Multicenter Validation of a Deep Learning Detection Algorithm for Focal Cortical Dysplasia

Abstract

Objective.To test the hypothesis that a multicenter-validated computer deep learning algorithm detects MRI-negative focal cortical dysplasia (FCD).Methods.We used clinically-acquired 3D T1-weighted and 3D FLAIR MRI of 148 patients (median age, 23 years [range, 2-55]; 47% female) with histologically-verified FCD at nine centers to train a deep convolutional neural network (CNN) classifier. Images were initially deemed as MRI-negative in 51% of cases, in whom intracranial EEG determined the focus. For risk stratification, the CNN incorporated Bayesian uncertainty estimation as a measure of confidence. To evaluate performance, detection maps were compared to expert FCD manual labels. Sensitivity was tested in an independent cohort of 23 FCD cases (13±10 years). Applying the algorithm to 42 healthy and 89 temporal lobe epilepsy disease controls tested specificity.Results.Overall sensitivity was 93% (137/148 FCD detected) using a leave-one-site-out cross-validation, with an average of six false positives per patient. Sensitivity in MRI-negative FCD was 85%. In 73% of patients, the FCD was among the clusters with the highest confidence; in half it ranked the highest. Sensitivity in the independent cohort was 83% (19/23; average of five false positives per patient). Specificity was 89% in healthy and disease controls.Conclusions.This first multicenter-validated deep learning detection algorithm yields the highest sensitivity to date in MRI-negative FCD. By pairing predictions with risk stratification this classifier may assist clinicians to adjust hypotheses relative to other tests, increasing diagnostic confidence. Moreover, generalizability across age and MRI hardware makes this approach ideal for pre-surgical evaluation of MRI-negative epilepsy.Classification of evidence.This study provides Class III evidence that deep learning on multimodal MRI accurately identifies FCD in epilepsy patients initially diagnosed as MRI-negative.

Interpretable surface-based detection of focal cortical dysplasias: a MELD study

“…However, being limited by small numbers of patients and data acquired from only one or two MRI scanners can lead to large error bars on estimates of sensitivity and specificity (Varoquaux, 2018) and limited generalisability due lack of diversity in training data. Progress is also being made on automated volumetric MRI methods (David et al, 2021;Gill et al, 2021;House et al, 2021), however datasets have tended to be restricted either to histopathologically confirmed FCD type II cohorts, or single sites. Through creating a large dataset including both paediatric and adult patients across multiple sites and MRI scanners as well as including all FCD histopathological subtypes, we aimed to address these limitations.…”

“…Radiologically FCDs are characterised by alterations in cortical thickness, blurring at the greywhite matter boundary, folding abnormalities, and T2 or FLAIR signal intensity changes (Colombo et al, 2012). Approaches to improving the detection of FCDs have involved improved scanner protocols (Bernasconi et al, 2019) and field strengths (Bartolini et al, 2019;Wang et al, 2020) as well as automated volumetric (David et al, 2021;Gill et al, 2021Gill et al, , 2018House et al, 2021;Huppertz et al, 2005) and surface-based (Adler et al, 2017;Ahmed et al, 2015;Hong et al, 2014;Jin et al, 2018) post-processing methods.…”

Artificial Intelligence Applications in the Imaging of Epilepsy and Its Comorbidities: Present and Future

“… 22 – 25 A recent multicenter-validated study showed that DL using multimodal MRI data could reliably identify previous MRI-negative FCD lesions, suggesting that DL shows promise for assisting non-expert clinicians in this challenging diagnosis. 26 …”

“…[22][23][24][25] A recent multicentervalidated study showed that DL using multimodal MRI data could reliably identify previous MRI-negative FCD lesions, suggesting that DL shows promise for assisting non-expert clinicians in this challenging diagnosis. 26 AI methods can also combine imaging and clinical data to build models for predicting clinical outcomes in patients with epilepsy. 27 For example, automated volumetric MRI measurements incorporated into statistical models help to predict postoperative seizure outcomes in TLE and frontal lobe epilepsy (FLE), revealing that subtle cortical atrophy beyond the surgical resection influences seizure outcome.…”

Multicenter Validation of a Deep Learning Detection Algorithm for Focal Cortical Dysplasia

“…In the Research Article “Multicenter Validation of a Deep Learning Detection Algorithm for Focal Cortical Dysplasia” by Gill et al, 1 several links to the supplemental content were omitted in the final manuscript. The article has now been replaced by a corrected version with the links included.…”