Machine learning identifies “rsfMRI epilepsy networks” in temporal lobe epilepsy

Authors

Rose Dawn Bharath, National Institute of Mental Health and Neuro Sciences
Rajanikant Panda, National Institute of Mental Health and Neuro Sciences
Jeetu Raj, Indian Institute of Technology Delhi
Sujas Bhardwaj, National Institute of Mental Health and Neuro Sciences
Sanjib Sinha, National Institute of Mental Health and Neuro Sciences
Ganne Chaitanya, National Institute of Mental Health and Neuro Sciences
Kenchaiah Raghavendra, National Institute of Mental Health and Neuro Sciences
Ravindranadh C. Mundlamuri, National Institute of Mental Health and Neuro Sciences
Arivazhagan Arimappamagan, National Institute of Mental Health and Neuro Sciences
Malla Bhaskara Rao, National Institute of Mental Health and Neuro Sciences
Jamuna Rajeshwaran, National Institute of Mental Health and Neuro Sciences
Kandavel Thennarasu, National Institute of Mental Health and Neuro Sciences
Kaushik K. Majumdar, Indian Statistical Institute Bangalore
Parthasarthy Satishchandra, National Institute of Mental Health and Neuro Sciences
Tapan K. Gandhi, Indian Institute of Technology Delhi

Article Type

Research Article

Publication Title

European Radiology

Abstract

Objectives: Experimental models have provided compelling evidence for the existence of neural networks in temporal lobe epilepsy (TLE). To identify and validate the possible existence of resting-state “epilepsy networks,” we used machine learning methods on resting-state functional magnetic resonance imaging (rsfMRI) data from 42 individuals with TLE. Methods: Probabilistic independent component analysis (PICA) was applied to rsfMRI data from 132 subjects (42 TLE patients + 90 healthy controls) and 88 independent components (ICs) were obtained following standard procedures. Elastic net-selected features were used as inputs to support vector machine (SVM). The strengths of the top 10 networks were correlated with clinical features to obtain “rsfMRI epilepsy networks.” Results: SVM could classify individuals with epilepsy with 97.5% accuracy (sensitivity = 100%, specificity = 94.4%). Ten networks with the highest ranking were found in the frontal, perisylvian, cingulo-insular, posterior-quadrant, thalamic, cerebello-thalamic, and temporo-thalamic regions. The posterior-quadrant, cerebello-thalamic, thalamic, medial-visual, and perisylvian networks revealed significant correlation (r > 0.40) with age at onset of seizures, the frequency of seizures, duration of illness, and a number of anti-epileptic drugs. Conclusions: IC-derived rsfMRI networks contain epilepsy-related networks and machine learning methods are useful in identifying these networks in vivo. Increased network strength with disease progression in these “rsfMRI epilepsy networks” could reflect epileptogenesis in TLE. Key Points: • ICA of resting-state fMRI carries disease-specific information about epilepsy. • Machine learning can classify these components with 97.5% accuracy. • “Subject-specific epilepsy networks” could quantify “epileptogenesis” in vivo.

First Page

3496

Last Page

3505

DOI

10.1007/s00330-019-5997-2

Publication Date

7-1-2019

Comments

Open Access, Green

Recommended Citation

Bharath, Rose Dawn; Panda, Rajanikant; Raj, Jeetu; Bhardwaj, Sujas; Sinha, Sanjib; Chaitanya, Ganne; Raghavendra, Kenchaiah; Mundlamuri, Ravindranadh C.; Arimappamagan, Arivazhagan; Rao, Malla Bhaskara; Rajeshwaran, Jamuna; Thennarasu, Kandavel; Majumdar, Kaushik K.; Satishchandra, Parthasarthy; and Gandhi, Tapan K., "Machine learning identifies “rsfMRI epilepsy networks” in temporal lobe epilepsy" (2019). Journal Articles. 794.
https://digitalcommons.isical.ac.in/journal-articles/794