Deep learning based clinico-radiological model for paediatric brain tumor detection and subtype prediction

Abhishek Mahajan; Mayur Burrewar; Ujjwal Agarwal; Bharadwaj Kss; Apparao Mlv; Amrita Guha; Arpita Sahu; Amit Choudhari; Vivek Pawar; Vivek Punia; Sridhar Epari; Ayushi Sahay; Tejpal Gupta; Girish Chinnaswamy; Prakash Shetty; Aliasgar Moiyadi

doi:10.37349/etat.2023.00159

Deep learning based clinico-radiological model for paediatric brain tumor detection and subtype prediction

Explor Target Antitumor Ther. 2023;4(4):669-684. doi: 10.37349/etat.2023.00159. Epub 2023 Aug 30.

Authors

Affiliations

¹ Clatterbridge Centre for Oncology NHS Foundation Trust, L7 8YA, Liverpool, UK.
² Department of Radiodiagnosis, Tata Memorial Hospital, Parel, Mumbai 400012, Maharashtra, India.
³ Endimension Technology Pvt Ltd, Maharashtra, India.
⁴ Department of Pathology, Tata Memorial Hospital, Parel, Mumbai 400012, India.
⁵ Department of Paediatric Oncology, Tata Memorial Hospital, Parel, Mumbai 400012, India.
⁶ Department of Surgical Oncology, Tata Memorial Hospital, Parel, Mumbai 400012, India.

Abstract

Aim: Early diagnosis of paediatric brain tumors significantly improves the outcome. The aim is to study magnetic resonance imaging (MRI) features of paediatric brain tumors and to develop an automated segmentation (AS) tool which could segment and classify tumors using deep learning methods and compare with radiologist assessment.

Methods: This study included 94 cases, of which 75 were diagnosed cases of ependymoma, medulloblastoma, brainstem glioma, and pilocytic astrocytoma and 19 were normal MRI brain cases. The data was randomized into training data, 64 cases; test data, 21 cases and validation data, 9 cases to devise a deep learning algorithm to segment the paediatric brain tumor. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the deep learning model were compared with radiologist's findings. Performance evaluation of AS was done based on Dice score and Hausdorff95 distance.

Results: Analysis of MRI semantic features was done with necrosis and haemorrhage as predicting features for ependymoma, diffusion restriction and cystic changes were predictors for medulloblastoma. The accuracy of detecting abnormalities was 90%, with a specificity of 100%. Further segmentation of the tumor into enhancing and non-enhancing components was done. The segmentation results for whole tumor (WT), enhancing tumor (ET), and non-enhancing tumor (NET) have been analyzed by Dice score and Hausdorff95 distance. The accuracy of prediction of all MRI features was compared with experienced radiologist's findings. Substantial agreement observed between the classification by model and the radiologist's given classification [K-0.695 (K is Cohen's kappa score for interrater reliability)].

Conclusions: The deep learning model had very high accuracy and specificity for predicting the magnetic resonance (MR) characteristics and close to 80% accuracy in predicting tumor type. This model can serve as a potential tool to make a timely and accurate diagnosis for radiologists not trained in neuroradiology.

Keywords: Deep learning model; artificial intelligence; brainstem glioma; ependymoma; medulloblastoma; paediatric brain tumors; pilocytic astrocytoma.