Diagnosis Of Breast Tumors Research Articles

MSA-GCN: A Multi-information Selection Aggregation Graph Convolutional Network for Breast Tumor Grading.

Physicians typically combine multi-modal data to make a graded diagnosis of breast tumors. However, most existing breast tumor grading methods rely solely on image information, resulting in limited accuracy in grading. This paper proposes a Multi-information Selection Aggregation Graph Convolutional Networks (MSA-GCN) for breast tumor grading. Firstly, to fully utilize phenotypic data reflecting the clinical and pathological characteristics of tumors, an automatic combination screening and weight encoder is proposed for phenotypic data, which can construct a population graph with improved structural information. Then, a graph structure is designed through similarity learning to reflect the correlation between patient image features. Finally, a multi-information selection aggregation mechanism is employed in the graph convolution model to extract the effective features of multi-modal data and enhance the classification performance of the model. The proposed method is evaluated on different clinical datasets from the Digital Database for Screening Mammography (DDSM) and INbreast. The average classification accuracies are 90.74% and 85.35%, respectively, surpassing the performance of existing methods. In conclusion, our method effectively fuses image and non-image information, leading to a significant improvement in the accuracy of breast tumor grading.

Experience in Using Breast Single-Photon Emission Computed Tomography with <sup>99m</sup>Tc-MIBI

Objective: evaluation of the results of single-photon emission computed tomography (SPECT) with technetium-99m-methoxyisobutylisonitrile (99mTc-MIBI) in women with breast masses according to mammography (MMG) and ultrasound (US) examination in comparison with the results of morphology analysis.Material and methods. SPECT with 99mTc-MIBI was performed in 11 women aged from 30 to 63 years (median age 40 [33.5; 46.5]) with BI-RADS (Breast Imaging Reporting and Data System) categories 4 or 5 mammary masses according to bilateral MMG and US diagnostics. Two patients underwent repeated SPECT and US to analyze the primary tumor response to neoadjuvant chemotherapy.Results. With a multimodal approach to the diagnosis of breast tumors, including MMG and US, 11 pathological masses were identified. SPECT made it possible to find an additional mass that had not been previously visualized. All 12 masses were subsequently verified morphologically. With benign changes, hyperfixation of 99mTc-MIBI was not observed. SPECT allowed to assess the accumulation of the radiopharmaceutical in metastatically affected lymph nodes. However, the most sensitive method for detecting the axillary lymph node lesions is US. According to the results of repeated SPECT in 2 patients, no radiopharmaceutical accumulation was noted in mammary gland masses after two courses of neoadjuvant chemotherapy, but according to the histological examination, the tumor response was different: I and IV degrees of pathomorphosis according to G.A. LavnikovaConclusion. The experience of using mammary SPECT with 99mTc-MIBI confirms an increase in the accuracy of radiologic diagnosis of newly identified breast cancer, expands the possibilities of examination in cases of ambiguous MMG and US data, as well as the contraindications or patient’s refusal from magnetic resonance imaging.

The outstanding diagnostic value of DKI in multimodal magnetic resonance imaging for benign and malignant breast tumors: A diagnostic accuracy study.

To explore the value of applying different magnetic resonance imaging MRI sequences in the differential diagnosis of benign and malignant breast tumors. Routine breast magnetic resonance scans (T1-weighted image, T1WI; T2-weighted image, T2WI), dynamically enhanced scans, diffusion-weighted Imaging, and diffusion kurtosis imaging (DKI) scans were performed on 63 female patients with breast-occupying lesions. The benign and malignant lesions were confirmed by biopsy, excision-histopathology reports. There are 70 lesions, of which 46 are benign and 24 are malignant. Analyze the primary conditions, such as the shape, size, and boundary of the lesion, and determine the apparent diffusion coefficient (ADC), mean kurtosis (MK), and mean diffusion (MD) values. The receiver operating characteristic curve was used to evaluate the value and difference in differentiating benign and malignant lesions. In this study, the results of the 2 testers both showed that the MK of malignant lesions was significantly higher than that of benign lesions (P < .001), and the MD of benign lesions was higher than that of malignant lesions (P < .05). The ADC of benign lesions was higher than that of malignant lesions (P < .05). For MK, the area under the curve of the 2 testers was 0.855/0.869, respectively. When the best cutoff value of MK for tester 1 was 0.515, the sensitivity and specificity of MK for diagnosing malignant tumors were 83.3%/87.0%, respectively. For the 2 testers MD, and ADC, the area under the curve was < 0.5, and the diagnostic value was low. The MK value obtained by DKI has a specific value in the differential diagnosis of benign and malignant breast lesions. DKI is helpful in the identification of benign and malignant breast tumors. The diagnostic value is outstanding, and its importance to the changes in the microstructure of the organization needs to be further explored.

The study of ultrasonography based on deep learning in breast cancer

ObjectiveThrough the dynamic observation, treatment, and nursing experience of patients with advanced breast cancer, we can accumulate rich clinical experience and do a better job in nursing. MethodsDeep learning radiology (DLR) extracted in-depth features from medical images through different frameworks and further analyzed the extracted deep features. The dynamic observation, treatment, and nursing of 50 patients with advanced breast cancer from hospital diagnosis provided patients with rapid, effective, and definite treatment and nursing measures. As a result, 49 patients (99%) were discharged from the hospital with improved and stable conditions. Conclusion Effective treatment and careful nursing for patients with advanced breast cancer as soon as possible will help to improve their quality of life, prolong their life cycle, reduce mortality, and assist clinical practice. ResultsCompared to traditional imaging omics, ultrasound-based DLR also needs some help, as it requires direct learning of features from data and a large dataset. Although there are more medical images than ever before, it is difficult to collect a large amount of data due to issues such as data privacy protection and complex artificial image annotation. In this comparative study, DLR can automatically extract deep features without relying on manual labeling by doctors, further improving its accuracy and reliability in tumor diagnosis and prognosis prediction. Ultrasound is the primary way of early diagnosis of breast cancer. ConclusionBy analyzing the research status of ultrasound-based DLR in the differential diagnosis of benign and malignant breast tumors, the prediction of molecular typing of breast cancer, the evaluation of axillary lymph node status, and the evaluation of neoadjuvant chemotherapy efficacy in recent years, Ultrasound-based DLR directly extracts high-order features from datasets; although it is still in the development stage, it also demonstrates its superior performance, helping to reduce workload and improve diagnostic efficiency.

An integrated network based on 2D/3D feature correlations for benign-malignant tumor classification and uncertainty estimation in digital breast tomosynthesis

Objective. Classification of benign and malignant tumors is important for the early diagnosis of breast cancer. Over the last decade, digital breast tomosynthesis (DBT) has gradually become an effective imaging modality for breast cancer diagnosis due to its ability to generate three-dimensional (3D) visualizations. However, computer-aided diagnosis (CAD) systems based on 3D images require high computational costs and time. Furthermore, there is considerable redundant information in 3D images. Most CAD systems are designed based on 2D images, which may lose the spatial depth information of tumors. In this study, we propose a 2D/3D integrated network for the diagnosis of benign and malignant breast tumors. Approach. We introduce a correlation strategy to describe feature correlations between slices in 3D volumes, corresponding to the tissue relationship and spatial depth features of tumors. The correlation strategy can be used to extract spatial features with little computational cost. In the prediction stage, 3D spatial correlation features and 2D features are both used for classification. Main results. Experimental results demonstrate that our proposed framework achieves higher accuracy and reliability than pure 2D or 3D models. Our framework has a high area under the curve of 0.88 and accuracy of 0.82. The parameter size of the feature extractor in our framework is only 35% of that of the 3D models. In reliability evaluations, our proposed model is more reliable than pure 2D or 3D models because of its effective and nonredundant features. Significance. This study successfully combines 3D spatial correlation features and 2D features for the diagnosis of benign and malignant breast tumors in DBT. In addition to high accuracy and low computational cost, our model is more reliable and can output uncertainty value. From this point of view, the proposed method has the potential to be applied in clinic.

Identification of canine mammary tumor-associated metabolites using untargeted metabolomics

The canine mammary tumor is the most common tumor type in female dogs and seriously threatens their life. Currently, no effective treatments are available for this condition. Hence, it is essential to identify biomarkers that positively influence the early diagnosis and treatment and prognosis of this disease. To provide a basis for early diagnosis of canine breast tumors, in this study, 23 dogs with mammary tumors were identified via histopathological examination combined with ancillary diagnoses via blood examinations and diagnostic imaging. The canine mammary tumor and tumor-adjacent healthy tissues were collected, and their metabolites were identified utilizing a UHPLC-qTOF-MS-based untargeted metabolomics approach. The metabolic results revealed a total of 979 ion features in the positive polarity mode and 371 ion features in the negative polarity mode in the tissues of two groups; among them, 536 differential metabolites (385 in the positive and 151 in the negative polarity mode) were analyzed by PCA and PLS-DA. Subsequently, the enrichment pathways purine metabolism, alpha-linolenic acid metabolism, vitamin B6 metabolism, and fatty acid biosynthesis were analyzed using Metaboanalyst 4.0, which suggested that these pathways were valuable diagnostic and therapeutic targets. Moreover, the receiver operating characteristic curves further confirmed 13Z,16Z-docosadienoic acid, 23-nordeoxycholic acid, and (±)12(13)-DiHOME as expected candidate biomarkers of canine mammary tumors. In conclusion, the discovery of tumor biomarkers based on untargeted metabolomics is informative for pathological mechanism studies and facilitates the early diagnosis of canine mammary tumors.

Immunohistochemistry in the Diagnosis and Classification of Breast Tumors.

In the clinical practice of breast pathology, immunohistochemistry (IHC) of different markers is widely used for the diagnosis and classification of breast lesions. To provide an overview of currently used and recently identified IHC stains that have been implemented in the field of diagnostic breast pathology. Data were obtained from literature review and clinical experience of the authors as breast pathologists. In the current review, we summarize the common uses of IHC stains for diagnosing different types of breast lesions, especially invasive and noninvasive breast lesions, and benign and malignant spindle cell lesions. In addition, the cutting-edge knowledge of diagnostic carcinoma markers will lead us to further understand the different types of breast carcinoma and differentiate breast carcinomas from other carcinomas of similar morphology. Knowing the strengths and limitations of these markers is essential to the clinical practice of breast pathology.

Breast Cancer Prediction based on Support Vector Machine

Breast cancer is the most common malignant tumor among women in the world, and its mortality rate ranks second. In this paper, the breath cancer Wisconsin (diagnostic) data set is used for prediction. First, exploratory data analysis was carried out, and 10 indicators related to breast cancer were selected as independent variables. Then, support vector machine (SVM) and logistic regression model were used as classifiers, and 75% of data sets were divided into training sets to build models. Finally, 25% of test sets were used as input models for prediction, and accuracy and recall were used as evaluation indicators to compare and analyze the advantages and disadvantages of the two algorithms. The results show that the accuracy of SVM is 16.2% higher than that of logistic regression, and the recall rate is 9.2% lower. The prediction application of SVM algorithm shows that SVM algorithm has low computational complexity and strong generalization ability. Therefore, support vector machine can be used to improve the diagnostic accuracy in the diagnosis of breast tumors.

Deep Learning-assisted Diagnosis of Breast Lesions on US Images: A Multivendor, Multicenter Study.

To evaluate the diagnostic performance of a deep learning (DL) model for breast US across four hospitals and assess its value to readers with different levels of experience. In this retrospective study, a dual attention-based convolutional neural network was built and validated to discriminate malignant tumors from benign tumors by using B-mode and color Doppler US images (n = 45 909, March 2011-August 2018), acquired with 42 types of US machines, of 9895 pathologic analysis-confirmed breast lesions in 8797 patients (27 men and 8770 women; mean age, 47 years ± 12 [SD]). With and without assistance from the DL model, three novice readers with less than 5 years of US experience and two experienced readers with 8 and 18 years of US experience, respectively, interpreted 1024 randomly selected lesions. Differences in the areas under the receiver operating characteristic curves (AUCs) were tested using the DeLong test. The DL model using both B-mode and color Doppler US images demonstrated expert-level performance at the lesion level, with an AUC of 0.94 (95% CI: 0.92, 0.95) for the internal set. In external datasets, the AUCs were 0.92 (95% CI: 0.90, 0.94) for hospital 1, 0.91 (95% CI: 0.89, 0.94) for hospital 2, and 0.96 (95% CI: 0.94, 0.98) for hospital 3. DL assistance led to improved AUCs (P < .001) for one experienced and three novice radiologists and improved interobserver agreement. The average false-positive rate was reduced by 7.6% (P = .08). The DL model may help radiologists, especially novice readers, improve accuracy and interobserver agreement of breast tumor diagnosis using US.Keywords: Ultrasound, Breast, Diagnosis, Breast Cancer, Deep Learning, Ultrasonography Supplemental material is available for this article. © RSNA, 2023.

Impact of the COVID-19 Pandemic on Histopathological Diagnosis of Breast Tumours in Calabar, Nigeria.

The Coronavirus-19 transmitted through physical contact, droplets, and fomites caused severe respiratory disease resulting in high mortality worldwide. The COVID-19 pandemic caused innumerable hardships, panic, and restrictions of movement which negatively affected the assessment of healthcare services like breast cancer diagnosis in many countries. The results from the histopathological diagnosis of breast tumours have been routinely employed for the treatment and management of these diseases. This study investigated the impact of the COVID-19 pandemic on the histopathological diagnosis of breast tumours in Calabar. A retrospective study of the newly diagnosed breast tumours recorded in the Histopathology Laboratory register during the COVID-19 and the post-COVID-19 recovery from January 2020-February 2021 was compared with cases diagnosed before the pandemic from January 2018 to February 2019. Descriptive and inferential statistics and the Artificial Neural Network (ANN) of Statistical Package for Social Sciences (SPSS) were used for data analysis. New breast tumours diagnosed based on month showed low rates of 2.4% and 1.2% during the first and second waves of the pandemic respectively. The diagnosed cases increased to 11.8% and 8.2% after the first and second waves of the virus respectively. There was a significantly strong negative correlation between the COVID-19 pandemic and lockdown measures with breast tumour diagnosis (r=-0.919, p=0.001). More benign tumours of 56(58.3%) cases with a mean age of 25.3±11.1 years were recorded before the pandemic and were statistically significant (F=64.260, p=0.004). More malignant cases of 48(57.1%) with a mean age of 47.5±11.7 years were recorded during the pandemic. The diagnosis of malignant tumours was statistically significant between both periods (F=183.550, p=0.001). The ANN model predicted a 25% reduction in breast tumour diagnosis during the pandemic. There was a 100% impact of the pandemic on tumour type, the nature of specimen, and mean age of subjects. The COVID-19 pandemic disrupted the assessment of healthcare services as a smaller number of women were diagnosed with breast tumours during the period. This may have caused delays and late presentation leading to the diagnosis of more malignant tumours. There is a need to put adequate measures to encourage the assessment of diagnostic services during pandemics as delays may lead to an increase in morbidity and mortality.

The value of contrast-enhanced ultrasound in the diagnosis of BI-RADS-US 4a lesions less than 2 cm in diameter.

Breast cancer is the most common malignant tumor in women. Early diagnosis of benign and malignant breast tumors is of great significance. To retrospectively analyze the value of contrast-enhanced ultrasonography (CEUS) in the diagnosis of Breast Imaging-Reporting and Data System (BI-RADS) 4a breast lesions less than 2 cm in diameter. CEUS was performed for 143 breast masses less than 2 cm in diameter that were diagnosed as BI-RADS 4a by ultrasound and reclassified. Considering pathological diagnosis as the gold standard, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of reclassified lesions after CEUS for the diagnosis of benign and malignant masses were analyzed. BI-RADS 4a breast masses with a diameter less than 2 cm (n = 143) were confirmed by pathology; 103 and 40 were classified as benign and malignant, respectively. The sensitivity, specificity, PPV, and NPV of CEUS for the diagnosis were 90%, 86%, 72%, and 95%, respectively. The area under the receiver operating characteristic (ROC) curve of CEUS for the diagnosis of benign and malignant tumors after CEUS was 0.904. CEUS can help to improve the diagnostic accuracy of BI-RADS 4a masses with a diameter less than 2 cm.

Signet ring cell-like diffuse large B-cell lymphoma involving the breast: a case report

BackgroundDiffuse large B-cell lymphoma (DLBCL) with signet ring cell components is extremely rare. Here, we present a case of DLBCL with signet ring cell components involving the breast, which can be easily confused with invasive lobular carcinoma of the breast or metastatic signet ring cell carcinoma of gastrointestinal origin.Case presentationA 66-year-old woman presented with a painless mass in her left breast. Enhanced magnetic resonance imaging (MRI) of the breast revealed a 42 × 29 × 28 mm mass in the left breast. Histological examination revealed a diffuse or scattered arrangement of round cells mixed with signet ring-like cells. Immunohistochemically, the neoplastic cells were positive for PAX-5, CD79a, CD20, Bcl-6, and MUM-1 but and negative for cytokeratin, ER, PR, E-cadherin, and P120. The Ki-67 proliferation index was approximately 70%. Fluorescence in situ hybridisation (FISH) demonstrated non-rearrangement of Bcl-2, Bcl-6, and c-MYC genes. Immunohistochemistry and FISH examination confirmed the diagnosis of DLBCL. Subsequently, immunofluorescence showed both IgM and IgG deposits in the signet ring-like lymphocytes. After confirming the diagnosis, the patient received four courses of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisolone) chemotherapy in a specialist hospital and achieved partial remission; however, she unfortunately died of secondary pneumocystis pneumonia infection 3 months later.ConclusionMalignant lymphoma with signet ring cell morphology is quite uncommon, and this variant can be a diagnostic pitfall. We emphasise that pathologists should consider lymphoma in the differential diagnosis of malignant breast tumours.

Secondary Breast Malignancy from Renal Cell Carcinoma: Challenges in Diagnosis and Treatment—Case Report

Renal cell carcinoma represents about 2% of all malignant tumours in adults. Metastases of the primary tumour in the breast make up to about 0.5-2% of the cases. Renal cell carcinoma metastases in the breast are extremely rare and have been sporadically recorded in the literature. In this paper, we present the case of a patient with breast metastasis of renal cell carcinoma 11 years after primary treatment. Case presentation: An 82-year-old female who had right nephrectomy due to renal cancer in 2010 felt a lump in her right breast in August 2021, whereby a clinical examination revealed a tumour at the junction of the upper quadrants of her right breast, about 2 cm, movable toward the base, vaguely limited, and with a rough surface. The axillae were without palpable lymph nodes. Mammography showed a circular and relatively clearly contoured lesion in the right breast. Ultrasound showed an oval lobulated lesion of 19 × 18 mm at the upper quadrants, with strong vascularisation and without posterior acoustic phenomena. A core needle biopsy was performed, and the histopathological findings and obtained immunophenotype indicated a metastatic clear cell carcinoma of renal origin. A metastasectomy was performed. Histopathologically, the tumour was without desmoplastic stroma, comprising predominantly solid-type alveolar arrangements of large moderately polymorphic cells, bright and abundant cytoplasm, and round vesicular cores with focally prominent nuclei. Immunohistochemically, tumour cells were diffusely positive for CD10, EMA, and vimentin, and negative for CK7, TTF-1, renal cell antigen, and E-cadherin. With a normal postoperative course, the patient was discharged on the third postoperative day. After 17 months, there were no new signs of the underlying disease spreading at regular follow-ups. Conclusion: Metastatic involvement of the breast is relatively rare and should be suspected in patients with a prior history of other cancers. Core needle biopsy and pathohistological analysis are required for the diagnosis of breast tumours.

A Novel Nomogram Based on Imaging Biomarkers of Shear Wave Elastography, Angio Planewave Ultrasensitive Imaging, and Conventional Ultrasound for Preoperative Prediction of Malignancy in Patients with Breast Lesions.

Several studies have demonstrated the difficulties in distinguishing malignant lesions of the breast from benign lesions owing to overlapping morphological features on ultrasound. Consequently, we aimed to develop a nomogram based on shear wave elastography (SWE), Angio Planewave Ultrasensitive imaging (Angio PLUS (AP)), and conventional ultrasound imaging biomarkers to predict malignancy in patients with breast lesions. This prospective study included 117 female patients with suspicious lesions of the breast. Features of lesions were extracted from SWE, AP, and conventional ultrasound images. The least absolute shrinkage and selection operator (Lasso) algorithms were used to select breast cancer-related imaging biomarkers, and a nomogram was developed based on six of the 16 imaging biomarkers. This model exhibited good discrimination (area under the receiver operating characteristic curve (AUC): 0.969; 95% confidence interval (CI): 0.928, 0.989) between malignant and benign breast lesions. Moreover, the nomogram also showed demonstrated good calibration and clinical usefulness. In conclusion, our nomogram can be a potentially useful tool for individually-tailored diagnosis of breast tumors in clinical practice.

HISTOPATHOLOGICAL STUDY OF NEOPLASTIC BREAST LESIONS IN A TERTIARY CARE CENTRE.

We have studied 194 cases of neoplastic breast lesions over a period of 2 years (1 Jan 2020 to 31 December 2021) in MGM Medical college and Hospital, Aurangabad. The specimens were received in histopathology section of pathology department. Specimens were received in different forms such as Lumpectomy, trucut biopsy, simple mastectomy and MRM. Detailed gross examination of specimens (grossing of MRM: - according to AJCC protocol) followed by xation, thorough sampling and tissue processing. The different lesions were studied by histopathological examination and analysed. Neoplastic lesions were classied according to WHO classication (2017).Out of 194 cases, 75 cases had benign breast tumours and 119 had malignant tumours. Fibroadenoma (68 cases) was most common benign tumour followed by benign phyllodes (4cases), lactating adenoma (2 cases) and multiple ductal papilloma(1 case). Invasive carcinoma with no special type (104 cases) was most common malignant tumour. Special subtypes of invasive carcinoma found in our study were:- Invasive metaplastic carcinoma (1 case) Invasive lobular carcinomas (2cases), Invasive papillary carcinoma (1 case), Invasive mucinous adenocarcinoma (1 case), Solid papillary carcinoma (1 case), Invasive carcinoma with mucinous component (1 case) , Invasive carcinoma with neuroendocrine features (1 case) and Intracystic papillary carcinoma (2cases). Other malignant tumours included malignant phyllodes (2 cases) and malignant spindle cell neoplasm (3 cases). In our study, we found that malignant neoplasms were more common than benign neoplasms with Invasive breast carcinoma with no special type being the commonest malignant tumour. Histopathological study plays an important role in diagnosis of breast neoplasms and hence in treatment and prognosis.

Breast Tumor Classification in Ultrasound Images by Fusion of Deep Convolutional Neural Network and Shallow LBP Feature.

Breast cancer is one of the most dangerous and common cancers in women which leads to a major research topic in medical science. To assist physicians in pre-screening for breast cancer to reduce unnecessary biopsies, breast ultrasound and computer-aided diagnosis (CAD) have been used to distinguish between benign and malignant tumors. In this study, we proposed a CAD system for tumor diagnosis using a multi-channel fusion method and feature extraction structure based on multi-feature fusion on breast ultrasound (BUS) images. In the pre-processing stage, the multi-channel fusion method completed the color conversion of the BUS image to make it contain richer information. In the feature extraction stage, the pre-trained ResNet50 network was selected as the basic network, and three levels of features were combined based on adaptive spatial feature fusion (ASFF), and finally, the shallow local binary pattern (LBP) texture features were fused. Support vector machine (SVM) was used for comparative analysis. A retrospective analysis was carried out, and 1615 breast tumor images (572 benign and 1043 malignant) confirmed by pathological examinations were collected. After data processing and augmentation, for an independent test set consisting of 874 breast ultrasound images (457 benign and 417 malignant), the accuracy, precision, recall, specificity, F1 score, and AUC of our method were 96.91%, 98.75%, 94.72%, 98.91%, 0.97, and 0.991, respectively. The results show that the integration of shallow LBP texture features and multi-level depth features can more effectively improve the comprehensive performance of breast tumor diagnosis, and has strong clinical application value. Compared with the past methods, our proposed method is expected to realize the automatic diagnosis of breast tumors and provide an auxiliary tool for radiologists to accurately diagnose breast diseases.

Pre-Trained Deep Neural Network-Based Computer-Aided Breast Tumor Diagnosis Using ROI Structures

Deep neural network (DNN) based computer-aided breast tumor diagnosis (CABTD) method plays a vital role in the early detection and diagnosis of breast tumors. However, a Brightness mode (B-mode) ultrasound image derives training feature samples that make closer isolation toward the infection part. Hence, it is expensive due to a meta-heuristic search of features occupying the global region of interest (ROI) structures of input images. Thus, it may lead to the high computational complexity of the pre-trained DNN-based CABTD method. This paper proposes a novel ensemble pre-trained DNN-based CABTD method using global- and local-ROI-structures of B-mode ultrasound images. It conveys the additional consideration of a local-ROI-structures for further enhancing the pre-trained DNN-based CABTD method’s breast tumor diagnostic performance without degrading its visual quality. The features are extracted at various depths (18, 50, and 101) from the global and local ROI structures and feed to support vector machine for better classification. From the experimental results, it has been observed that the combined local and global ROI structure of small depth residual network ResNet18 (0.8 in %) has produced significant improvement in pixel ratio as compared to ResNet50 (0.5 in %) and ResNet101 (0.3 in %), respectively. Subsequently, the pre-trained DNN-based CABTD methods have been tested by influencing local and global ROI structures to diagnose two specific breast tumors (Benign and Malignant) and improve the diagnostic accuracy (86%) compared to Dense Net, Alex Net, VGG Net, and Google Net. Moreover, it reduces the computational complexity due to the small depth residual network ResNet18, respectively.

Ultrasonic Diagnosis of Breast Tumors

Ultrasonic Diagnosis of Breast Tumors

Portable Microwave-Acoustic Coaxial Thermoacoustic Probe With Miniaturized Vivaldi Antennas for Breast Tumor Screening.

Microwave-induced thermoacoustic (TA) imaging (MTAI), which exploits dielectric contrasts to provide images with high contrast and spatial resolution, holds the potential to serve as an additional means of clinical diagnosis and treatment. However, conventional MTAI usually uses large and heavy metal antennas to radiate pulsed microwaves, making it challenging to image different target areas flexibly. In this work, we presented the design and evaluation of a portable microwave-acoustic coaxial TA probe (51 mm × 63 mm × 138 mm) that can flexibly image the region of interest. The TA probe contains two miniaturized symmetrically distributed Vivaldi antennas (7.5 g) and a 128-element linear ultrasonic transducer. By adjusting the geometry of the antennas and the ultrasonic transducer, the TA probe's acoustic field and microwave field can be designed to be coaxial, which helps achieve homogeneous microwave illumination and high-sensitivity ultrasonic detection. The practical feasibility of the proposed probe was tested on an in vitro ewe breast and a healthy volunteer. The results demonstrate that the MTAI system with the proposed TA probe can visualize the anatomical structure of the breast tumor in ewe breast and a healthy volunteer breast with resolutions in hundreds of microns (transverse: 910 μm, axial: 780 μm) and an excellent signal-to-noise ratio can be obtained in deep adipose tissue (10 dB in 6 cm fat). The miniaturized portable TA probe takes a solid step forward in translating MTAI technology to clinical breast tumor diagnosis.

СЛУЧАЙ НЕХОДЖКИНСКОЙ ЛИМФОМЫ МОЛОЧНОЙ ЖЕЛЕЗЫ

Abstract. Primary breast lymphoma is an unusual clinical entity accounting for 0.05-0.53% of all breast neoplasms. The usual presentation includes a painless palpable mass similar to that of breast carcinoma. Diffuse large B-cell lymphoma is the most common identifiable type of plasmablastic lymphoma based on the histopathological examination. We report an unusual case of a 47-year-old Belarusian woman who presented with a 3-month history of a painless mass in the left breast. A biopsy revealed diffuse infiltration of lymphoid cells. The immunohistochemical biomarkers confirmed the diagnosis of a non-Hodgkin's diffuse large B-cell lymphoma. After completing the chemotherapy, the patient is on follow-up, and at 48 months she is alive without disease activity. This case demonstrates the importance of histological verification for treatment. An integrated approach, including the analysis of clinical and anamnestic data, the results of histological and immunohistochemical analysis, allows for modern differential diagnosis of rare breast tumors.