An Efficient AI-Based Framework for Early Breast Cancer Detection Using Mammographic Image Analysis

Le résumé fourni par la source

Breast cancer continues to be the primary cause of death among women globally, and early detection remains the most significant factor in raising the survival rates. Despite the fact that traditional diagnostic methods are still quite effective, they still face issues related to accuracy, computational efficiency, and accessibility, with low-resource areas being the worst affected. The present study proposes a novel algorithmic approach for the reliable detection of breast cancer that combines state-of-the-art imaging techniques and machine learning models that are made to the point for the purpose of accuracy and efficiency in calculation. The creation of an original hybrid algorithm that melds together the various feature extraction techniques and a lightweight classification model that is specifically tailor-made for mammogram image analysis is a significant achievement of the researchers. The developed method addresses the primary issues of the existing detection systems such as the occurrence of high false-positive rates, long processing times, and the demand for more powerful computing resources. Through comprehensive trials executed on standardized breast cancer datasets, it has been established that the algorithm is superior in terms of accuracy, sensitivity, and specificity, and also faster in processing compared to traditional methods. The methodology of the study includes collecting data from mammography databases that are publicly available, preprocessing data to improve the quality of images, extracting features using new descriptors, and classifying them with the help of an optimized ensemble learning method. Validation is done extensively through the application of cross-validation techniques and by comparing the results with those from the established benchmark algorithms. The outcomes of the study show that the method proposed in the study reaches an accuracy of 96.8%, sensitivity of 95.4%, and specificity of 97.2%, while the computational time is reduced by 43% compared to the state-of-the-art methods. This research not only adds to the body of knowledge but also provides a practical solution by offering a new algorithmic framework that wraps up detection performance with computational efficiency, hence making it a viable option for medical settings with limited computational resources. The results have far-reaching consequences for the enhancement of early breast cancer screening programs, especially in developing countries where the access to advanced diagnostic resources is still a challenge.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.