Development of a feature vector for accurate breast cancer detection in mammographic images

The aim of this article was to provide early detection of breast cancer by using both mammography and histopathology images of the same patient. When the studies in the literature are examined, it is seen that the mammography and histopathology images of the same patient are not used together for early diagnosis of breast cancer. Mammographic and microscopic images can be limited when using only one of them for the early detection of the breast cancer. Therefore, multi-modality solutions that give more accuracy results than single solutions have been realized in this paper. 3 × 50 microscopic (histopathology) and 3 × 50 mammography image sets have been taken from Firat University Medicine Faculty Pathology and Radiology Laboratories, respectively. Optimum feature space has been obtained by minimum redundancy and maximum relevance via mutual information method applying to the 3 × 50 microscopic and mammography images. Then, probabilistic values of suspicious lesions in the image for selected features have been found by exponential curve fitting. Jensen Shannon, Hellinger, and Triangle measurements have been used for the diagnosis of breast cancer. It has been proved that these measures have been related to each other. Weight values for selected each feature have been found using these measures. These weight values have been used in object function. Afterward, histopathology and mammography images have been classified as normal, malign, and benign utilizing object function. In the result of this classifier, the accuracy of diagnosis of breast cancer has been estimated probabilistically. Furthermore, classifications have been probabilistically visualized on a pie chart. Consequently, the performances of Jensen Shannon, Hellinger, and Triangle measures have been compared with ROC analysis using histopathology and mammography test images. It has been observed that Jensen Shannon measure has higher performance than Hellinger and Triangle measures. Accuracy rates of histopathology and mammography images in Jensen Shannon measure have been found to 99 and 98%, respectively.

The growing rate of breast cancer necessitates immediate global attention. Mammography images are used to determine the stage of malignancy. Breast cancer stages must be identified in order to save a person's life. This article's main goal is to identify different techniques to obtain the difference between two breast cancer mammography images taken of the same individual at different times. This is the first effort to identify breast cancer in mammography images using change detection techniques. The Mammogram Image Change Detection (ICD) technique is also a recent advancement to prevent breast cancer in the early stage and precancerous level in medical images. The main purpose of this work is to observe the changes between breast cancer images in different screening periods using different techniques. Mammogram Breast Cancer Image Change Detection (MBCICD) methods usually start with a Difference Image (DI) and classify the pixels in the DI into changed and unaffected classes using unsupervised fuzzy c means (FCM) clustering methods based on texture features taken from the log and mean ratio difference pictures. Two operators, mean ratio and log ratio, were used to check the changes in the images. The Gabor wavelet is utilized as a feature extraction technique among several standards. Using the Gabor wavelet ratio operators is a useful method for altering the detection of breast cancer in mammography images. Currently, it is challenging to obtain real malignant images of the same person for testing or training. In this study, two images are utilized. To clearly see the changes, one is an image from the MIAS breast cancer mammography images dataset, and the other is a self-generated change image. The research aims to examine the image results and other quantitative analysis results of proposed change detection methods on cancer images. The Mean Ratio Accuracy result is 0.9738, and the Log ratio PCC is 0.9737. The classification results are the Log Ratio + Gabor Filter + FCM is 0.9737, and Mean Ratio +Gabor Filter + FCM is 0.9719. The mean Ratio Accuracy result is 0.9738, Log ratio is 0.9737. Log Ratio + Gabor Filter + FCM is 0.9737, Mean Ratio +Gabor Filter + FCM is 0.9719. Comparing the PCC of proposed change detection methods with the FDA-RMG method on the same dataset, the accuracy is 0.9481 only. The study concludes that variations in mammography breast cancer images could be successfully identified using the ratio operators with Gabor wavelet features.

Breast cancer is one of the major causes of death among women in recent decades. Screening mammography is currently the best available radiological technique for early detection of breast cancer. In recent years, several methods have been used for automated tumor detection in mammography images. In some methods, due to a variety of processing and multiple operations on images, there are many computational complexities and much time overhead. In other methods the recognition accuracy is relatively low. In this paper, a new method to detect cancerous lesions in mammography images is presented using cellular learning automata algorithm. Cellular learning automata algorithm is well suited for image processing, because it is cellular and belongs pixels like an image. Distributed performance and parallel processing properties of this method has optimal results in image processing. Experimental results show the effectiveness of the proposed method.

Breast cancer is one of the most frequently diagnosed solid cancers. Mammography is the most commonly used screening technology for detecting breast cancer. Traditional machine learning methods of mammographic image classification or segmentation using manual features require a great quantity of manual segmentation annotation data to train the model and test the results. But manual labeling is expensive, time-consuming, and laborious, and greatly increases the cost of system construction. To reduce this cost and the workload of radiologists, an end-to-end full-image mammogram classification method based on deep neural networks was proposed for classifier building, which can be constructed without bounding boxes or mask ground truth label of training data. The only label required in this method is the classification of mammographic images, which can be relatively easy to collect from diagnostic reports. Because breast lesions usually take up a fraction of the total area visualized in the mammographic image, we propose different pooling structures for convolutional neural networks(CNNs) instead of the common pooling methods, which divide the image into regions and select the few with high probability of malignancy as the representation of the whole mammographic image. The proposed pooling structures can be applied on most CNN-based models, which may greatly improve the models' performance on mammographic image data with the same input. Experimental results on the publicly available INbreast dataset and CBIS dataset indicate that the proposed pooling structures perform satisfactorily on mammographic image data compared with previous state-of-the-art mammographic image classifiers and detection algorithm using segmentation annotations.

Automated breast ultrasound imaging (ABUS) results in a reduction in breast cancer stage at diagnosis beyond that seen with mammographic screening in women with increased breast density or who are at a high risk of breast cancer. It is unknown if the addition of ABUS to mammography or ABUS imaging alone, in this population, is a cost-effective screening strategy. A discrete event simulation (Monte Carlo) model was developed to assess the costs of screening, diagnostic evaluation, biopsy, and breast cancer treatment. The number of quality-adjusted life years gained through each screening method is assessed using previously published quality of life measures. Incremental cost-effectiveness ratios for screening with the combination of mammographic and ABUS imaging, and for ABUS imaging alone are calculated as compared to standard mammographic imaging. Combined screening with both mammographic and ABUS imaging results in an incremental cost-effectiveness ratio of $7,071 ($6,332-$7,809) when compared to traditional mammographic imaging (p < 0.05). ABUS screening alone results in an incremental cost-effectiveness ratio of $3,559 ($-965-$8,082) when compared to mammographic imaging (p < 0.05). ABUS screening alone is more likely to be cost-effective for a willingness-to-pay of less than $7,100. The addition of ABUS to mammographic imaging is a cost-effective screening strategy in women with increased breast density or who are at a high risk of developing breast cancer. ABUS imaging alone is also a cost-effective strategy in this population, particularly in resource-poor areas.

The prompt and accurate diagnosis of breast lesions, including the distinction between cancer, non-cancer, and suspicious cancer, plays a crucial role in the prognosis of breast cancer. In this paper, we introduce a novel method based on feature extraction and reduction for the detection of breast cancer in mammography images. First, we extract features from multiple pre-trained convolutional neural network (CNN) models, and then concatenate them. The most informative features are selected based on their mutual information with the target variable. Subsequently, the selected features can be classified using a machine learning algorithm. We evaluate our approach using four different machine learning algorithms: neural network (NN), k-nearest neighbor (kNN), random forest (RF), and support vector machine (SVM). Our results demonstrate that the NN-based classifier achieves an impressive accuracy of 92% on the RSNA dataset. This dataset is newly introduced and includes two views as well as additional features like age, which contributed to the improved performance. We compare our proposed algorithm with state-of-the-art methods and demonstrate its superiority, particularly in terms of accuracy and sensitivity. For the MIAS dataset, we achieve an accuracy as high as 94.5%, and for the DDSM dataset, an accuracy of 96% is attained. These results highlight the effectiveness of our method in accurately diagnosing breast lesions and surpassing existing approaches.

Expectation–maximization technique for fibro-glandular discs detection in mammography images

Detection of the area withholding the mitotic cell growth is a vital marker in breast cancer detection. This paper aims to fabricate an automatic Computer-Aided Detection (CAD) [2] model that helps to locate the region of mitotic cell growth and signify the type of breast cancer found: benign or malignant. Contrary to the legacy literature [1], which uses partially supervised learning models applied to histopathology images, we devise a model which involves a fully supervised convolution model applied to mammographic images. The model trained with image datasets: benign and malignant breast cancer images. This model exploits the MIAS database and datasets collected privately from hospitals, consisting of mammographic images available as samples for breast cancer detection. Applying image segmentation techniques to the datasets, we highlight the region of interest, and thereby using classification methodologies, we separate the results as benign or malignant. The developed model equips us to yield results with an accuracy of 97.96% on the dataset.

Breast cancer is one of the non-contagious diseases that tends to increase every year. This disease occurs almost entirely in women, but can also occur in men. One way to detect this disease is by observing mammography images. However, mammography images often tend to be blurry with low quality so that it is possible to detect them incorrectly. Therefore, in this study, automatic classification of breast cancer on mammographic images was carried out using the Convolutional Neural Network (CNN). This proposed system uses the VGG16 architecture with a transfer learning system. The proposed system is then optimized using Adam optimizers and RMSprop optimizers. The results of system testing for normal, benign, and malignant classifications obtained an accuracy value of 80% - 90% with the highest accuracy achieved using Adam's optimizers. With this proposed system, it is hoped that it can help in the clinical diagnosis of breast cancer.

Breast cancer detection in mammography images using Neighborhood Attention transformer and Shearlet Transform.

Background. Triple negative breast cancer has a poor prognosis. Therefore, it is vital to detect this subtype of breast cancer in its early stage. The imaging features of this clinically important subtype of breast cancer are not well known. There have been no published reports about radiological diagnostics of triple negative breast tumour in Lithuania. The purpose of this study was to review the imaging characteristics of triple receptor negative cancers in mammography, ultrasonography and magnetic resonance imaging (MRI). Materials and methods. The published data for the period 2006–2011 concerning the imaging of triple negative breast cancer were analyzed. There were ten retrospective, ten prospective studies and five reviews. Five studies were on mammography imaging, three on both mammography and ultrasonography imaging, and five studies dealt with MR imaging data. Two studies analysed all three diagnostic methods. Results. In mammography, triple negative breast (TRN) cancers often present as a mass and are most frequently round, oval or lobular in shape, less frequently being irregular. TRN tumours aren’t associated with calcifications. Moreover, architectural distortion is not a characteristic feature of triple negative breast cancer. In ultrasonography, TRN cancer appears as a parallel. TRN breast tumours mostly are irregular in shape and have a circumscribed margin. Attenuating posterior echoes and hypervascularity are not their characteristic features. In MR imaging, TRN breast cancer tends to have a lobulated, round or oval mass shape. Rim enhancement is identified in most of TRN tumours. Initially, rapid enhancement with a washout pattern (a sign of malignancy) does not usually apply to triple-negative breast cancers. Conclusions. TRN breast cancer is difficult to diagnose, because usually it has no specific imaging signs typical of breast cancer. In mammography, TRN cancers aren’t associated with microcalcifications. In ultrasonography, attenuating posterior echoes and hypervascularity are not characteristic features of TRN tumours. In MRI, initially rapid enhancement with a washout pattern does not usually apply to triple-negative breast cancers. Keywords: triple negative breast cancer, mammography, ultrasonography, magnetic resonance imaging

Despite the most significant cancer that occurs worldwide is breast cancer for women. All almost all parts of the country, breast cancer detection increase every year, and the death rate of breast cancer is nearly (20–27) % in India. The input mammography image i.e left side and right side of the breast segment and optimize concerning pixel shape and intensity. Furthermore, the proposed method of segmentation thresholding algorithm optimizes and classifies the mammographic image through two methods. The first method of optimization of segmentation image through the convex border of mammographic images and detect the edges and boundary. The second method of optimization image through the non-convex border and detect edges and boundaries. Finally, the result is processed through a segmentation thresholding algorithm to delineate the edges and boundaries in the mammographic images and helps the radiologist to detect earlier stages of breast cancer. The net result outer performs the existing algorithm and proposed in the state of art to detect breast cancer in earlier stage and help the Radiologist for accuracy in cancer detection. The achieved result of the segmentation thresholding algorithm through the convex and non-convex border of image optimization is about 93% of the accuracy.

Breast cancer is the uncontrolled proliferation of a group of cells in the breast and is the second leading cause of death for women in the world. The disease can be cured if detected in the early stages. A lot of research has been done to correctly detect the tumor, but a 100% accurate method has not been found. Research on breast cancer detection using digital image processing is not new, but many new approaches are being considered in this field to accurately predict the tumor area. The current approach consists of detecting the tumor area visually and also finding out in which area the tumor is most concentrated. CC and MLO dualview mammographic screening images are widely used in the diagnostic process. This project presents a method for detecting a tumor area and classifying a normal and oncological patient. Preprocessing operations are performed on the input mammographic image and unwanted parts are removed from the image. Tumor regions are segmented from the image using a morphological operation and are highlighted on the original mammographic image. If the image on the mammogram is normal, it means that the patient is healthy. This work mainly focuses on finding the best algorithms for detecting tumors present in the breast. A number of algorithms were used in the proposed work, but the most suitable for cancer detection is the combination of K-Means clustering algorithm. K-Means classification accuracy is 95% accurate output will be predicted. Keywords: Image Processing, Breast Cancer, K-Means clustering, Dilation, Closing, Edge Detection, Mammography screening images

Breast cancer is a prevalent and potentially life-threatening medical condition characterized by uncontrolled cell proliferation within breast tissue. This global health concern predominantly among women, although it can affect men as well. The timely and accurate detection of breast cancer is critical, as it significantly influenced treatment outcomes and survival rates. Despite advancements in human-based diagnostic methods, inherent limitations such as subjectivity, human error, and fatigue persist. To address these constraints, computer-vision-based techniques have been explored for breast cancer detection, aiming to enhance early detection, reduce diagnostic errors, and improve patient outcomes. This study advances previous approaches by integrating group convolution and the Special Euclidean motion group as a features extractor into the Faster Region Convolutional Neural Networks framework of Detectron2. This integration offers the advantage of enhancing the Convolutional Neural Network by incorporating a method to preserve invariant and equivariant features, effectively leveraging symmetries in mammography images. The INbreast dataset, comprising 410 images from 115 cases, including 90 instances with bilateral breast involvement in women, was utilized. To ensure data compatibility, DICOM to PNG and XML to JSON format conversions were performed. A data enhancement pipeline, encompassing techniques such as image cropping, truncation normalization, contrast-limited adaptive histogram equalization, and image synthesis, was employed for data preprocessing and cleaning. The proposed technique demonstrated competitive results in breast cancer detection, achieving a recall rate of 97.22. This underscores the efficacy of the integrated approach in improving diagnostic accuracy and holds promise for advancing computer-vision-based breast cancer detection methodologies.

Breast cancer is one of the most common cancer among women and the survival rate tends to be low when its stage found high when treated. To improve breast cancer survival, early detection is critical. There are two ways of detection for breast cancer: early diagnosis and screening. To make an accurate diagnosis in the early stage of breast cancer, the appearance of masses and micro-calcifications on the mammography image are two important indicators. It is time-consuming and challenging to identify micro-calcification from mammogram images by the human eye because of its size and appearance. Several Computer-Aided Detection (CADe) have been developed to support radiologists because the automatic detection of micro-calcification is important for diagnosis and the next recommended treatment. Most of the current CADe systems at this time started using Convolutional Neural Network (CNN) to implement the micro-calcification detection in mammograms and their quantitative results are very satisfying, the average level of accuracy is more than 90%. However, most of the methods used are image fragments from a complete image which are then included in the program. This research conducts an automated approach to detect the location of any micro-calcification in the mammogram images with the complete image and in a simple way. At first, the image preprocessing algorithms were applied to enhance the image quality. After that, the micro-calcification region was labeled using image segmentation based on the Radiologist's expertise. The positive label which contains micro-calcification pixels was taken to train with segmentation network. A total of 354 images from INbreast dataset were used in this research study. Finally, the trained network was utilized to detect the micro-calcification area automatically from the mammogram images. This process can help as an assistant to the radiologist for early diagnosis and increase the detection accuracy of the micro-calcification regions. The proposed system performance is measured according to the error values of Mean Squared Logarithmic Error (MSLE) as the technique to find out the difference between the values predicted by the proposed model and the actual values. The best MSLE loss value obtained was achieved in 0.05 with accuracy 0.95.