Open Access

ARTICLE

An Advanced Medical Diagnosis of Breast Cancer Histopathology Using Convolutional Neural Networks

Ahmed Ben Atitallah^1,*, Jannet Kamoun^2,3, Meshari D. Alanazi¹, Turki M. Alanazi⁴, Mohammed Albekairi¹, Khaled Kaaniche¹

1 Department of Electrical Engineering, College of Engineering, Jouf University, Sakaka, 72388, Saudi Arabia
2 Laboratory of Biochemistry and Enzymatic Engineering of Lipases, National Engineering School of Sfax, University of Sfax, Sfax, 3038, Tunisia
3 LIPONOV, Biological Engineering Department, HealthTech Industry, Sfax, 3038, Tunisia
4 Department of Electrical Engineering, College of Engineering, University of Hafr Al Batin, Hafr Al Batin, 39524, Saudi Arabia

* Corresponding Author: Ahmed Ben Atitallah. Email:

(This article belongs to the Special Issue: Advanced Medical Imaging Techniques Using Generative Artificial Intelligence)

Computers, Materials & Continua 2025, 83(3), 5761-5779. https://doi.org/10.32604/cmc.2025.063634

Received 20 January 2025; Accepted 10 April 2025; Issue published 19 May 2025

Abstract

Breast Cancer (BC) remains a leading malignancy among women, resulting in high mortality rates. Early and accurate detection is crucial for improving patient outcomes. Traditional diagnostic tools, while effective, have limitations that reduce their accessibility and accuracy. This study investigates the use of Convolutional Neural Networks (CNNs) to enhance the diagnostic process of BC histopathology. Utilizing the BreakHis dataset, which contains thousands of histopathological images, we developed a CNN model designed to improve the speed and accuracy of image analysis. Our CNN architecture was designed with multiple convolutional layers, max-pooling layers, and a fully connected network optimized for feature extraction and classification. Hyperparameter tuning was conducted to identify the optimal learning rate, batch size, and number of epochs, ensuring robust model performance. The dataset was divided into training (80%), validation (10%), and testing (10%) subsets, with performance evaluated using accuracy, precision, recall, and F1-score metrics. Our CNN model achieved a magnification-independent accuracy of 97.72%, with specific accuracies of 97.50% at 40×, 97.61% at 100×, 99.06% at 200×, and 97.25% at 400× magnification levels. These results demonstrate the model’s superior performance relative to existing methods. The integration of CNNs in diagnostic workflows can potentially reduce pathologist workload, minimize interpretation errors, and increase the availability of diagnostic testing, thereby improving BC management and patient survival rates. This study highlights the effectiveness of deep learning in automating BC histopathological classification and underscores the potential for AI-driven diagnostic solutions to improve patient care.

---

Keywords

---