Dataset Information

Development and Validation of MRI-Based Radiomics Models for Diagnosing Juvenile Myoclonic Epilepsy.

ABSTRACT:

Objective

Radiomic modeling using multiple regions of interest in MRI of the brain to diagnose juvenile myoclonic epilepsy (JME) has not yet been investigated. This study aimed to develop and validate radiomics prediction models to distinguish patients with JME from healthy controls (HCs), and to evaluate the feasibility of a radiomics approach using MRI for diagnosing JME.

Materials and methods

A total of 97 JME patients (25.6 ± 8.5 years; female, 45.5%) and 32 HCs (28.9 ± 11.4 years; female, 50.0%) were randomly split (7:3 ratio) into a training (n = 90) and a test set (n = 39) group. Radiomic features were extracted from 22 regions of interest in the brain using the T1-weighted MRI based on clinical evidence. Predictive models were trained using seven modeling methods, including a light gradient boosting machine, support vector classifier, random forest, logistic regression, extreme gradient boosting, gradient boosting machine, and decision tree, with radiomics features in the training set. The performance of the models was validated and compared to the test set. The model with the highest area under the receiver operating curve (AUROC) was chosen, and important features in the model were identified.

Results

The seven tested radiomics models, including light gradient boosting machine, support vector classifier, random forest, logistic regression, extreme gradient boosting, gradient boosting machine, and decision tree, showed AUROC values of 0.817, 0.807, 0.783, 0.779, 0.767, 0.762, and 0.672, respectively. The light gradient boosting machine with the highest AUROC, albeit without statistically significant differences from the other models in pairwise comparisons, had accuracy, precision, recall, and F1 scores of 0.795, 0.818, 0.931, and 0.871, respectively. Radiomic features, including the putamen and ventral diencephalon, were ranked as the most important for suggesting JME.

Conclusion

Radiomic models using MRI were able to differentiate JME from HCs.

SUBMITTER: Kim KM

PROVIDER: S-EPMC9747272 | biostudies-literature | 2022 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development and Validation of MRI-Based Radiomics Models for Diagnosing Juvenile Myoclonic Epilepsy.

Kim Kyung Min KM Hwang Heewon H Sohn Beomseok B Park Kisung K Han Kyunghwa K Ahn Sung Soo SS Lee Wonwoo W Chu Min Kyung MK Heo Kyoung K Lee Seung-Koo SK

Korean journal of radiology 20221201 12

<h4>Objective</h4>Radiomic modeling using multiple regions of interest in MRI of the brain to diagnose juvenile myoclonic epilepsy (JME) has not yet been investigated. This study aimed to develop and validate radiomics prediction models to distinguish patients with JME from healthy controls (HCs), and to evaluate the feasibility of a radiomics approach using MRI for diagnosing JME.<h4>Materials and methods</h4>A total of 97 JME patients (25.6 ± 8.5 years; female, 45.5%) and 32 HCs (28.9 ± 11.4 y ...[more]

PMID: 36447416

Similar Datasets

Project description:ObjectiveAccording to Panksepp's hierarchical emotion model, emotion processing relies on three functionally and neuroanatomically distinct levels. These levels comprise subcortical networks (primary level), the limbic system (secondary level), and the neocortex (tertiary level) and are suggested to serve differential emotional processing. We aimed to validate and extend previous evidence of discrete and dimensional emotion processing in patient with juvenile myoclonic epilepsy (JME).MethodsWe recorded brain activity of patients with JME and healthy controls in response to lexical decisions to words reflecting the discrete emotion fear and the affective dimension negativity previously suggested to rely on different brain regions and to reflect different levels of processing. In all study participants, we tested verbal cognitive functions, as well as the relationship of psychiatric conditions, seizure types and duration of epilepsy and emotional word processing.ResultsIn support of the hierarchical emotion model, we found an interaction of discrete emotion and affective dimensional processing in the right amygdala likely to reflect secondary level processing. Brain activity related to affective dimensional processing was found in the right inferior frontal gyrus and is suggested to reflect tertiary level processing. Psychiatric conditions, type of seizure nor mono- vs. polytherapy and duration of epilepsy within patients did not have any effect on the processing of emotional words. In addition, no differences in brain activity or response times between patients and controls were observed, despite neuropsychological testing revealed slightly decreased verbal intelligence, verbal fluency and reading speed in patients with JME.SignificanceThese results were interpreted to be in line with the hierarchical emotion model and to highlight the amygdala's role in processing biologically relevant stimuli, as well as to suggest a semantic foundation of affective dimensional processing in prefrontal cortex. A lack of differences in brain activity of patients with JME and healthy controls in response to the emotional content of words could point to unaffected implicit emotion processing in patients with JME.

Project description:ObjectiveTo study the effectiveness and tolerability of antiepileptic drugs (AEDs) commonly used in juvenile myoclonic epilepsy (JME).MethodsPeople with JME were identified from a large database of individuals with epilepsy, which includes detailed retrospective information on AED use. We assessed secular changes in AED use and calculated rates of response (12-month seizure freedom) and adverse drug reactions (ADRs) for the five most common AEDs. Retention was modeled with a Cox proportional hazards model. We compared valproate use between males and females.ResultsWe included 305 people with 688 AED trials of valproate, lamotrigine, levetiracetam, carbamazepine, and topiramate. Valproate and carbamazepine were most often prescribed as the first AED. The response rate to valproate was highest among the five AEDs (42.7%), and significantly higher than response rates for lamotrigine, carbamazepine, and topiramate; the difference to the response rate to levetiracetam (37.1%) was not significant. The rates of ADRs were highest for topiramate (45.5%) and valproate (37.5%). Commonest ADRs included weight change, lethargy, and tremor. In the Cox proportional hazards model, later start year (1.10 [1.08-1.13], P < 0.001) and female sex (1.41 [1.07-1.85], P = 0.02) were associated with shorter trial duration. Valproate was associated with the longest treatment duration; trials with carbamazepine and topiramate were significantly shorter (HR [CI]: 3.29 [2.15-5.02], P < 0.001 and 1.93 [1.31-2.86], P < 0.001). The relative frequency of valproate trials shows a decreasing trend since 2003 while there is an increasing trend for levetiracetam. Fewer females than males received valproate (76.2% vs 92.6%, P = 0.001).SignificanceIn people with JME, valproate is an effective AED; levetiracetam emerged as an alternative. Valproate is now contraindicated in women of childbearing potential without special precautions. With appropriate selection and safeguards in place, valproate should remain available as a therapy, including as an alternative for women of childbearing potential whose seizures are resistant to other treatments.

Dataset Information

Development and Validation of MRI-Based Radiomics Models for Diagnosing Juvenile Myoclonic Epilepsy.

Objective

Materials and methods

Results

Conclusion

Publications

Development and Validation of MRI-Based Radiomics Models for Diagnosing Juvenile Myoclonic Epilepsy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets