Same same but different: A Web‐based deep learning application revealed classifying features for the histopathologic distinction of cortical malformations

Abstract

Objective The microscopic review of hematoxylin‐eosin–stained images of focal cortical dysplasia type IIb and cortical tuber of tuberous sclerosis complex remains challenging. Both entities are distinct subtypes of human malformations of cortical development that share histopathological features consisting of neuronal dyslamination with dysmorphic neurons and balloon cells. We trained a convolutional neural network (CNN) to classify both entities and visualize the results. Additionally, we propose a new Web‐based deep learning application as proof of concept of how deep learning could enter the pathologic routine. Methods A digital processing pipeline was developed for a series of 56 cases of focal cortical dysplasia type IIb and cortical tuber of tuberous sclerosis complex to obtain 4000 regions of interest and 200 000 subsamples with different zoom and rotation angles to train a neural network. Guided gradient‐weighted class activation maps (Guided Grad‐CAMs) were generated to visualize morphological features used by the CNN to distinguish both entities. Results Our best‐performing network achieved 91% accuracy and 0.88 area under the receiver operating characteristic curve at the tile level for an unseen test set. Novel histopathologic patterns were found through the visualized Guided Grad‐CAMs. These patterns were assembled into a classification score to augment decision‐making in routine histopathology workup. This score was successfully validated by 11 expert neuropathologists and 12 nonexperts, boosting nonexperts to expert level performance. Significance Our newly developed Web application combines the visualization of whole slide images with the possibility of deep learning–aided classification between focal cortical dysplasia IIb and tuberous sclerosis complex. This approach will help to introduce deep learning applications and visualization for the histopathologic diagnosis of rare and difficult‐to‐classify brain lesions.

A Whole Slide Image Managing Library Based on Fastai for Deep Learning in the Context of Histopathology: Two Use-Cases Explained

“…We believe that these rather big networks with lots of parameters worked well, because of their big input image size of 512x512 pixels. On smaller images networks with less parameters tend to work better in our experience [3] . A crucial step in our pipeline was among sufficient training data the way of image preprocessing.…”

“…Medical and non-medical imageclassification tasks have been remarkably successful utilizing DL. Successful examples range from utilization of different types of cancer detection/classification/grading [1,2] , Focal Cortical Dysplasia in human focal epilepsies [3], classification of liver cirrhosis, heart failure detection, and classification of Alzheimer plaques [4] . Disease grading, prognosis prediction and imaging biomarkers for genetic subtype identification are more challenging tasks but have also been successfully established [5,6].…”

Isolated cortical tuber in an infant with genetically confirmed tuberous sclerosis complex 1 presenting with symptomatic West syndrome

“…Calcifications on neuroimaging are more common in cortical tubers, reported in as many as 50% of cases, 16 whereas they are rarely present in FCDs 17,18 . A recent study utilizing deep learning demonstrates 91% accuracy in distinguishing HE‐stained images of FCD type IIb from genetically confirmed cortical tubers achieved by a convolutional neural network‐based model, which was much better than 72.3% baseline accuracy among 11 experienced neuropathologists 19 . In addition, novel‐distinguished histologic features were extracted from guided gradient‐weighted class activation maps (Guided Grad‐CAMs), 19 including: (i) matrix reaction (fibrillar and strand‐like in cortical tuber vs diffuse and granular in FCD type IIb); (ii) astrocytes (larger nuclei with uncondensed chromatin structure in cortical tuber vs smaller nuclei with more condensed chromatin in FCD type IIb); and (iii) ballooned/giant cells (halo‐like artifact around giant cells in cortical tuber [Fig.…”

A Whole Slide Image Managing Library Based on Fastai for Deep Learning in the Context of Histopathology: Two Use-Cases Explained

“…Medical and non-medical imageclassification tasks have been remarkably successful utilizing DL. Successful examples range from utilization of different types of cancer detection/classification/grading [1,2] , Focal Cortical Dysplasia in human focal epilepsies [3], classification of liver cirrhosis, heart failure detection, and classification of Alzheimer plaques [4] . Disease grading, prognosis prediction and imaging biomarkers for genetic subtype identification are more challenging tasks but have also been successfully established [5,6].…”