Cancer Medical Image Analysis Across Modalities Using Attention Mechanism and Capsule Network

Medical image analysis is a critical component of accurate disease diagnosis and treatment planning which relies on traditional methods such as MRI, CT scans, and X-rays. However, challenges arise from these traditional methods due to human errors in image interpretation, lack of standardization in...

Full description

Saved in:
Bibliographic Details
Main Authors: A, Sharon Deborah, Beryl Princess, P. Joyce, Leelipushpam Paulraj, Getzi Jeba
Format: Conference Proceeding
Language:English
Subjects:
Accuracy
Adaptation models
Analytical models
attention mechanism
Attention mechanisms
cancer
capsule network
deep neural network
histology
Image analysis
Magnetic resonance imaging
Medical diagnosis
Medical diagnostic imaging
medical image analysis
medical modalities
Microscopy
MRI scans
X-rays
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Medical image analysis is a critical component of accurate disease diagnosis and treatment planning which relies on traditional methods such as MRI, CT scans, and X-rays. However, challenges arise from these traditional methods due to human errors in image interpretation, lack of standardization in medical image analysis, varying expertise levels among healthcare professionals, and the potential oversight of critical details which leads to variability in diagnosis and treatment outcomes across facilities. To address these issues, advanced computational methods are imperative for medical conditions that are hard to detect visually. This paper proposes an integrated approach using the Deep neural network with Attention mechanism and Capsule Network to deal with diverse imaging modalities, especially histology, microscopy, and MR scans. This innovative approach represents a superior alternative to traditional methods by emphasizing its potential to focus on crucial regions within medical images and to enhance disease prediction which holds a significant progress to contribute to a faster and more accurate diagnostic process, ultimately benefiting both the healthcare industry and patients.
ISSN:2469-5556
DOI:10.1109/ICACCS60874.2024.10717118