-
Asian Pacific Journal of Cancer... Feb 2023Early detection and precise diagnosis of breast cancer (BC) plays an essential part in enhancing the diagnosis and improving the breast cancer survival rate of patients...
OBJECTIVE
Early detection and precise diagnosis of breast cancer (BC) plays an essential part in enhancing the diagnosis and improving the breast cancer survival rate of patients from 30 to 50%. Through the advances of technology in healthcare, deep learning takes a significant role in handling and inspecting a great number of X-ray, MRI, CTR images. The aim of this study is to propose a deep learning model (BCCNN) to detect and classify breast cancers into eight classes: benign adenosis (BA), benign fibroadenoma (BF), benign phyllodes tumor (BPT), benign tubular adenoma (BTA), malignant ductal carcinoma (MDC), malignant lobular carcinoma (MLC), malignant mucinous carcinoma (MMC), and malignant papillary carcinoma (MPC).
METHODS
Breast cancer MRI images were classified into BA, BF, BPT, BTA, MDC, MLC, MMC, and MPC using a proposed Deep Learning model with additional 5 fine-tuned Deep learning models consisting of Xception, InceptionV3, VGG16, MobileNet and ResNet50 trained on ImageNet database. The dataset was collected from Kaggle depository for breast cancer detection and classification. That Dataset was boosted using GAN technique. The images in the dataset have 4 magnifications (40X, 100X, 200X, 400X, and Complete Dataset). Thus we evaluated the proposed Deep Learning model and 5 pre-trained models using each dataset individually. That means we carried out a total of 30 experiments. The measurement that was used in the evaluation of all models includes: F1-score, recall, precision, accuracy.
RESULTS
The classification F1-score accuracies of Xception, InceptionV3, ResNet50, VGG16, MobileNet, and Proposed Model (BCCNN) were 97.54%, 95.33%, 98.14%, 97.67%, 93.98%, and 98.28%, respectively.
CONCLUSION
Dataset Boosting, preprocessing and balancing played a good role in enhancing the detection and classification of breast cancer of the proposed model (BCCNN) and the fine-tuned pre-trained models' accuracies greatly. The best accuracies were attained when the 400X magnification of the MRI images due to their high images resolution.
Topics: Humans; Female; Breast Neoplasms; Deep Learning; Carcinoma, Ductal, Breast; Carcinoma, Lobular; Carcinoma, Papillary; Fibroadenoma; Fibroma
PubMed: 36853302
DOI: 10.31557/APJCP.2023.24.2.531 -
The Bulletin of Tokyo Dental College 2018Sclerosing polycystic adenosis (SPA) is a rare benign lesion of the salivary glands which appears histologically similar to sclerosing adenosis and fibrocystic disease...
Sclerosing polycystic adenosis (SPA) is a rare benign lesion of the salivary glands which appears histologically similar to sclerosing adenosis and fibrocystic disease of the mammary gland. To date, 67 cases of SPA have been reported in the literature, with the lesion arising in the minor salivary glands in only 9. The present report describes the 10th case of SPA. The patient was a 39-year-old Brazilian man who presented with an asymptomatic nodule on the ventral surface of the tongue. Based on a clinical diagnosis of benign salivary gland neoplasm, an excisional biopsy of the lesion was performed. Histopathological examination showed lobular proliferation of ductal and acinar elements surrounded by a fibrosclerotic stroma. Many of the ductal structures exhibited cystic dilatation and were surrounded by periductal fibrosis, which is consistent with SPA findings. No recurrence of the disease was observed after a 5-year follow-up. A literature review is also discussed, focusing on both the etiology of SPA and the treatment options available.
Topics: Adult; Cysts; Fibrosis; Follow-Up Studies; Humans; Male; Salivary Gland Diseases; Salivary Gland Neoplasms; Salivary Glands, Minor; Sclerosis; Tongue Diseases
PubMed: 29962419
DOI: 10.2209/tdcpublication.2017-0029 -
Radiologia 2020Triple-negative tumors are the most aggressive type of breast cancer. We aimed to analyze the main radiologic and histopathologic factors of these tumors to create a... (Observational Study)
Observational Study
OBJECTIVE
Triple-negative tumors are the most aggressive type of breast cancer. We aimed to analyze the main radiologic and histopathologic factors of these tumors to create a risk profile.
MATERIALS AND METHODS
We analyzed data from 140 patients diagnosed with triple-negative breast cancer between January 2007 and December 2016, with follow-up through April 2018. We analyzed the following variables in the breast MRI done for staging: size, necrosis, associated findings, adenopathies, and perfusion and diffusion parameters. We analyzed the following variables in histopathologic studies of biopsy specimens: histological type, Scarf-Bloom, Ki67, and p53 in the infiltrating component as well as in the in situ component. We analyzed the following variables in histopathologic studies of positive lymph nodes and surgical specimens: size, lymphovascular/perineural invasion, and microglandular adenosis. We analyzed the relation between the radiologic and histopathologic factors and recurrence and disease-free survival.
RESULTS
MRI tumor size>25mm, non-nodular enhancement, breast edema, areola-nipple complex retraction, and lymph-node involvement were associated with recurrence and lower disease-free survival. Invasive lobular carcinoma, postsurgical size>20mm, and p53<15% were also associated with recurrence and lower disease-free survival. Histologically positive lymph nodes were associated with a greater percentage of recurrence and lymphovascular invasion and with lower disease-free survival. The multivariate analysis found that the variables MRI size>25mm, non-nodular enhancement, adenopathies on MRI, and p53 expression <15% were independent predictors of lower disease-free survival.
CONCLUSIONS
In triple-negative breast tumors, factors associated with lower disease-free survival are non-nodular enhancement, size>25mm, and adenopathies on MRI, and p53 expression <15% on histopathologic study.
Topics: Adult; Aged; Disease-Free Survival; Female; Humans; Magnetic Resonance Imaging; Middle Aged; Retrospective Studies; Risk Assessment; Triple Negative Breast Neoplasms
PubMed: 32093905
DOI: 10.1016/j.rx.2020.01.001 -
Surgical Case Reports May 2023Neurofibroma of the breast is extremely rare, with only a few reported cases. Here, we report a case of solitary neurofibroma of the breast in a 95-year-old woman.
BACKGROUND
Neurofibroma of the breast is extremely rare, with only a few reported cases. Here, we report a case of solitary neurofibroma of the breast in a 95-year-old woman.
CASE PRESENTATION
A 95-year-old woman presented with a palpable mass in the left breast. Mammography revealed a well-defined mass. A 1.6-cm round mass was found in the lower outer quadrant of the left breast on ultrasonography. The internal echo of the tumor was a mixture of relatively uniform hypoechoic areas with posterior enhancement and heterogeneous hyperechoic areas. She underwent a core needle biopsy. The pathological findings revealed a spindle cell lesion with no malignant findings. At 2 months follow-up, repeat breast ultrasonography showed that the mass had enlarged to be 2.7 cm in size. A repeat core needle biopsy, however, revealed no particularly new information. Because the tumor was growing and a definite diagnosis was not made, lumpectomy was performed. We found bland-spindled cells with shredded-carrot collagen bundles. Immunohistochemical antibody markers (S100, SOX10, and CD34) were positive for the spindle cells. Some of the tumors maintained the bilayer nature of luminal cells and myoepithelial cells, which might be the reason for internal heterogeneity on ultrasound. A histological diagnosis of neurofibroma with adenosis was made. At 6 months follow-up, no recurrent lesions were found.
CONCLUSIONS
Ultrasound and pathological images revealed an extremely rare case of neurofibroma combined with adenosis. Tumor resection was performed because it was difficult to make a definitive diagnosis using needle biopsy. Even when a benign tumor is suspected, short-term follow-up is necessary, and if an enlargement is observed, early tumor resection is recommended.
PubMed: 37246203
DOI: 10.1186/s40792-023-01673-0 -
Diagnostic Pathology Jun 2016Enteric-type glandular lesions are extremely rare in the vagina. Their histological origin remains a matter of speculation at present.
BACKGROUND
Enteric-type glandular lesions are extremely rare in the vagina. Their histological origin remains a matter of speculation at present.
METHOD
We review two rectal mucosal prolapse-like polyps and one intestinal-type adenosis in the vagina.
RESULTS
Case 1, a 64-year-old woman, presented with a vaginal polypoid lesion with a size of 4 × 3 × 3 cm. Case 2, an 8-year-old girl, had a 1.5 × 1.5 × 0.8-cm pedunculated polyp in the vaginal navicular fossa and a clinically suspected rectovaginal fistula. Case 1 and 3 had an obsolete severe perineal laceration. On histopathological examination, cases 1 and 2 resembled rectal mucosal prolapse or inflammatory cloacogenic polyp (rectal mucosal prolapse-like polyp). Case 3 had an incidental intestinal-type adenosis in the removed vaginal wall. Immunohistochemistry confirmed the intestinal differentiation in all 3 lesions by showing diffuse CDX2-positive, CK20-positive, and scattered chromogranin A-positive neuroendocrinal cells in the lower compartment of the crypt.
CONCLUSIONS
In summary, we report herein three unusual cases of benign intestinal-type glandular lesions in the vagina including two rectal mucosal prolapse-like polyps and one case of intestinal-type adenosis, and discuss possibilities for their histogenetic basis.
Topics: Biomarkers, Tumor; CDX2 Transcription Factor; Cell Differentiation; Child; Chromogranin A; Female; Humans; Immunohistochemistry; Intestinal Mucosa; Keratin-20; Middle Aged; Neoplasms, Glandular and Epithelial; Polyps; Rectal Prolapse; Treatment Outcome; Vagina; Vaginal Neoplasms
PubMed: 27315791
DOI: 10.1186/s13000-016-0503-5 -
Medicine Dec 2015Sclerosing adenosis (SA) is a less common histopathological lesion of the breast that can coexist with proliferative lesions as well as malignancies. We aimed to analyze... (Observational Study)
Observational Study
Sclerosing adenosis (SA) is a less common histopathological lesion of the breast that can coexist with proliferative lesions as well as malignancies. We aimed to analyze the clinicopathological characteristics of SA and to investigate the radiological features of SA.Patients who underwent breast surgery at our institute from 2007 to 2013 were retrospectively reviewed. A total of 815 breasts (722 patients) were included in the final analysis. Synchronous bilateral SA was defined as the detection of another SA arising in the contralateral breast within 1 month after surgery for the initial breast lesion. Baseline characteristics, imaging records (ultrasonography, mammography, and magnetic resonance imaging [MRI]), and pathology were included in the analysis.The median age at diagnosis was 47 years old. The majority of patients had unilateral non-Bc-SA (457/722). Among 102 patients with bilateral SA, 78.4% were diagnosed synchronously. In total, 26 patients suffered from synchronous bilateral breast cancer. Upon final pathological investigation, 226 cases were SA involving breast cancer (Bc-SA), most (56.2%) of which were ductal carcinoma in situ (DCIS). In addition, lobular carcinoma in situ (LCIS) and diseases that involved LCIS also comprised up to 11.1% of cases. The majority of SA cases (405; 49.7%) had no obvious symptoms except for imaging changes in mammography or ultrasound. Compared with non-Bc-SA cases, Bc-SA cases were more likely to exhibit features of mass (32.8% vs. 28.6%) and architectural distortion (20.4% vs. 13.0%) on mammography. Ultrasonography, mammography, and MRI revealed unsatisfactory sensitivity and specificity to differentiate Bc-SA from non-Bc-SA. MRI exhibited the highest sensitivity and lowest specificity, whereas the specificity of mammography was as low as 50.0%.A tendency for synchronous bilaterality in both Bc-SA and non-Bc-SA was noted. DCIS was the most commonly observed malignancy involved in Bc-SA. Although most patients with SA were asymptomatic, the ability of imaging studies to accurately differentiate non-Bc-SA from Bc-SA remained unsatisfactory.
Topics: Adult; Breast Neoplasms; Carcinoma, Intraductal, Noninfiltrating; Carcinoma, Lobular; China; Diagnostic Imaging; Female; Histiocytoma, Benign Fibrous; Humans; Magnetic Resonance Imaging; Mammography; Middle Aged; Retrospective Studies; Sensitivity and Specificity; Ultrasonography, Mammary
PubMed: 26656378
DOI: 10.1097/MD.0000000000002298 -
Gynecologic Oncology Reports Nov 2020•Vaginal adenosis is a non-obligate pre-cursor for vaginal clear cell carcinoma.•Vaginal adenosis is rare and presents with a variety of signs and...
•Vaginal adenosis is a non-obligate pre-cursor for vaginal clear cell carcinoma.•Vaginal adenosis is rare and presents with a variety of signs and symptoms.•Unclear link between adenosis and carcinoma without diethylstilbestrol exposure.•Surveillance with physical examinations, imaging and biopsies is recommended.
PubMed: 33294576
DOI: 10.1016/j.gore.2020.100672 -
Journal of Thoracic Disease Dec 2019Deep learning-based computer-aided diagnosis (CAD) is an important method in aiding diagnosis for radiologists. We investigated the accuracy of a deep learning-based CAD...
BACKGROUND
Deep learning-based computer-aided diagnosis (CAD) is an important method in aiding diagnosis for radiologists. We investigated the accuracy of a deep learning-based CAD in classifying breast lesions with different histological types.
METHODS
A total of 448 breast lesions were detected on ultrasound (US) and classified by an experienced radiologist, a resident and deep learning-based CAD respectively. The pathological results of the lesions were chosen as the golden standard. The diagnostic performances of the three raters in different pathological types were analyzed.
RESULTS
For the overall diagnostic performance, deep learning-based CAD presented a significantly higher specificity (76.96%) compared with the two radiologists. The area under ROC of CAD was almost equal with the experienced radiologist (0.81 0.81), while significantly higher than the resident (0.81 0.70, P<0.0001). In the benign lesions, deep learning-based CAD had a higher accuracy than both the two radiologists, which correctly classified as benign lesions in 119/135 of fibroadenomas (88.1%), 25/35 of adenosis (71.4%), 14/27 of intraductal papillary tumors (51.9%), 5/10 of inflammation (50%), and 4/8 of sclerosing adenosis (50%). But only the differences between CAD and the two radiologists in fibroadenomas had statistical significance (P=0.0011 and P=0.0313), and the differences between CAD and the resident in adenosis had statistical significance (P=0.012). In the malignant lesions, 151/168 of invasive ductal carcinomas (89.9%), 21/29 of ductal carcinoma in situ (DCIS) (72.4%) and 6/7 of invasive lobular carcinomas (85.7%) were diagnosed as malignancies by deep learning-based CAD, with no significant differences between CAD and the two radiologists.
CONCLUSIONS
In the diagnosis of these common types of breast lesions, deep learning-based CAD had a satisfying performance. Deep learning-based CAD had a better performance in the breast benign lesions, especially in fibroadenomas and adenosis. Therefore, deep learning-based CAD is a promising supplemental tool to US to increase the specificity and avoid unnecessary benign biopsies.
PubMed: 32030218
DOI: 10.21037/jtd.2019.12.10 -
The Journal of Pathology. Clinical... May 2021Microglandular adenosis (MGA) represents a rare neoplasm of the mammary gland, which in a subset of cases may be associated with triple-negative breast cancer (BC). The...
Microglandular adenosis (MGA) represents a rare neoplasm of the mammary gland, which in a subset of cases may be associated with triple-negative breast cancer (BC). The biology of MGA is poorly understood. In this study, eight MGA cases (n = 4 with and n = 4 without associated BC) were subjected to a comprehensive characterization using immunohistochemistry, genome-wide DNA copy number (CN) profiling, fluorescence in situ hybridization (FISH), next-generation sequencing (NGS), and DNA methylation profiling using 850 K arrays and bisulfite pyrosequencing. Median patient age was 61 years (range 57-76 years). MGA lesions were estrogen receptor (ER)-negative, progesterone receptor-negative, HER2-negative, and S100-positive. DNA CN alterations (CNAs) were complex or limited to few gains and losses. CN gain on chromosome 2q was the most common CNA and was validated by FISH in five of eight cases. NGS demonstrated an average of two mutations per case (range 0-5) affecting 10 different genes (ARID1A, ATM, CTNNB1, FBXW7, FGFR2, MET, PIK3CA, PMS2, PTEN, and TP53). CNAs and mutations were similar in MGA and adjacent BC, indicating clonal relatedness. DNA methylation profiling identified aberrant hypermethylation of CpG sites within GATA3, a key transcription factor required for luminal differentiation. Immunohistochemistry showed regular GATA3 protein expression in the normal mammary epithelium and in ER-positive BC. Conversely, GATA3 was reduced or lost in all MGA cases tested (8/8). In conclusion, MGA is characterized by common CN gain on chromosome 2q and loss of GATA3. Epigenetic inactivation of GATA3 may provide a new clue to the peculiar biology of this rare neoplasia.
Topics: Aged; Biomarkers, Tumor; Chromosome Aberrations; Chromosomes, Human, Pair 2; DNA Methylation; Female; Fibrocystic Breast Disease; GATA3 Transcription Factor; Gene Silencing; Humans; Middle Aged; Molecular Diagnostic Techniques; Precancerous Conditions; Triple Negative Breast Neoplasms
PubMed: 33382535
DOI: 10.1002/cjp2.195 -
Human Pathology Jan 2019Magnetic resonance imaging (MRI)/ultrasound fusion-targeted biopsy (TB) has been shown to more accurately identify higher-grade prostate cancers compared with...
Magnetic resonance imaging (MRI)/ultrasound fusion-targeted biopsy (TB) has been shown to more accurately identify higher-grade prostate cancers compared with standard-of-care systematic sextant prostate biopsy (SB). However, occasional false-positive imaging findings occur. We investigated the histologic findings associated with false-positive prostate MRI findings. A retrospective review was performed on our surgical pathology database from 2014 to 2017 selecting patients with no cancer detected on TB with concurrent SB after at least 1 prior benign SB session. Histologic features evaluated included percentage of core involvement by chronic inflammation, percentage of core composed of stroma, percentage of glands involved by atrophy, and presence of the following features: acute or granulomatous inflammation, stromal nodular hyperplasia, adenosis, squamous metaplasia, basal cell hyperplasia, and presence of skeletal muscle. Histologic findings were compared between TB and concurrent SB. We identified 544 patients who underwent TB. Of these, 41 patients, including 62 targeted lesions, met criteria. Compared with SB tissue, the mean percentage of stroma was increased in TB (P = .02). Basal cell hyperplasia was also found to be more common on TB (P = .02). Both high percentage of stroma (P = .046) and presence of basal cell hyperplasia (P = .038) were independent predictors on multivariate analysis. The combination of high chronic inflammation, high stroma, acute inflammation, and basal cell hyperplasia was associated with TB (P = .001). Atrophic glands and chronic inflammation showed a positive correlation (r = 0.67, P = .003), which was especially seen in high prostate imaging reporting and data system lesions. Specific benign histologic entities are associated with false-positive findings on prostate MRI.
Topics: Aged; Aged, 80 and over; False Positive Reactions; Humans; Image-Guided Biopsy; Magnetic Resonance Imaging, Interventional; Male; Middle Aged; Prostatic Neoplasms; Retrospective Studies; Ultrasonography, Interventional
PubMed: 30179687
DOI: 10.1016/j.humpath.2018.08.021