Advances in Machine, Deep, and Explainable AI-based Semantic Segmentation for Brain Tumor and Lesion Detection

View at publisher Summarise with VIRAT
Authors: Divya Boddukoori, Arwinder Dhillon, Vishal Walia
Cited by: 0
References: 59

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Pinpointing and dividing brain tumors is crucial in the patient’s treatment process. It is necessary to develop techniques for the automatic, precise and interpretable analysis of medical images. Machine and deep learning techniques that are optimally used for segmentation allows for the segmentation of pathological region ‘pixel-wise’. Standard ML methods require hand-designed features and use classical classifiers. Alternatively, deep learning methods use convolutional and recurrent architectures to achieve end-to-end activations to improve the accuracy and robustness of these methods. Although DL models offer great accuracy and precision advantages, they have a patent black box which hinders clinical uptake. XAI offers a clear view of the working conditions of an automated decision system. The automated process is explained, model transparency is achieved and the performance of the system is showcased through counterfactuals. The research work is an extensive survey on the Semantic Segmentation ML, DL, XAI Based Techniques for brain tumor lesion. The paper outlines how they generated the dataset and the challenges faced. Findings from the survey indicates that integration of XAI in clinical workflow, transfer learning within limited data capacity and multi-modal imaging solutions are the future directions and open research issues related to XAI. The aim of this study is to provide the sufficient information on ending segmentation so as to come up with high performing models that are interpretable for medical image analysis.

Fields of Study & Subjects

Brain Tumor Detection and Classification Medical Image Segmentation Techniques AI in cancer detection

Publication details

Journal / source
International Conference on Image Processing and Capsule Networks (ICIPCN)
Pages
646–654
Publication date
27 January 2026
Language
EN