CDAE-R: Multifunctional End-to-End Model for Brain Abnormality Images Classification and Denoising

DOI

10.2991/978-94-6463-540-9_93How to use a DOI?

Keywords

Autoencoder; Brain Abnormality Images; Residual Learning

Abstract

Traditionally, medical image classification and denoising tasks are conducted and evaluated separately, which may waste computational resources and incur excessive expenses. Besides, the features extracted by different models cannot be shared and utilized effectively. Therefore, an end-to-end multimodal neural network model is required and of importance. For the first time, this study attempted and proposed a multimodal autoencoder-based model, Classification, and Denoising Autoencoder with Residual blocks (CDAE-R) on the dataset composed of three brain abnormalities, aneurysm, tumor, and cancer for image denoising and abnormality classification. The study implemented the basic structure of an autoencoder, consisting of an encoder and an interconnected decoder, as the model’s foundation. The concept of residual learning was incorporated into CDAE-R in the form of residual blocks, enabling the preservation of essential features. On top of the traditional denoising autoencoder, the study employed and connected the 2D convolutional dense layers and pooling layers after the latent space at the end of the autoencoder component for classification. CDAE-R was then trained and evaluated on joint loss of classification and denoising with equal weight coefficient. The experiment results achieved 100% classification accuracy, competitive denoising, and reconstruction performance on the test dataset, indicating the effectiveness and feasibility of CDAE-R.

Copyright

© 2024 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Zezhou Wang
PY  - 2024
DA  - 2024/10/16
TI  - CDAE-R: Multifunctional End-to-End Model for Brain Abnormality Images Classification and Denoising
BT  - Proceedings of the 2024 2nd International Conference on Image, Algorithms and Artificial Intelligence (ICIAAI 2024)
PB  - Atlantis Press
SP  - 930
EP  - 938
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-540-9_93
DO  - 10.2991/978-94-6463-540-9_93
ID  - Wang2024
ER  -