EMS-Net: Ensemble of Multiscale Convolutional Neural Networks for Classification of Breast Cancer Histology Images

Publication date: Available online 30 July 2019Source: NeurocomputingAuthor(s): Zhanbo Yang, Lingyan Ran, Shizhou Zhang, Yong Xia, Yanning ZhangAbstractHistology images analysis resulted from needle biopsy serves as the gold standard for breast cancer diagnosis. Deep learning-based classification of breast tissues in histology images, however, is less accurate, due to the lack of adequate training data and ignoring the multiscale structural and textural information. In this paper, we propose the Ensemble of MultiScale convolutional neural Networks (EMS-Net) to classify hematoxylin-eosin stained breast histopathological microscopy images into four categories, including normal tissue, benign lesion, in situ carcinoma, invasive carcinoma. We first convert each image to multiple scales, and then use the training patches cropped and augmented at each scale to fine-tune the pre-trained DenseNet-161, ResNet-152, and ResNet-101, respectively. We find that a combination of three fine-tuned models is more accurate than other combinations, and use them to form an ensemble model. We evaluated our algorithm against three recent methods on the BACH challenge dataset. It shows that the proposed EMS-Net algorithm achieved an accuracy of 91.75 ± 2.32% in the five-fold cross validation using 400 training images, which is higher than the accuracy of other three algorithms, and also achieved an accuracy of 90.00% in the online verification using 100 testing images.
Source: Neurocomputing - Category: Neuroscience Source Type: research

Related Links:

AbstractPurposeTo investigate the performance of an imaging and biopsy parameters-based multivariate model in decreasing unnecessary surgeries for high-risk breast lesions.MethodsIn an IRB-approved study, we retrospectively reviewed all high-risk lesions (HRL) identified at imaging-guided biopsy in our institution between July 1, 2014-July 1, 2017. Lesions were categorized high-risk-I (HR-I  = atypical ductal hyperplasia, atypical lobular hyperplasia, lobular carcinoma in situ and atypical papillary lesion) and II (HR-II = Flat epithelial atypia, radial scar, benign papilloma). Patient risk fact...
Source: Breast Cancer Research and Treatment - Category: Cancer & Oncology Source Type: research
Invasive breast cancer deriving from the milk duct and lobule that develops in the nipple is extremely rare, except in Paget ’s disease and skin cancer. This is the second reported case of the development of invasive cancer confined to the nipple after breast-conserving surgery. A 69-year-old woman visited our department due to redness, swelling, and bloody discharge of the right nipple in the last month. A needle biops y was suggestive of invasive ductal carcinoma; we performed a removal surgery of the right residual breast tissue and a second sentinel lymph node biopsy. She underwent these procedures 10 years previ...
Source: Case Reports in Oncology - Category: Cancer & Oncology Source Type: research
ConclusionPatients with radial scar with atypia have a higher risk of upgrade to malignancy. Further research is needed to identify which patients may safely avoid excision of radial scar. Patients with a diagnosis of radial scar on CNB are at increased subsequent risk of breast cancer and may benefit from additional screening.
Source: Breast Cancer Research and Treatment - Category: Cancer & Oncology Source Type: research
Conclusion: Cluster gray image-fractal analysis evaluating the darkness of clusters, cluster unevenness, and complexity of cluster density in breast FNAB HCG is a useful cytology assistance system for breast FNA.Acta Cytologica
Source: Acta Cytologica - Category: Cytology Source Type: research
Authors: Gilmore R, Prasath V, Habibi M Abstract Paget's disease of the breast (PDB) is a rare breast carcinoma believed to arise from an underlying in situ or invasive ductal cancer that migrates through the epidermis causing characteristic skin changes including scaling, redness, and itching of the nipple, areola, and sometimes the surrounding skin. Although Paget's may mimic benign conditions such as contact or allergic eczema and mastitis, it should remain a strong consideration in the differential diagnosis , especially in peripartum women for whom benign conditions such as bacterial mastitis from breastfeedin...
Source: Advances in Experimental Medicine and Biology - Category: Research Tags: Adv Exp Med Biol Source Type: research
CONCLUSION: In our cohort, expectant management of FEA alone appears to be a safe option as our upstaging rate to DCIS or invasive cancer for FEA diagnosed on core biopsy was only 4%. Our study suggests that close follow-up is a safe and feasible option for pure FEA without a radiographic discordance found on core biopsy. PMID: 32816560 [PubMed - as supplied by publisher]
Source: The American Surgeon - Category: Surgery Authors: Tags: Am Surg Source Type: research
AbstractAlveolar soft part sarcoma (ASPS) is a rare soft tissue tumor of unknown histogenesis generally characterized by the der(17)t(X;17)(p11.2;q25) translocation which results in theASPSCR1-TFE3 gene fusion. Primary ASPS of the thyroid gland has not yet been reported. During oncology follow-up for breast cancer, a pulmonary nodule and thyroid gland mass  were identified in a 71-year-old Korean male. Thyroid ultrasound showed a 5.7 cm left thyroid gland mass. After several fine needle aspirations, a thyroid gland lobectomy was performed after documenting only non-caseating granulomatous inflammation in a biopsy...
Source: Head and Neck Pathology - Category: Pathology Source Type: research
Conclusion: The combination of acellular dermal matrix and oxidized regenerated cellulose for partial breast reconstruction can be a good option based on oncological safety and cosmetic outcome.
Source: Medicine - Category: Internal Medicine Tags: Research Article: Clinical Case Report Source Type: research
ConclusionsValidation of the SNUH BCS nomogram was not successful in the current study as much as its original publication. However, we could improve its predictive power by including surgeon-related factor. Before applying a published nomogram as a preoperative predictive model, we suggest each institution to validate the model and adjust it with surgeon-related factor. Addition of new factors to currently available nomograms holds promise for improving its applicability for breast cancer patients at the actual clinical level.
Source: Breast Cancer Research and Treatment - Category: Cancer & Oncology Source Type: research
AbstractPurposeThis proof-of-concept study investigates gene expression in core needle biopsies (CNB) to predict whether individuals diagnosed with ductal carcinoma in situ (DCIS) on CNB were affected by invasion at the time of diagnosis.MethodsUsing a QuantiGene Plex 2.0 assay, 14 gene expression profiling was performed in 303 breast tissue samples. Preoperative diagnostic performance of a gene was measured by area under receiver-operating characteristic curve (AUC) with 95% confidence interval (CI). The gene mRNA positivity cutoff was computed using Gaussian mixture model (GMM); protein expression was measured by immunoh...
Source: Breast Cancer Research and Treatment - Category: Cancer & Oncology Source Type: research
More News: Brain | Breast Cancer | Breast Carcinoma | Breast Needle Aspiration | Cancer | Cancer & Oncology | Carcinoma | Learning | Neurology | Neuroscience | Training | Universities & Medical Training