ABSTRACT
This study presents a new diagnostic method for breast cancer that combines Histogram of Oriented Gradients (HOG) feature extraction with Support Vector Machine (SVM) classification and adopts a hybrid strategy. Compared to traditional methods, the HOG- SVM combination enables classification of intensity gradients to detect subtle patterns in medical images. Hybrid strategies achieve this by combining knowledge of the subject and the legal system, allowing for an expanded and defined diagnostic process. This integration aims to improve the accuracy and meaning of cancer diagnosis and provide doctors with the most advanced tools. Statistical analysis of different datasets demonstrates the effectiveness of HOG-SVM and hybrid techniques in achieving significance and specificity. This approach provides an effective way to increase the efficiency and reliability of breast cancer diagnosis by providing a unique combination of computer vision models and techniques as a rule in treatment.
