Related Experiment Video
Updated: Jul 9, 2025
04:23
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
1.9K
Employing Atrous Pyramid Convolutional Deep Learning Approach for Detection to Diagnose Breast Cancer Tumors
Ehsan Sadeghi Pour1, Mahdi Esmaeili1, Morteza Romoozi1
1Department of Electrical and Computer Engineering, Kashan Branch, Islamic Azad University, Kashan 8715998151, Iran.
Computational Intelligence and Neuroscience
|November 29, 2023
Summary
This study introduces a novel quantum wavelet transform (QWT) filtering and Atrous pyramid convolutional neural network (APCNN) method for improved breast cancer detection in mammograms. The QWT-APCNN model enhances noise reduction and mass segmentation, leading to more accurate diagnoses.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Signal Processing
Background:
- Breast cancer is a leading cause of death in women globally, necessitating early detection for improved survival rates.
- Computer-aided diagnosis systems can assist radiologists in accurately identifying and locating breast tumors.
- Mammographic image quality is often affected by noise, hindering precise tumor detection.
Purpose of the Study:
- To evaluate noise reduction techniques for mammographic images, specifically targeting salt and pepper, Gaussian, and Poisson noise.
- To develop and assess a deep learning model for precise mass detection and segmentation in mammography.
- To introduce a hybrid methodology combining quantum wavelet transform filtering and an Atrous pyramid convolutional neural network.
Main Methods:
- Utilized the MIAS (Mammographic Image Analysis Society) dataset for the study.
- Implemented quantum wavelet transform (QWT) filtering for noise reduction in mammographic images.
- Employed an Atrous pyramid convolutional neural network (APCNN) for image classification and mass segmentation, forming the QWT-APCNN hybrid model.
Main Results:
- The proposed QWT-APCNN method demonstrated superior performance in noise reduction and mass segmentation compared to existing approaches.
- Achieved high accuracy rates, including 98.57% accuracy, 92% sensitivity, 88% specificity, and 90% Duke-Scott (DSS).
- The model also showed promising results in Receiver Operating Characteristic (ROC) and Area Under the Curve (AUC) analysis, reaching 88.77%.
Conclusions:
- The QWT-APCNN hybrid methodology offers a robust solution for enhancing mammographic image quality and improving the accuracy of breast cancer detection.
- This approach has the potential to significantly aid radiologists in making faster and more precise diagnoses, ultimately improving patient outcomes.
- The study highlights the effectiveness of combining advanced signal processing techniques with deep learning for medical image analysis.
