Diffusion Tractography Biomarker for Epilepsy Severity in Children With Drug-Resistant Epilepsy
Jeong-Won Jeong1,2,3,4, Min-Hee Lee1,2, Hiroshi Uda1
1Department of Pediatrics, Wayne State University, Detroit, Michigan, USA.
Insights
A novel deep learning model accurately predicts epilepsy severity in pediatric drug-resistant epilepsy (DRE) patients. This tool aids in identifying neurocognitive impairments, enabling timely interventions for better outcomes.
Area of Science:
- Neuroscience
- Artificial Intelligence
- Medical Imaging
Background:
- Pediatric drug-resistant epilepsy (DRE) often co-occurs with neurocognitive impairments.
- Accurate assessment of epilepsy severity is crucial for predicting these impairments.
Purpose of the Study:
- To develop a deep learning model using diffusion-weighted imaging (DWI) tractography to predict epilepsy severity.
- To evaluate the model's ability to screen for neurocognitive impairments in pediatric DRE.
Main Methods:
- Constructed an epilepsy severity network (ESN) correlated with clinical GASE scores in DRE children.
- Utilized a dilated deep convolutional neural network with a relational network (dilated DCNN+RN) to generate a predicted GASE score biomarker.
- Validated the model on separate development and independent test sets, including a random score learning experiment to assess overfitting.
Main Results:
- The dilated DCNN+RN model achieved high correlation with clinical GASE scores (r=0.92 development, r=0.83 test).
- The model demonstrated minimal overfitting, outperforming state-of-the-art methods.
- Predicted GASE scores showed superior model fit and discriminatory ability compared to clinical GASE scores.
Conclusions:
- The developed deep learning biomarker shows significant potential for early identification of neurocognitive impairment risk in pediatric DRE.
- This tool can facilitate timely, personalized interventions to mitigate long-term effects of DRE.
Objective:
To develop a novel deep-learning model of clinical DWI tractography that can accurately predict the general assessment of epilepsy severity (GASE) in pediatric drug-resistant epilepsy (DRE) and test if it can screen diverse neurocognitive impairments identified through neuropsychological assessments.
Methods:
DRE children and age-sex-matched healthy controls were enrolled to construct an epilepsy severity network (ESN), whose edges were significantly correlated with GASE scores of DRE children. An ESN-based biomarker called the predicted GASE score was obtained using dilated deep convolutional neural network with a relational network (dilated DCNN+RN) and used to quantify the risk of neurocognitive impairments using global/verbal/non-verbal neuropsychological assessments of 36/37/32 children performed on average 3.2 ± 2.7 months prior to the MRI scan. To warrant the generalizability, the proposed biomarker was trained and evaluated using separate development and independent test sets, with the random score learning experiment included to assess potential overfitting.
Results:
The dilated DCNN+RN outperformed other state-of-the art methods to create the predicted GASE scores with significant correlation (r = 0.92 and 0.83 for development and test sets with clinical GASE scores) and minimal overfitting (r = -0.25 and 0.00 for development and test sets with random GASE scores). Both univariate and multivariate models demonstrated that compared with the clinical GASE scores, the predicted GASE scores provide better model fit and discriminatory ability, suggesting more adjusted and accurate estimate of epilepsy severity contributing to the overall risk.
Interpretation:
The proposed biomarker shows strong potential for early identification of DRE children at risk of neurocognitive impairments, enabling timely, personalized interventions to prevent long-term effects.
