive SE prognosticator—sleep architecture during cEEG in adults with electrographic SE was associated with increased survival, possibly due to the resilience of younger brains generating sleep architecture in the face of SE, acting as a marker for improved prognosis.
- 1 October 2025
- Published by Wolters Kluwer Health in Neurology Clinical Practice
- Vol. 15 (5) , e200522
Continuous EEG Monitoring in Canadian Hospitals
Continuous EEG (cEEG) is the gold standard for diagnosing nonconvulsive seizures (NCSs) and nonconvulsive status epilepticus (NCSE) in critically ill patients, with NCSE occurring in 8%-10% of patients with unexplained coma. Untreated NCSs are associated with secondary brain injury, as well as increased mortality and morbidity. cEEG monitoring allows clinicians to identify more than twice the number of seizures compared with a 30-min routine EEG recording. However, there are limited data on cEEG practices in Canadian hospitals. The aim of this study was to evaluate the availability, indications, and barriers to cEEG access in Canada. A national cross-sectional survey was distributed to EEG laboratory directors and physicians who interpret cEEGs to assess cEEG monitoring practices in Canadian adult hospitals. The survey evaluated institutional cEEG availability, clinical applications, and technical infrastructure. Among 1,267 adult hospitals in Canada, only 92 hospital networks (9%) were identified as having an EEG laboratory. Twenty-four were identified as potentially offering cEEG monitoring, and a survey was sent to a physician at these institutions. Responses were received from 22 institutions (92% response rate), with 19 hospital networks reporting cEEG availability-representing just 2% of Canadian hospitals. Geographic disparities were significant, with 3 provinces and all 3 territories lacking cEEG access. Among tertiary care hospitals, only 68% reported cEEG availability. Barriers included insufficient EEG technologist coverage and prolonged processing periods for 24-hour EEG recordings. Most institutions lacked standardized guidelines, were unable to perform new cEEG hookups after regular work hours, and did not have access to abbreviated montages when cEEG was unavailable. cEEG availability in Canada is highly limited, including at tertiary care centers, with significant geographic inequities and operational barriers. Most Canadian hospitals do not meet guideline standards for cEEG use. These findings highlight the need for systemic changes to improve cEEG access and align Canadian cEEG practices with international standards.
- 1 October 2025
- Published by Wolters Kluwer Health in Neurology Clinical Practice
- Vol. 15 (5) , e200522
Continuous EEG Monitoring in Canadian Hospitals
Continuous EEG (cEEG) is the gold standard for diagnosing nonconvulsive seizures (NCSs) and nonconvulsive status epilepticus (NCSE) in critically ill patients, with NCSE occurring in 8%-10% of patients with unexplained coma. Untreated NCSs are associated with secondary brain injury, as well as increased mortality and morbidity. cEEG monitoring allows clinicians to identify more than twice the number of seizures compared with a 30-min routine EEG recording. However, there are limited data on cEEG practices in Canadian hospitals. The aim of this study was to evaluate the availability, indications, and barriers to cEEG access in Canada. A national cross-sectional survey was distributed to EEG laboratory directors and physicians who interpret cEEGs to assess cEEG monitoring practices in Canadian adult hospitals. The survey evaluated institutional cEEG availability, clinical applications, and technical infrastructure. Among 1,267 adult hospitals in Canada, only 92 hospital networks (9%) were identified as having an EEG laboratory. Twenty-four were identified as potentially offering cEEG monitoring, and a survey was sent to a physician at these institutions. Responses were received from 22 institutions (92% response rate), with 19 hospital networks reporting cEEG availability-representing just 2% of Canadian hospitals. Geographic disparities were significant, with 3 provinces and all 3 territories lacking cEEG access. Among tertiary care hospitals, only 68% reported cEEG availability. Barriers included insufficient EEG technologist coverage and prolonged processing periods for 24-hour EEG recordings. Most institutions lacked standardized guidelines, were unable to perform new cEEG hookups after regular work hours, and did not have access to abbreviated montages when cEEG was unavailable. cEEG availability in Canada is highly limited, including at tertiary care centers, with significant geographic inequities and operational barriers. Most Canadian hospitals do not meet guideline standards for cEEG use. These findings highlight the need for systemic changes to improve cEEG access and align Canadian cEEG practices with international standards.
- Open Access
- 30 July 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e67446
Machine Learning–Based Cognitive Assessment With The Autonomous Cognitive Examination: Randomized Controlled Trial
Background: The rising prevalence of dementia necessitates a scalable solution to cognitive assessments. The Autonomous Cognitive Examination (ACoE) is a foundational cognitive test for the phenotyping of cognitive symptoms across the primary cognitive domains. However, while the ACoE has been internally validated, it has not been externally validated in a clinical population, and its ability to render accurate appraisals of cognition is unknown. Further, it is unclear if these phenotypic assessments are useful in clinical tasks such as screening patients with and those without impairments. Objective: The objective of this study is to validate the ability of the ACoE to reliably phenotype cognition and to act as a screening examination relative to standard paper-based tests. Methods: To compare the evaluations of the ACoE to established paper-based tests, 46 patients with neurological disorders were enrolled in a randomized crossover study and received either the ACoE or a standard paper-based cognitive test. Patients received either the Addenbrooke Cognitive Examination-3 (ACE-3; n=35) or the Montreal Cognitive Examination (MoCA; n=11). We evaluated 3 primary metrics of the ACoE’s performance relative to paper-based tests: (1) interrater reliability of overall cognitive scores, (2) interrater reliability of cognitive domain scores, and (3) ability to classify patients similarly to paper-based tests. Results: The ACoE’s overall cognitive assessments were significantly reliable (ICC [intraclass correlation coefficient]=0.89; P<.001). Each cognitive domain’s assessments were also significantly reliable, including attention (ICC=0.74; PFWE<.001), language (ICC=0.89; PFWE<.001), memory (ICC=0.91; PFWE<.001), fluency (ICC=0.74; PFWE<.001), and visuospatial function (ICC=0.78; PFWE<.001). The ACoE was also able to successfully diagnose patients similarly to both paper-based tests (area under the receiver operating characteristic curve=0.96; PFWE<.001). Conclusions: In this study, we evaluated if the ACoE could reliably phenotype cognitive symptoms relative to the assessments of established standard paper-based cognitive assessments. We found that the ACoE reliably phenotypes patient cognition, which can be used to screen patients. In the future, these cognitive phenotypes may be used to diagnose specific etiologies.
- 30 July 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e67446
Open AccessMachine Learning–Based Cognitive Assessment With The Autonomous Cognitive Examination: Randomized Controlled Trial
Background: The rising prevalence of dementia necessitates a scalable solution to cognitive assessments. The Autonomous Cognitive Examination (ACoE) is a foundational cognitive test for the phenotyping of cognitive symptoms across the primary cognitive domains. However, while the ACoE has been internally validated, it has not been externally validated in a clinical population, and its ability to render accurate appraisals of cognition is unknown. Further, it is unclear if these phenotypic assessments are useful in clinical tasks such as screening patients with and those without impairments. Objective: The objective of this study is to validate the ability of the ACoE to reliably phenotype cognition and to act as a screening examination relative to standard paper-based tests. Methods: To compare the evaluations of the ACoE to established paper-based tests, 46 patients with neurological disorders were enrolled in a randomized crossover study and received either the ACoE or a standard paper-based cognitive test. Patients received either the Addenbrooke Cognitive Examination-3 (ACE-3; n=35) or the Montreal Cognitive Examination (MoCA; n=11). We evaluated 3 primary metrics of the ACoE’s performance relative to paper-based tests: (1) interrater reliability of overall cognitive scores, (2) interrater reliability of cognitive domain scores, and (3) ability to classify patients similarly to paper-based tests. Results: The ACoE’s overall cognitive assessments were significantly reliable (ICC [intraclass correlation coefficient]=0.89; P<.001). Each cognitive domain’s assessments were also significantly reliable, including attention (ICC=0.74; PFWE<.001), language (ICC=0.89; PFWE<.001), memory (ICC=0.91; PFWE<.001), fluency (ICC=0.74; PFWE<.001), and visuospatial function (ICC=0.78; PFWE<.001). The ACoE was also able to successfully diagnose patients similarly to both paper-based tests (area under the receiver operating characteristic curve=0.96; PFWE<.001). Conclusions: In this study, we evaluated if the ACoE could reliably phenotype cognitive symptoms relative to the assessments of established standard paper-based cognitive assessments. We found that the ACoE reliably phenotypes patient cognition, which can be used to screen patients. In the future, these cognitive phenotypes may be used to diagnose specific etiologies.
- 26 June 2025
- Published by Massachusetts Medical Society in NEJM AI
- Vol. 2 (7)
Expert-Level Detection of Epilepsy Markers in EEG on Short and Long Timescales
Epileptiform discharges, or spikes, within electroencephalogram (EEG) recordings are essential for diagnosing epilepsy and localizing seizure origins. Artificial intelligence (AI) offers a promising approach to automating detection, but current models are often hindered by artifact-related false positives and often target either event- or EEG-level classification, thus limiting clinical utility. We developed SpikeNet2, a deep-learning model based on a residual network architecture, and enhanced it with hard-negative mining to reduce false positives. Our study analyzed 17,812 EEG recordings from 13,523 patients across multiple institutions, including Massachusetts General Brigham (MGB) hospitals. Data from the Human Epilepsy Project (HEP) and SCORE-AI (SAI) were also included. A total of 32,433 event-level samples, labeled by experts, were used for training and evaluation. Performance was assessed using the area under the receiver operating characteristic curve (AUROC), the area under the precision–recall curve (AUPRC), calibration error, and a modified area under the curve (mAUC) metric. The model’s generalizability was evaluated using external datasets. SpikeNet2 demonstrated strong performance in event-level spike detection, achieving an AUROC of 0.973 and an AUPRC of 0.995, with 44% of experts surpassing the model on the MGB dataset. In external validation, the model achieved an AUROC of 0.942 and an AUPRC of 0.948 on the HEP dataset. For EEG-level classification, SpikeNet2 recorded an AUROC of 0.958 and an AUPRC of 0.959 on the MGB dataset, an AUROC of 0.888 and an AUPRC of 0.823 on the HEP dataset, and an AUROC of 0.995 and an AUPRC of 0.991 on the SAI dataset, with 32% of experts outperforming the model. The false-positive rate was reduced to an average of nine spikes per hour. SpikeNet2 offers expert-level accuracy in both event-level spike detection and EEG-level classification, while significantly reducing false positives. Its dual functionality and robust performance across diverse datasets make it a promising tool for clinical and telemedicine applications, particularly in resource-limited settings. (Funded by the National Institutes of Health and others.)
- 26 June 2025
- Published by Massachusetts Medical Society in NEJM AI
- Vol. 2 (7)
Expert-Level Detection of Epilepsy Markers in EEG on Short and Long Timescales
Epileptiform discharges, or spikes, within electroencephalogram (EEG) recordings are essential for diagnosing epilepsy and localizing seizure origins. Artificial intelligence (AI) offers a promising approach to automating detection, but current models are often hindered by artifact-related false positives and often target either event- or EEG-level classification, thus limiting clinical utility. We developed SpikeNet2, a deep-learning model based on a residual network architecture, and enhanced it with hard-negative mining to reduce false positives. Our study analyzed 17,812 EEG recordings from 13,523 patients across multiple institutions, including Massachusetts General Brigham (MGB) hospitals. Data from the Human Epilepsy Project (HEP) and SCORE-AI (SAI) were also included. A total of 32,433 event-level samples, labeled by experts, were used for training and evaluation. Performance was assessed using the area under the receiver operating characteristic curve (AUROC), the area under the precision–recall curve (AUPRC), calibration error, and a modified area under the curve (mAUC) metric. The model’s generalizability was evaluated using external datasets. SpikeNet2 demonstrated strong performance in event-level spike detection, achieving an AUROC of 0.973 and an AUPRC of 0.995, with 44% of experts surpassing the model on the MGB dataset. In external validation, the model achieved an AUROC of 0.942 and an AUPRC of 0.948 on the HEP dataset. For EEG-level classification, SpikeNet2 recorded an AUROC of 0.958 and an AUPRC of 0.959 on the MGB dataset, an AUROC of 0.888 and an AUPRC of 0.823 on the HEP dataset, and an AUROC of 0.995 and an AUPRC of 0.991 on the SAI dataset, with 32% of experts outperforming the model. The false-positive rate was reduced to an average of nine spikes per hour. SpikeNet2 offers expert-level accuracy in both event-level spike detection and EEG-level classification, while significantly reducing false positives. Its dual functionality and robust performance across diverse datasets make it a promising tool for clinical and telemedicine applications, particularly in resource-limited settings. (Funded by the National Institutes of Health and others.)
- Open Access
- 19 June 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e66735
The Rapid Online Cognitive Assessment for the Detection of Neurocognitive Disorder: Open-Label Study
Background: The rising prevalence of dementia necessitates a scalable solution to cognitive screening. Paper-based cognitive screening examinations are well-validated but minimally scalable. If a digital cognitive screening examination could replicate paper-based screening, it may improve scalability while potentially maintaining the performance of these well-validated paper-based tests. Here, we evaluate the Rapid Online Cognitive Assessment (RoCA), a remote and self-administered digital cognitive screening examination. Objective: The objective of this study was to validate the ability of RoCA to reliably evaluate patient input, identify patients with cognitive impairment relative to the established tests, and evaluate its potential as a screening tool. Methods: RoCA uses a convolutional neural network to evaluate a patient’s ability to perform common cognitive screening tasks: wireframe diagram copying and clock drawing tests. To evaluate RoCA, we compared its evaluations with those of established paper-based tests. This open-label study consists of 46 patients (age range 33-82 years) who were enrolled from neurology clinics. Patients completed the RoCA screening examination and either Addenbrooke’s Cognitive Examination-3 (ACE-3, n=35) or Montreal Cognitive Assessment (MoCA, n=11). We evaluated 3 primary metrics of RoCA’s performance: (1) ability to correctly evaluate patient inputs, (2) ability to identify patients with cognitive impairment compared to ACE-3 and MoCA, and (3) performance as a screening tool. Results: RoCA classifies patients similarly to gold standard paper-based tests, with a receiver operating characteristic area under the curve of 0.81 (95% CI 0.67‐0.91; P<.001). RoCA achieved sensitivity of 0.94 (95% CI 0.80‐1.0; P<.001). This was robust to multiple control analyses. Approximately 83% (16/19) of the patient respondents reported RoCA as highly intuitive, with 95% (18/19) perceiving it as adding value to their care. Conclusions: RoCA may act as a simple and highly scalable digital cognitive screening examination. However, due to the limitations of this study, further work is required to evaluate the ability of RoCA to be generalizable across patient populations, assess its performance in an entirely remote manner, and analyze the effect of digital literacy.
- 19 June 2025
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 27 (1) , e66735
Open AccessThe Rapid Online Cognitive Assessment for the Detection of Neurocognitive Disorder: Open-Label Study
Background: The rising prevalence of dementia necessitates a scalable solution to cognitive screening. Paper-based cognitive screening examinations are well-validated but minimally scalable. If a digital cognitive screening examination could replicate paper-based screening, it may improve scalability while potentially maintaining the performance of these well-validated paper-based tests. Here, we evaluate the Rapid Online Cognitive Assessment (RoCA), a remote and self-administered digital cognitive screening examination. Objective: The objective of this study was to validate the ability of RoCA to reliably evaluate patient input, identify patients with cognitive impairment relative to the established tests, and evaluate its potential as a screening tool. Methods: RoCA uses a convolutional neural network to evaluate a patient’s ability to perform common cognitive screening tasks: wireframe diagram copying and clock drawing tests. To evaluate RoCA, we compared its evaluations with those of established paper-based tests. This open-label study consists of 46 patients (age range 33-82 years) who were enrolled from neurology clinics. Patients completed the RoCA screening examination and either Addenbrooke’s Cognitive Examination-3 (ACE-3, n=35) or Montreal Cognitive Assessment (MoCA, n=11). We evaluated 3 primary metrics of RoCA’s performance: (1) ability to correctly evaluate patient inputs, (2) ability to identify patients with cognitive impairment compared to ACE-3 and MoCA, and (3) performance as a screening tool. Results: RoCA classifies patients similarly to gold standard paper-based tests, with a receiver operating characteristic area under the curve of 0.81 (95% CI 0.67‐0.91; P<.001). RoCA achieved sensitivity of 0.94 (95% CI 0.80‐1.0; P<.001). This was robust to multiple control analyses. Approximately 83% (16/19) of the patient respondents reported RoCA as highly intuitive, with 95% (18/19) perceiving it as adding value to their care. Conclusions: RoCA may act as a simple and highly scalable digital cognitive screening examination. However, due to the limitations of this study, further work is required to evaluate the ability of RoCA to be generalizable across patient populations, assess its performance in an entirely remote manner, and analyze the effect of digital literacy.
- Open Access
- 25 June 2024
- Published by Elsevier in Epilepsy & Behavior
- Vol. 158, 109901
The challenges of treating status epilepticus in rural Canada
- 25 June 2024
- Published by Elsevier in Epilepsy & Behavior
- Vol. 158, 109901
Open AccessThe challenges of treating status epilepticus in rural Canada
- Open Access
- 19 September 2023
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 10 (11) , 2166-2170
Circannual seizure provocation as the day lengthens in the northern and southern hemispheres
Circannual status epilepticus (SE) patterns in communities near Earth's poles best test the hypothesis that SE susceptibility varies with light exposure because these communities are routinely subject to large changes in annual light exposure, which may result in changes to daily sleep time. We compared northern hemispheric circannual SE occurrence in Kivalliq, Canada (latitude‐62.8° N) to southern hemispheric Auckland, New Zealand (latitude‐36.9° S). Instead of peaking at a similar calendar time, SE peaked at a similar solar time during the increasing daylight phase after each region's respective winter solstice. This demonstrates that cumulative effects of increasing light exposure can mediate SE susceptibility.
- 19 September 2023
- Published by Wiley in Annals of Clinical and Translational Neurology
- Vol. 10 (11) , 2166-2170
Open AccessCircannual seizure provocation as the day lengthens in the northern and southern hemispheres
Circannual status epilepticus (SE) patterns in communities near Earth's poles best test the hypothesis that SE susceptibility varies with light exposure because these communities are routinely subject to large changes in annual light exposure, which may result in changes to daily sleep time. We compared northern hemispheric circannual SE occurrence in Kivalliq, Canada (latitude‐62.8° N) to southern hemispheric Auckland, New Zealand (latitude‐36.9° S). Instead of peaking at a similar calendar time, SE peaked at a similar solar time during the increasing daylight phase after each region's respective winter solstice. This demonstrates that cumulative effects of increasing light exposure can mediate SE susceptibility.