BC2NetRF: Breast Cancer Classification from Mammogram Images Using Enhanced Deep Learning Features and Equilibrium-Jaya Controlled Regula Falsi-Based Features Selection

Abstract

One of the most frequent cancers in women is breast cancer, and in the year 2022, approximately 287,850 new cases have been diagnosed. From them, 43,250 women died from this cancer. An early diagnosis of this cancer can help to overcome the mortality rate. However, the manual diagnosis of this cancer using mammogram images is not an easy process and always requires an expert person. Several AI-based techniques have been suggested in the literature. However, still, they are facing several challenges, such as similarities between cancer and non-cancer regions, irrelevant feature extraction, and weak training models. In this work, we proposed a new automated computerized framework for breast cancer classification. The proposed framework improves the contrast using a novel enhancement technique called haze-reduced local-global. The enhanced images are later employed for the dataset augmentation. This step aimed at increasing the diversity of the dataset and improving the training capability of the selected deep learning model. After that, a pre-trained model named EfficientNet-b0 was employed and fine-tuned to add a few new layers. The fine-tuned model was trained separately on original and enhanced images using deep transfer learning concepts with static hyperparameters’ initialization. Deep features were extracted from the average pooling layer in the next step and fused using a new serial-based approach. The fused features were later optimized using a feature selection algorithm known as Equilibrium-Jaya controlled Regula Falsi. The Regula Falsi was employed as a termination function in this algorithm. The selected features were finally classified using several machine learning classifiers. The experimental process was conducted on two publicly available datasets—CBIS-DDSM and INbreast. For these datasets, the achieved average accuracy is 95.4% and 99.7%. A comparison with state-of-the-art (SOTA) technology shows that the obtained proposed framework improved the accuracy. Moreover, the confidence interval-based analysis shows consistent results of the proposed framework.

BCUIS‐Net: A breast cancer ultrasound image segmentation network via boundary‐aware and shape feature fusion

“…To effectively integrate features from different semantic information, feature fusion and selection technology have great potential for a wide range of applications. [47][48][49] Fatima et al simultaneously combined the original images and the contrast-enhanced images as network inputs for breast cancer classification. It implemented the canonical correlation analysis-based feature fusion approach to combine the average pooling layer features generated by two deep learning frameworks and thus enabled comprehensive utilization of different semantic information and improved classification accuracy.…”

“…47 Similarly, Jabeen et al employed two parallel networks to learn features from the original and enhanced images, extracting deep features with distinct semantics from the average pooling layers and then fusing and selecting the optimal features. 48 However, the features obtained after global average pooling lack morphological representation and positional information. Therefore, these methods are not suitable for segmentation tasks.…”

An improved breast cancer classification with hybrid chaotic sand cat and Remora Optimization feature selection algorithm

“…Worldwide, BrC is a prevalent and deadly illness that affects women. Among many cancers kinds, including brain, liver, and lung cancer, breast cancer ranks third in terms of mortality [1][2][3]. Furthermore, throughout the next 20 years, there will likely be a 70% rise in the number of new BrC patients.…”

A novel framework of multiclass skin lesion recognition from dermoscopic images using deep learning and explainable AI

“…Several lesion segmentation techniques, such as thresholding, saliency and region growing, and clustering, have been developed in the literature. In traditional techniques, the segmented images are used for feature extraction; however, the problem of irrelevant features is solved by computer vision researchers using feature selection techniques ( 24 ). Using feature selection techniques, the best features are chosen from the original extracted features, resulting in a reduction in computational time ( 25 , 26 ).…”