Accurate Diagnosis Of Breast Cancer Research Articles

Artificial intelligence in breast imaging: potentials and challenges.

Breast cancer, which is the most common type of malignant tumor among humans, is a leading cause of death in females. Standard treatment strategies, including neoadjuvant chemotherapy, surgery, postoperative chemotherapy, targeted therapy, endocrine therapy, and radiotherapy, are tailored for individual patients. Such personalized therapies have tremendously reduced the threat of breast cancer in females. Furthermore, early imaging screening plays an important role in reducing the treatment cycle and improving breast cancer prognosis. The recent innovative revolution in artificial intelligence (AI) has aided radiologists in the early and accurate diagnosis of breast cancer. In this review, we introduce the necessity of incorporating AI into breast imaging and the applications of AI in mammography, ultrasonography, magnetic resonance imaging, and positron emission tomography/computed tomography based on published articles since 1994. Moreover, the challenges of AI in breast imaging are discussed.

Photon Absorption Remote Sensing Imaging of Breast Needle Core Biopsies Is Diagnostically Equivalent to Gold Standard H&E Histologic Assessment.

Photon absorption remote sensing (PARS) is a new laser-based microscope technique that permits cellular-level resolution of unstained fresh, frozen, and fixed tissues. Our objective was to determine whether PARS could provide an image quality sufficient for the diagnostic assessment of breast cancer needle core biopsies (NCB). We PARS imaged and virtually H&E stained seven independent unstained formalin-fixed paraffin-embedded breast NCB sections. These identical tissue sections were subsequently stained with standard H&E and digitally scanned. Both the 40× PARS and H&E whole-slide images were assessed by seven breast cancer pathologists, masked to the origin of the images. A concordance analysis was performed to quantify the diagnostic performances of standard H&E and PARS virtual H&E. The PARS images were deemed to be of diagnostic quality, and pathologists were unable to distinguish the image origin, above that expected by chance. The diagnostic concordance on cancer vs. benign was high between PARS and conventional H&E (98% agreement) and there was complete agreement for within-PARS images. Similarly, agreement was substantial (kappa > 0.6) for specific cancer subtypes. PARS virtual H&E inter-rater reliability was broadly consistent with the published literature on diagnostic performance of conventional histology NCBs across all tested histologic features. PARS was able to image unstained tissues slides that were diagnostically equivalent to conventional H&E. Due to its ability to non-destructively image fixed and fresh tissues, and the suitability of the PARS output for artificial intelligence assistance in diagnosis, this technology has the potential to improve the speed and accuracy of breast cancer diagnosis.

A novel breast cancer detection system using SDM-WHO-RNN classifier with LS-CED segmentation

Breast cancer (BC) is caused by the abnormal and rapid growth of breast cells. Accurate diagnosis of BC at an early stage could minimize the mortality related to this disease. Because of this, researchers have recently developed an automatic recognition system. However, poor image resolution, lower accuracy for deep tumors, biased nature and incapability to localize a lesion, greater rate of False Positives (FPs), and inefficiency towards the overlapped cells are the restrictions encompassed in the majority of the prevailing techniques. The main motive of this proposed work is for detecting the nuclei and classifying the tumor. Finding the nuclei's location is the main goal of the proposed endeavour. The structure of the cell is measured in relation to the nuclei. Hence, superior accuracy in complexity classification could be attained by this incentive. The input image is preprocessed and the contrast is enhanced. By utilizing Linear Scaling centered Canny Edge Detection (LS-CED), these Contrast-Enhanced Images (CEI) are further segmented into several partitions. The nuclei are sensed by utilizing a Soft Plus-Max Region-centered Fully Convolutional Network (SPM-R-FCN) from the segmented images. Then, as per the size and shape of the tumor, the Tumor Cells (TCs) are gauged and classified by utilizing the adaptive threshold function. After that, the most momentous and needed features are extorted as of the gauged TCs. This Extracted Feature (EF) is inputted into the SDM-WHO-RNN classifier that classifies cancer as benign, malignant, and normal. Simulation results demonstrated that the SDM-WHO-RNN approach acquires perfect outcomes as of the experiment and accomplishes 97.9% accuracy.

Blockchain and Classification of Mammograms and Histopathology Images in Breast Cancer Lesions

This work provides an overview of the research that has been done on using Computer-Aided Detection (CAD) for the diagnosis of breast cancer. The focus is on mammographic and histopathology images and the different techniques used for image pre-processing, segmentation, and classification. The accuracy of the different algorithms was evaluated on different datasets, including MIAS, IRMA, DDSM, and CBIS-DDSM, and the results showed that deep learning models such as Convolutional Neural Networks (CNNs) and Random Forest, along with Multi-Layer Perception and Nave Bayes, were effective in detecting and classifying breast cancer. The results of these studies show the potential of CAD in making the diagnosis of breast cancer easier and more accurate. In the segmentation stage, the U-Net model has been modified to perform better in terms of accuracy. The DDSM database has been found to have a higher accuracy percentage compared to the MIAS and CBIS-DDSM databases. This indicates that the modified U-Net model has performed well on the DDSM database in terms of accurately segmenting the images. The use of Computer Aided Diagnosis (CAD) has improved the accuracy of breast cancer diagnosis in mammography and histopathology images. The four-step process of preprocessing, segmentation, feature extraction, and classification has proven to be effective in detecting malignant and benign tumors. Different algorithms like Naive Bayes, Multilayer Perception, Random Forest, and Convolutional Neural Network (CNN) have been used with varying accuracy levels. The best results were obtained by using the modified U-Net model for segmentation and the Random Forest algorithm for classification, as they showed higher accuracy compared to other methods and databases. In this study it has been analyzed that the use of Computer-Aided Diagnosis (CAD) in mammography and histopathology images has shown promising results in the diagnosis of breast cancer. Different algorithms and techniques have been used to improve the accuracy of the classification of malignant and benign tumors. The MIAS database has shown the best results when using a Convolutional Neural Network approach. However, histopathology images have some limitations in comparison to radiologist images, but the use of CAD is still an important tool in cancer diagnosis.

The effect of image resolution on convolutional neural networks in breast ultrasound

PurposeThe objective of this research was to investigate the efficacy of various parameter combinations of Convolutional Neural Networks (CNNs) models, namely MobileNet and DenseNet121, and different input image resolutions (REZs) ranging from 64×64 to 512×512 pixels, for diagnosing breast cancer. Materials and methodsDuring the period of June 2015 to November 2020, two hospitals were involved in the collection of two-dimensional ultrasound breast images for this retrospective multicenter study. The diagnostic performance of the computer models MobileNet and DenseNet 121 was compared at different resolutions. ResultsThe results showed that MobileNet had the best breast cancer diagnosis performance at 320×320pixel REZ and DenseNet121 had the best breast cancer diagnosis performance at 448×448pixel REZ. ConclusionOur study reveals a significant correlation between image resolution and breast cancer diagnosis accuracy. Through the comparison of MobileNet and DenseNet121, it is highlighted that lightweight neural networks (LW-CNNs) can achieve model performance similar to or even slightly better than large neural networks models (HW-CNNs) in ultrasound images, and LW-CNNs' prediction time per image is lower.

Survey and comparative analysis of machine learning algorithms for breast cancer diagnosis: A comprehensive review

Breast cancer is a significant health concern worldwide, and early and accurate diagnosis plays a crucial role in improving patient outcomes. Machine learning algorithms have emerged as powerful tools for analyzing complex medical data and aiding in the diagnosis of breast cancer. This paper provides an overview of the application of machine learning algorithms in breast cancer diagnosis. The findings indicate that machine learning algorithms, such as support vector machines (SVM), random forests, artificial neural networks (ANN), and deep learning models, have been extensively explored for breast cancer diagnosis. These algorithms leverage the vast amounts of available data, including patient demographics, medical history, imaging data (mammography, ultrasound, MRI), and genetic profiles, to identify patterns and make predictions. One of the primary applications of machine learning algorithms in breast cancer diagnosis is the classification of tumors as malignant or benign. By training on labeled datasets, these algorithms can learn to differentiate between cancerous and non-cancerous cases, thus assisting in accurate tumor diagnosis. Additionally, machine learning algorithms can be used to predict the likelihood of cancer recurrence, which helps guide treatment decisions and post-treatment monitoring. Feature selection and extraction techniques also play a vital role in breast cancer diagnosis using machine learning algorithms. These techniques aim to identify the most relevant features or biomarkers associated with breast cancer, reducing the dimensionality of the data and enhancing the performance of the models. Feature selection algorithms, such as recursive feature elimination and correlation-based feature selection, contribute to the identification of critical indicators for accurate diagnosis. Furthermore, the integration of different data sources and modalities, such as combining clinical data with imaging data or genetic data, has shown promise in improving breast cancer diagnosis accuracy. By fusing multiple types of information, machine learning algorithms can leverage the complementary nature of these data sources to enhance diagnostic capabilities. Despite the advancements made, challenges remain in the field of breast cancer diagnosis using machine learning algorithms. Issues such as data quality, interpretability of models, and generalizability to diverse populations need to be addressed to ensure the reliable and equitable application of these algorithms in clinical practice.

A Light Gradient-Boosting Machine algorithm with Tree-Structured Parzen Estimator for breast cancer diagnosis

Breast cancer is a common and potentially life-threatening disease. Early and accurate diagnosis of breast cancer is crucial for effective treatment and improved patient outcomes. This study proposed using the Light Gradient-Boosting Machine (LightGBM) algorithm, Borderline- Synthetic Minority Oversampling Technique (SMOTE), and the Tree-Structured Parzen Estimator (TPE) for hyperparameter tuning to enhance the effectiveness of the Machine Learning (ML) model for diagnosing breast cancer. A 10-fold cross-validated TPE optimized Borderline-SMOTE LightGBM classifier was modelled on the Wisconsin Diagnostic Breast Cancer (WDBC) Dataset and evaluated for its performance compared to a baseline LightGBM model. The TPE-optimized Borderline-SMOTE LightGBM model exhibited a significant improvement in performance over the baseline model, achieving an average accuracy of 99.12%, specificity of 100%, precision of 100%, recall of 97.62%, F1-score of 98.80%, and a Mathews Correlation Coefficient of 98.12%. Compared to previous studies, the TPE-optimized Borderline-SMOTE LightGBM model performed exceptionally well. The study demonstrates the effectiveness of using data augmentation and hyperparameter optimization techniques to improve the performance of ML models for breast cancer diagnosis, which has significant implications for the medical field where the accurate and efficient diagnosis of breast cancer is critical.

Feature Selection Using a Hybrid Approach Depends on Filter and Wrapper Methods for Accurate Breast Cancer Diagnosis

Breast cancer is the biggest cause of mortality in women, outscoring all other malignancies. Diagnosing breast cancer is hard because the disease is complicated, treatment methods change, and there are many different kinds of patients. Information technology and artificial intelligence contribute to improve diagnostic procedures, which are critical for care and treatment as well as reducing and controlling cancer recurrence. The primary part of this research is to develop a new feature selection strategy based on a hybrid approach that combines two methods for selecting features: the filter and the wrapper. In two stages, this method reduces the number of features from 30 to 15 to increase and improve classification accuracy. The suggested method was tested using the Wisconsin Breast Cancer Dataset dataset (WDBC). To enhance the classification of breast cancer tumors, a soft voting classifier was used in this study. The proposed methodology outperforms previous research, achieving 1 for the F1 score, 1 for AUC, 1 for recall, 1 for precision, and 100% for accuracy. Furthermore, 10-fold cross-validation has a 98.2% accuracy rate.

A survey: Breast Cancer Classification by Using Machine Learning Techniques

Breast cancer in general is a common and deadly disease. Early detection can significantly reduce the chances of death. Using automated feature extraction and classification algorithms, physicians' experience in diagnosing and detecting breast cancer can be aided. This paper focuses on various statistical and machine learning studies of mammography datasets for enhancing the accuracy of breast cancer diagnosis and classification based on various variables. The Naïve Bayes, the K-nearest neighbors (KNN), the Support Vector Machine (SVM), the Random Forest, the Logistic Regression, Multilayer Perceptron (MLP), fuzzy classifier, and Convolutional Neural Network (CNN) classifiers, are the most widely used technologies in this field. In this study, we provide an overview of the existing CAD systems based on artificial intelligence classification techniques and many types of medical image modalities. Potential research initiatives to build more efficient and accurate CAD systems have been investigated.

Machine Learning-Based Label-Free SERS Profiling of Exosomes for Accurate Fuzzy Diagnosis of Cancer and Dynamic Monitoring of Drug Therapeutic Processes.

Exosomes are a class of extracellular vesicles secreted by cells, which can be used as promising noninvasive biomarkers for the early diagnosis and treatment of diseases, especially cancer. However, due to the heterogeneity of exosomes, it remains a grand challenge to distinguish accurately and reliably exosomes from clinical samples. Herein, we achieve accurate fuzzy discrimination of exosomes from human serum samples for accurate diagnosis of breast cancer and cervical cancer through machine learning-based label-free surface-enhanced Raman spectroscopy (SERS), by using "hot spot" rich 3D plasmonic AuNPs nanomembranes as substrates. Due to the existence of some weak distinguishable SERS fingerprint signals and the high sensitivity of the method, the machine learning-based SERS analysis can precisely identify three (normal and cancerous) cell lines, two of which are different types of cancer cells, without specific labeling of biomarkers. The prediction accuracy based on the machine learning algorithm was up to 91.1% for the discrimination of different cell lines (H8, HeLa, and MCF-7 cell)-derived exosomes. Our model trained with SERS spectra of cell-derived exosomes could reach 93.3% prediction accuracy for clinical samples. Furthermore, the action mechanism of the chemotherapeutic process of MCF-7 cells can be revealed by dynamic monitoring of SERS profiling of the exosomes secreted. The method would be useful for noninvasive and accurate diagnosis and postoperative assessment of cancer or other diseases in the future.

Development of an anatomical breast phantom from polyvinyl chloride plastisol with lesions of various shape, elasticity and echogenicity for teaching ultrasound examination.

The WHO reported an increasing trend in the number of new cases of breast cancer, making it the most prevalent cancer in the world. This fact necessitates the availability of highly qualified ultrasonographers, which can be achieved by the widespread implementation of training phantoms. The goal of the present work is to develop and test an inexpensive, accessible, and reproducible technology for creating an anatomical breast phantom for practicing ultrasound diagnostic skills in grayscale and elastography imaging, as well as ultrasound-guided biopsy sampling. We used FDM 3D printer and PLA plastic for printing an anatomical breast mold. We made a phantom using a mixture of polyvinyl chloride plastisol, graphite powder, and metallic glitter to simulate soft tissues and lesions. Various degrees of elasticity were imparted using plastisols of stiffness ranging from 3 to 17 on the Shore scale. The lesions were shaped by hand. The materials and methods used are easily accessible and reproducible. Using the proposed technology, we have developed and tested a basic, differential, and elastographic versions of the breast phantom. The three versions of the phantom are anatomical and intended for use in medical education: the basic version is for practicing primary hand-eye coordination skills; the differential one is for practicing the differential diagnosis skills; the elastographic version helps developing the skills needed for assessing the stiffness of tissues. The proposed technology allows the creation of breast phantoms for practicing hand-eye coordination and develop the critical skills for navigation and assessment of the shape, margins, and size of the lesion, as well as performing an ultrasound-guided biopsy. It is cost-effective, reproducible, and easily implementable, and could be instrumental in generating ultrasonographers with crucial skills for accurate diagnosis of breast cancer, especially in low-resource settings.

Using Pearson Correlation and Mutual Information (PC-MI) to Select Features for Accurate Breast Cancer Diagnosis Based on a Soft Voting Classifier

Breast cancer is one of the most critical diseases suffered by many people around the world, making it the most common medical risk they will face. This disease is considered the leading cause of death around the world, and early detection is difficult. In the field of healthcare, where early diagnosis based on machine learning (ML) helps save patients’ lives from the risks of diseases, better-performing diagnostic procedures are crucial. ML models have been used to improve the effectiveness of early diagnosis. In this paper, we proposed a new feature selection method that combines two filter methods, Pearson correlation and mutual information (PC-MI), to analyse the correlation amongst features and then select important features before passing them to a classification model. Our method is capable of early breast cancer prediction and depends on a soft voting classifier that combines a certain set of ML models (decision tree, logistic regression, and support vector machine) to produce one model that carries the strengths of the models that have been combined, yielding the best prediction accuracy. Our work is evaluated by using the Wisconsin Diagnostic Breast Cancer datasets. The proposed methodology outperforms previous work, achieving 99.3% accuracy, an F1 score of 0.9922, a recall of 0.9846, a precision of 1, and an AUC of 0.9923. Furthermore, the accuracy of 10-fold cross-validation is 98.2%.

Accurate breast cancer diagnosis using a stable feature ranking algorithm

BackgroundBreast cancer (BC) is one of the most common cancers among women. Since diverse features can be collected, how to stably select the powerful ones for accurate BC diagnosis remains challenging.MethodsA hybrid framework is designed for successively investigating both feature ranking (FR) stability and cancer diagnosis effectiveness. Specifically, on 4 BC datasets (BCDR-F03, WDBC, GSE10810 and GSE15852), the stability of 23 FR algorithms is evaluated via an advanced estimator (S), and the predictive power of the stable feature ranks is further tested by using different machine learning classifiers.ResultsExperimental results identify 3 algorithms achieving good stability (S ge 0.55) on the four datasets and generalized Fisher score (GFS) leading to state-of-the-art performance. Moreover, GFS ranks suggest that shape features are crucial in BC image analysis (BCDR-F03 and WDBC) and that using a few genes can well differentiate benign and malignant tumor cases (GSE10810 and GSE15852).ConclusionsThe proposed framework recognizes a stable FR algorithm for accurate BC diagnosis. Stable and effective features could deepen the understanding of BC diagnosis and related decision-making applications.

Imaging of breast cancer–beyond the basics

Imaging of breast cancer is the backbone of breast cancer screening, diagnosis, preoperative/treatment assessment and follow-up. The main modalities are mammography, ultrasound and magnetic resonance imaging, each with its own advantages and disadvantages. New emerging technologies have also enabled each modality to improve on their weaknesses. Imaging-guided biopsies have allowed for accurate diagnosis of breast cancer, with low complication rates. The purpose of this article is to review the common modalities for breast cancer imaging in current practice with emphasis on the strengths and potential weaknesses, discuss the selection of the best imaging modality for the specific clinical question or patient population, and explore new technologies / future directions of breast cancer imaging.

HER2-low breast cancer: insights on pathological testing.

Human epidermal growth factor receptor 2 (HER2) is an important biomarker for predicting prognosis and effectiveness of HER2-targeted therapy in breast cancer. The emergence of novel HER2 antibody-drug conjugate has led to a shift from the binary categorization of HER2 status (i.e., negative and positive) to a ternary categorization gradually (i.e., HER2 0, HER2 low expression and HER2 positive). The heterogeneity of HER2 low expression in breast cancer has also recently aroused widespread concern, and the heterogeneity of tumors has led to differences in the efficacy of HER2-targeted therapy. Therefore, it is crucial to accurately identify the HER2 expression status of breast cancer, which can provide a basis for patients to formulate personalized treatment strategies. In recent years, artificial intelligence (AI) has developed rapidly and been widely used in the pathological accurate diagnosis of breast cancer. The research results show that AI can significantly improve the consistency and accuracy of HER2 interpreted by pathologists in breast cancer. This has provoked many discussions on HER2-low breast cancer, such as quality control prior to HER2-low expression detection, the latest progress of tumor heterogeneity, and the application of AI. In this paper, we discuss the latest testing guidelines and advances on HER2-low breast cancer, aiming to standardize and improve the pathological testing of HER2-low breast cancer.

Deep learning-enabled fully automated pipeline system for segmentation and classification of single-mass breast lesions using contrast-enhanced mammography: a prospective, multicentre study

Breast cancer is the leading cause of cancer-related deaths in women. However, accurate diagnosis of breast cancer using medical images heavily relies on the experience of radiologists. This study aimed to develop an artificial intelligence model that diagnosed single-mass breast lesions on contrast-enhanced mammography (CEM) for assisting the diagnostic workflow. A total of 1912 women with single-mass breast lesions on CEM images before biopsy or surgery were included from June 2017 to October 2022 at three centres in China. Samples were divided into training and validation sets, internal testing set, pooled external testing set, and prospective testing set. A fully automated pipeline system (FAPS) using RefineNet and the Xception+Pyramid pooling module (PPM) was developed to perform the segmentation and classification of breast lesions. The performances of six radiologists and adjustments in Breast Imaging Reporting and Data System (BI-RADS) category 4 under the FAPS-assisted strategy were explored in pooled external and prospective testing sets. The segmentation performance was assessed using the Dice similarity coefficient (DSC), and the classification was assessed using heatmaps, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. The radiologists' reading time was recorded for comparison with the FAPS. This trial is registered with China Clinical Trial Registration Centre (ChiCTR2200063444). The FAPS-based segmentation task achieved DSCs of 0.888±0.101, 0.820±0.148 and 0.837±0.132 in the internal, pooled external and prospective testing sets, respectively. For the classification task, the FAPS achieved AUCs of 0.947 (95% confidence interval [CI]: 0.916-0.978), 0.940 (95% [CI]: 0.894-0.987) and 0.891 (95% [CI]: 0.816-0.945). It outperformed radiologists in terms of classification efficiency based on single lesions (6s vs 3min). Moreover, the FAPS-assisted strategy improved the performance of radiologists. BI-RADS category 4 in 12.4% and 13.3% of patients was adjusted in two testing sets with the assistance of FAPS, which may play an important guiding role in the selection of clinical management strategies. The FAPS based on CEM demonstrated the potential for the segmentation and classification of breast lesions, and had good generalisation ability and clinical applicability. This study was supported by the Taishan Scholar Foundation of Shandong Province of China (tsqn202211378), National Natural Science Foundation of China (82001775), Natural Science Foundation of Shandong Province of China (ZR2021MH120), and Special Fund for Breast Disease Research of Shandong Medical Association (YXH2021ZX055).

Fusion Model for Classification Performance Optimization in a Highly Imbalance Breast Cancer Dataset

Accurate diagnosis of breast cancer using automated algorithms continues to be a challenge in the literature. Although researchers have conducted a great deal of work to address this issue, no definitive answer has yet been discovered. This challenge is aggravated further by the fact that most available datasets have imbalanced class issues, meaning that the number of cases in one class vastly outnumbers those of the others. The goal of this study was to (i) develop a reliable machine-learning-based prediction model for breast cancer based on the combination of the resampling technique and the classifier, which we called a ‘fusion model’; (ii) deal with a typical high-class imbalance problem, which is posed because the breast cancer patients’ class is significantly smaller than the healthy class; and (iii) interpret the model output to understand the decision-making mechanism. In a comparative analysis with three well-known classifiers representing classical learning, ensemble learning, and deep learning, the effectiveness of the proposed machine-learning-based approach was investigated in terms of metrics related to both generalization capability and prediction accuracy. Based on the comparative analysis, the fusion model (random oversampling techniques dataset + extreme gradient boosting classifier) affects the accuracy, precision, recall, and F1-score with the highest value of 99.9%. On the other hand, for ROC evaluation, the oversampling and hybrid sampling techniques dataset combined with extreme gradient boosting achieved 100% performance compared to the models combined with the undersampling techniques dataset. Thus, the proposed predictive model based on the fusion strategy can optimize the performance of breast cancer diagnosis classification.

A Study of Breast Cancer Classification Algorithms by Fusing Machine Learning and Deep Learning

Although breast cancer, with easy recurrence and high mortality, has become one of the leading causes of cancer death in women, early and accurate diagnosis of breast cancer can effectively increase the likelihood of a cure. Therefore, it is particularly important to improve the accuracy of early diagnosis of breast cancer. However, conventional early diagnosis relies on human experience and has a low accuracy rate. Therefore, many researchers have proposed various machine learning methods to improve the accuracy and efficiency of prediction. Most of the existing studies around breast cancer classification adopt a single algorithm to fit breast cancer data but ignore the applicability of different breast cancer data features to the model. In this paper, we adopt machine algorithms to strip the features of machine learning methods from the rest of the features and attempt to enhance the model effect by designing deep learning model structures to find the hidden patterns in the rest of the features. In addition, due to strict medical data privacy requirements and high collection difficulty and cost, the model designed in this paper will be trained on a small number of samples. As a result, we attempt to find a minimization model for a breast cancer classification algorithm that features both low cost and high efficiency. At the same time, the deep learning model is further designed to complement the original model when it is possible to introduce complex data indicators. Experimental values show that the design model in this paper performs best not only under limited data and limited indicators but also under limited data complex indicators, demonstrating the effectiveness of the approach of mixed comparison and feature selection of multiple classification algorithms. In summary, the fusion model designed and implemented in this paper performs well in the experiments, and the accuracy of the model test reaches 98.3%.

HCTNet: A hybrid CNN-transformer network for breast ultrasound image segmentation

Automatic breast ultrasound image segmentation helps radiologists to improve the accuracy of breast cancer diagnosis. In recent years, the convolutional neural networks (CNNs) have achieved great success in medical image analysis. However, it exhibits limitations in modeling long-range relations, which is unfavorable for ultrasound images with speckle noise and shadows, resulting in decreased accuracy of breast lesion segmentation. Transformer can obtain sufficient global information, but it is deficient in acquiring local details and needs to be pre-trained on large-scale datasets. In this paper, we propose a Hybrid CNN-Transformer network (HCTNet) for boosting the breast lesion segmentation in ultrasound images. In the encoder of HCTNet, Transformer Encoder Blocks (TEBlocks) are designed to learn the global contextual information, which are combined with CNNs to extract features. In the decoder of HCTNet, a Spatial-wise Cross Attention (SCA) module is developed based on the spatial attention mechanism, which reduces the semantic discrepancy with the encoder. Moreover, residual connection is used between decoder blocks to make the generated features more discriminative by aggregating contextual feature maps at different semantic scales. Extensive experiments on three public breast ultrasound datasets demonstrate that HCTNet outperforms other medical image segmentation methods and the recent semantic segmentation methods on breast ultrasound lesion segmentation.

SECS: An effective CNN joint construction strategy for breast cancer histopathological image classification

Breast cancer is one of the most prevalent cancers in women. Reliable pathology identification can help histopathologists make accurate diagnosis of breast cancer but require specialized histopathological knowledge and a significant amount of manpower and medical resources. In this study, we fuse the coordinated attention mechanism to enhance the image texture analysis capability of the DenseNet, and build the CA-BreastNet model to classify microscopic histopathological images of specific types of breast cancers in the BreakHis dataset. More crucially, convolutional decision trees based on the specialized enhanced classifying strategy(SECS) are built to increase the overall accuracy of the network by reducing the model's accuracy restriction imposed by dataset structures. The related experimental results show that our network has strong performance and the SECS offers researchers reliable and effective performance enhancement guidelines. The accuracy of the convolutional decision trees reaches 99.75% for binary classification and 95.69% for eight-class classification, which means our model and strategy will be useful in the field of automatic diagnosis of breast cancer.