Breast cancer is a common and fatal disease among women worldwide. Therefore, the early and precise diagnosis of breast cancer plays a pivotal role to improve the prognosis of patients with this disease. Several studies have developed automated techniques using different medical imaging modalities to predict breast cancer development. However, few review studies are available to recapitulate the existing literature on breast cancer classification. These studies provide an overview of the classification, segmentation, or grading of many cancer types, including breast cancer, by using traditional machine learning approaches through hand-engineered features. This review focuses on breast cancer classification by using medical imaging multimodalities through state-of-the-art artificial deep neural network approaches. It is anticipated to maximize the procedural decision analysis in five aspects, such as types of imaging modalities, datasets and their categories, pre-processing techniques, types of deep neural network, and performance metrics used for breast cancer classification. Forty-nine journal and conference publications from eight academic repositories were methodically selected and carefully reviewed from the perspective of the five aforementioned aspects. In addition, this study provided quantitative, qualitative, and critical analyses of the five aspects. This review showed that mammograms and histopathologic images were mostly used to classify breast cancer. Moreover, about 55% of the selected studies used public datasets, and the remaining used exclusive datasets. Several studies employed augmentation, scaling, and image normalization pre-processing techniques to minimize inconsistencies in breast cancer images. Several types of shallow and deep neural network architecture were employed to classify breast cancer using images. The convolutional neural network was utilized frequently to construct an effective breast cancer classification model. Some of the selected studies employed a pre-trained network or developed new deep neural networks to classify breast cancer. Most of the selected studies used accuracy and area-under-the-curve metrics followed by sensitivity, precision, and F-measure metrics to evaluate the performance of the developed breast cancer classification models. Finally, this review presented 10 open research challenges for future scholars who are interested to develop breast cancer classification models through various imaging modalities. This review could serve as a valuable resource for beginners on medical image classification and for advanced scientists focusing on deep learning-based breast cancer classification through different medical imaging modalities.

Deep learning-based breast cancer classification through medical imaging modalities: state of the art and research challenges

Role of image thermography in early breast cancer detection- Past, present and future

In this article, we present a new approach for breast thermogram image analysis by developing a fully automatic segmentation of right and left breast for asymmetry analysis, using shape features of the breast and Polynomial curve fitting. Segmentation results are validated with their respective Ground Truths. Histogram and grey level co-occurrence matrix-based texture features are extracted from the segmented images. Statistical test shows that features are highly significant in detection of breast cancer. We have obtained an accuracy of 90%, sensitivity of 87.5% and specificity of 92.5% for a set of eighty images with forty normal and forty abnormal using SVM RBF classifier.

Asymmetry analysis of breast thermograms using automated segmentation and texture features

Breast cancer is one of the most common types of cancer and early detection can significantly decrease the associated mortality rate. Different kinds of segmentation methods were applied to extract regions of interest from breast cancer images that are necessary to improve the classification. In this paper, a segmentation method for breast cancer from thermal images is introduced based on a proposed Chaotic Salp Swarm Algorithm (CSSA). Although the Salp Swarm Algorithm (SSA) shows superiority in single-objective optimization problems, it suffers from a low convergence rate and local optima stagnation. In the proposed method, a segmentation algorithm is formulated using the quick-shift method for superpixels extraction whose parameters are optimized by CSSA. The quick-shift method generates compact and nearly uniform superpixels by clustering the breast thermal image pixels. CSSA algorithm is developed based on ten chaotic maps to enhance the original SSA convergence rate while accuracy could be improved by controlling the balance between exploration and exploitation. The proposed algorithm is applied to real-world thermal images for the breast area. The results demonstrate that the proposed CSSA algorithm achieves fast convergence for the unimodal benchmark functions and outperforms the original SSA algorithm. Moreover, a dataset from Mastology Research with Infrared Image (DMR-IR) is used to test the performance of the proposed algorithm. In experiments, the proposed optimized segmentation algorithm extracts the breast area from the background accurately where the region of interest is focused on the breast area and removes the unwanted area such as underarms and stomach which intern can enhance the results of cancer detection. Furthermore, the proposed algorithms achieve robustness for the segmentation of different healthy and unhealthy cases images compared to the state-of-the-art methods.

/pdf/breast-cancer-segmentation-from-thermal-images-based-on-18glvsax0s.pdf

Breast Cancer Segmentation From Thermal Images Based on Chaotic Salp Swarm Algorithm

Ensuring the resiliency of critical infrastructures is essential in modern society, but much of the deployed infrastructure has yet to fully leverage modern technical developments. This pap...

Deep Learning for Critical Infrastructure Resilience

In this article, a study of various medical imaging modalities and survey of computer aided detection (CAD) approaches used for the detection of breast cancer are presented. Early breast cancer detection, is important as it can increase the rate of diagnosis, cure and survival of the affected women. Medical imaging techniques can detect early signs of breast cancer. They are also used to diagnose, stage breast cancer and guide treatments. This article explains the imaging procedure, analysis of images, benefits and limitations of standard medical imaging techniques such as mammography, ultrasound, magnetic resonance imaging, and thermography. Also some of modern imaging techniques are discussed here. Comparative study of mammography, ultrasound, thermography and magnetic resonance imaging describes the benefits of thermography in early detection of breast cancer. CAD provides a fast, reliable, and cost-effective second opinion to support doctors with the diagnostic management of breast cancer. This article reviews various CAD algorithms developed and studied for mammography, ultrasound and thermography. From the literature review, it is observed that accuracy of detection of breast abnormalities using CAD are decided by many factors, such as size of database, image quality, segmentation of region of interest, extracted features and classifiers used, etc.

Medical imaging techniques and computer aided diagnostic approaches for the detection of breast cancer with an emphasis on thermography - a review

Breast Thermography is an emerging medical imaging tool, used for early detection of breast cancer. The growing tumor in the breast will cause the occurrence of some asymmetric heat patterns in the left and right breast. Asymmetry analysis of left and right breast is used for detecting the breast cancer from the breast thermogram images. This paper presents a method for segmenting the left and right breast from breast thermogram images using Projection Profile Approach. Horizontal Projection Profile (HPP) method is used for locating the lower and upper borders of the breast by detecting the inframammary fold and axilla curves of the breast respectively. Vertical Projection Profile (VPP) method is used for locating the left and right borders of the breast thermogram image, which detects the parabolic shape of the breast. The results of segmentation are satisfactory. Generalization of this method can be done for various types of breast thermogram images by standardizing the height, background and removal of the noise present in the image.

/pdf/segmentation-of-breast-thermogram-images-for-the-detection-sf9p9qjqcy.pdf

K. V. Rajagopal

Papers

Medical imaging techniques and computer aided diagnostic approaches for the detection of breast cancer with an emphasis on thermography - a review

Segmentation of Breast Thermogram Images for the Detection of Breast Cancer: A Projection Profile Approach