OBJECTIVE: To develop a deep features-based model to classify benign and malignant breast lesions on full- filed digital mammography. METHODS: The data of full-filed digital mammography in both craniocaudal view and mediolateral oblique view from 106 patients with breast neoplasms were analyzed. Twenty-three handcrafted features (HCF) were extracted from the images of the breast tumors and a suitable feature set of HCF was selected using -test. The deep features (DF) were extracted from the 3 pre-trained deep learning models, namely AlexNet, VGG16 and GoogLeNet. With abundant breast tumor information from the craniocaudal view and mediolateral oblique view, we combined the two extracted features (DF and HCF) as the two-view features. A multi-classifier model was finally constructed based on the combined HCF and DF sets. The classification ability of different deep learning networks was evaluated. RESULTS: Quantitative evaluation results showed that the proposed HCF+DF model outperformed HCF model, and AlexNet produced the best performances among the 3 deep learning models. CONCLUSIONS The proposed model that combines DF and HCF sets of breast tumors can effectively distinguish benign and malignant breast lesions on full-filed digital mammography.
Breast Neoplasms ; classification ; diagnostic imaging ; Deep Learning ; Diagnosis, Computer-Assisted ; methods ; Female ; Humans ; Mammography ; methods

PURPOSE: The purpose of this study was to estimate the clinical utilization of radiation therapy (RT) in Korea between 2011 and 2015. MATERIALS AND METHODS: We analyzed the claims data from the Health Insurance Review and Assessment Service to estimate the clinical utilization of RT. The source population consisted of all patients who had any of the International Classification of Diseases 10th revision cancer diagnoses (C00-C97) and those with diagnostic codes D00-D48, who were also associated with at least one of the procedure codes related to RT. RESULTS: The total number of patients who received RT in 2011, 2012, 2013, 2014, and 2015 were 54,810, 59,435, 61,839, 64,062, and 66,183, respectively. Among them, the total numbers of male and female patients were 24,946/29,864 in 2011, 27,211/32,224 in 2012, 28,111/33,728 in 2013, 29,312/34,750 in 2014, and 30,266/35,917 in 2015. The utilization rate of RT in cancer patients has also increased steadily over the same period from 25% to 30%. The five cancers that were most frequently treated with RT between 2011 and 2012 were breast, lung, colorectal, liver, and uterine cervical cancers. However, the fifth most common cancer treated with RT that replaced uterine cervical cancer in 2013 was prostate cancer. More than half of cancer patients (64%) were treated with RT in the capital area (Seoul, Gyeonggi, and Incheon). CONCLUSION: The total number of patients who underwent RT increased steadily from 2011 to 2015 in Korea. The utilization rate of RT in cancer patients is also increasing.
Breast ; Diagnosis ; Female ; Gyeonggi-do ; Humans ; Insurance, Health ; International Classification of Diseases ; Korea* ; Liver ; Lung ; Male ; Prostatic Neoplasms ; Radiotherapy ; Uterine Cervical Neoplasms

OBJECTIVE: Among the various psychiatric disorders, depression is a common comorbid condition in cancer patients. Due to the distinct and diverse nature of cancer, the prevalence of depression may be assumed to be affected by differences in primary cancer sites. In this study, we explored the prevalence rates of depression among the ten most prevalent cancers in South Korea using a national patient sample. METHODS: This was a 1-year cross-sectional study using a national patient sample provided by the South Korean National Health Insurance in 2011. We selected all patients who had received ICD-10 codes of the 10 most prevalent cancers and major depressive disorder. Afterwards, the cancer and depression groups were merged and analyzed. RESULTS: The MDD prevalence rate was highest in lung cancer (11.0%), followed by non-Hodgkin's lymphoma (9.2%), prostate (9.1%), bladder (8.8%), breast (7.8%), cervix (7.8%), colorectal (7.7%), stomach (6.9%), liver (6.5%), and thyroid cancer (5.6%). Within all cancer groups, patients with a MDD diagnosis were significantly older (p<0.05) than non-MDD patients. Colorectal, stomach, and thyroid cancer displayed a higher female proportion in the MDD group than the non-MDD group. In the subgroup analysis, the prevalence rate differed by age and sex. CONCLUSION: The prevalence of depression varied according to cancer types. Patients with lung cancer were the most prone to experience depression. Because clinical and psychological factors may influence MDD prevalence, these factors will need to be studied more closely in the future.
Breast ; Cervix Uteri ; Cross-Sectional Studies ; Depression* ; Depressive Disorder, Major ; Diagnosis ; Epidemiology ; Female ; Humans ; International Classification of Diseases ; Korea* ; Liver ; Lung Neoplasms ; Lymphoma, Non-Hodgkin ; National Health Programs ; Prevalence* ; Prostate ; Psychology ; Stomach ; Thyroid Neoplasms ; Urinary Bladder

PURPOSE: Breast cancer survivors have slightly increased the risk of second primary cancers. Breast, colon, uterine, and ovarian cancers are common secondary cancers in breast cancer survivors. In this study, we assessed the development of second primary cancers of breast cancer survivors in Korea. METHODS: Medical records of patients with breast cancer in 3 tertiary medical institutions were reviewed retrospectively. We evaluated secondary malignancy diagnosed at least 2 months after the breast cancer diagnosis. Based on the International Classification of Disease-9 codes of malignancies, secondary primary breast cancer records were evaluated with person-year adjustment. The standardized incidence ratio (SIR) was assessed using national cancer incidence. RESULTS: A total of 3,444 treatment records were included from 3 medical centers. The cumulative incidence of overall second primary cancers was 2.8% (n = 93). The SIR was significantly higher in all sites (1.56; 95% confidence interval [CI], 1.26–1.91), endometrial cancer (5.65; 95% CI, 2.06–12.31), biliary tract cancer (3.96; 95% CI, 1.19–8.60), and thyroid cancer (2.29; 95% CI, 1.67–3.08). CONCLUSION: The incidence of cancer was higher in breast cancer survivors compared to general population. Surveillance of secondary cancer in this group should be recommended individually considering the benefit related to the prognosis of primary breast cancer.
Biliary Tract Neoplasms ; Breast Neoplasms* ; Breast* ; Classification ; Colon ; Diagnosis ; Early Detection of Cancer ; Endometrial Neoplasms ; Female ; Humans ; Incidence ; Korea ; Medical Records ; Neoplasms, Second Primary* ; Ovarian Neoplasms ; Prognosis ; Retrospective Studies ; Survivors* ; Thyroid Neoplasms

PURPOSE: Breast cancer survivors have slightly increased the risk of second primary cancers. Breast, colon, uterine, and ovarian cancers are common secondary cancers in breast cancer survivors. In this study, we assessed the development of second primary cancers of breast cancer survivors in Korea. METHODS: Medical records of patients with breast cancer in 3 tertiary medical institutions were reviewed retrospectively. We evaluated secondary malignancy diagnosed at least 2 months after the breast cancer diagnosis. Based on the International Classification of Disease-9 codes of malignancies, secondary primary breast cancer records were evaluated with person-year adjustment. The standardized incidence ratio (SIR) was assessed using national cancer incidence. RESULTS: A total of 3,444 treatment records were included from 3 medical centers. The cumulative incidence of overall second primary cancers was 2.8% (n = 93). The SIR was significantly higher in all sites (1.56; 95% confidence interval [CI], 1.26–1.91), endometrial cancer (5.65; 95% CI, 2.06–12.31), biliary tract cancer (3.96; 95% CI, 1.19–8.60), and thyroid cancer (2.29; 95% CI, 1.67–3.08). CONCLUSION: The incidence of cancer was higher in breast cancer survivors compared to general population. Surveillance of secondary cancer in this group should be recommended individually considering the benefit related to the prognosis of primary breast cancer.
Biliary Tract Neoplasms ; Breast Neoplasms* ; Breast* ; Classification ; Colon ; Diagnosis ; Early Detection of Cancer ; Endometrial Neoplasms ; Female ; Humans ; Incidence ; Korea ; Medical Records ; Neoplasms, Second Primary* ; Ovarian Neoplasms ; Prognosis ; Retrospective Studies ; Survivors* ; Thyroid Neoplasms

OBJECTIVETo investigate neuroendocrine differentiation and Wilms' tumor protein-1 (WT-1) expression in breast mucinous carcinoma and their clinicopathological significance.

METHODSThe clinicopathological data of 65 patients with breast mucinous carcinoma, including 31 cases of mixed mucinous carcinoma, 23 cases of hypocellular pure mucinous carcinoma and 11 cases of hypercellular pure mucinous carcinoma, admitted in Taizhou Hospital from January 2010 to June 2015 were retrospectively reviewed. The expression of neuroendocrine markers and WT-1 was detected by immunohistochemistry staining in all cases.

RESULTSThe mixed mucinous carcinomas and hypercelluar pure mucinous carcinomas had higher incidence of axillary lymph node metastasis and human epidermal recepter 2 (HER-2) positive than hypocellular pure mucinous carcinoma (all (P<0.01). However, the difference was not significant between mixed mucinous carcinomas and hypercellular pure mucinous carcinomas (all P>0.05). The expression of neuroendocrine marker was stronger in hypercellular mucinous carcinoma than that in mixed mucinous carcinoma and hypocellular mucinous carcinoma (all (P<0.05), but the difference was not statistically significant between mixed mucinous carcinoma and hypocellular pure mucinous carcinoma (P>0.05). The expression of WT-1 was weaker in mixed mucinous carcinoma than that in hypercellular and hypocellular pure mucinous carcinoma(all (P<0.05), but the difference was not statistically significant between hypercellular and hypocellular pure mucinous carcinoma (P>0.05). The mucinous carcinomas with lymph node metastasis had lower expression of neuroendocrine markers than those without lymph node metastasis ((P<0.01). The expression of WT-1 in breast mucinous carcinoma with lymph node metastasis trended lower than that in those without lymph node metastasis, but the difference was not statistically significant (P>0.05).

CONCLUSIONHypercellular pure mucinous breast carcinoma has higher rates of lymph node metastasis and HER-2 amplification than hypocellular pure mucinous carcinoma, the sub-classification of breast pure mucinous carcinoma should be considered. Neuroendocrine differentiation and WT-1 expression may be helpful in distinguishing the subtypes of breast mucinous carcinoma. Breast mucinous carcinoma with neuroendocrine differentiation trends to have less lymph node metastasis.

Adenocarcinoma, Mucinous ; classification ; diagnosis ; pathology ; Axilla ; Breast Neoplasms ; classification ; diagnosis ; pathology ; Female ; Humans ; Immunohistochemistry ; Incidence ; Lymph Nodes ; pathology ; Lymphatic Metastasis ; Neuroendocrine Tumors ; diagnosis ; pathology ; Receptor, ErbB-2 ; metabolism ; Retrospective Studies ; WT1 Proteins ; metabolism

OBJECTIVETo investigate the expression of p16INK4a protein in breast cancer and analyze its clinical significance.

METHODSA total of 132 surgical specimens of primary breast cancer obtained between 2014 and 2015 were examined for expressions of ER, PR, CK5/6, Her-2 and p16INK4a proteins using immunohistochemistry.

RESULTSThe breast cancer samples were classified into 5 molecular subtypes, namely Luminal A (58 cases), Luminal B (32 cases), Her-2-positive (21 cases), basal-like (12 cases) and normal-like (9 cases) types. p16INK4a expression was negative in 7/132 (5.30%) cases, weakly positive in 15/132 (11.36%) cases, positive in 40/132 (30.30%) cases, and strongly positive in 70/132 (53.03%) cases. When categorizing negative and weakly positive cases into negative group and the positive and strongly positive cases into positive group, the total negative and positive expression rates of p16INK4a were 16.67% (22/132) and 83.33% (110/132) in the carcinoma tissues. Statistical analysis showed the expression intensity of p16INK4a differed significantly between the age groups (P<0.05) but was not significantly correlated with ER, PR, Her-2, molecular subtypes or metastasis of the tumors.

CONCLUSIONThe compensatory high expression of p16INK4a is the main mechanism of cell cycle deregulation in invasive breast cancer and can be an important specific molecular marker for invasive breast cancer.

Biomarkers, Tumor ; metabolism ; Breast Neoplasms ; classification ; diagnosis ; metabolism ; Cyclin-Dependent Kinase Inhibitor p16 ; metabolism ; Female ; Humans ; Keratin-5 ; metabolism ; Keratin-6 ; metabolism ; Receptor, ErbB-2 ; metabolism ; Receptors, Estrogen ; metabolism ; Receptors, Progesterone ; metabolism

During the last 15 years, traditional breast cancer classifications based on histopathology have been reorganized into the luminal A, luminal B, human epidermal growth factor receptor 2 (HER2), and basal-like subtypes based on gene expression profiling. Each molecular subtype has shown varying risk for progression, response to treatment, and survival outcomes. Research linking the imaging phenotype with the molecular subtype has revealed that non-calcified, relatively circumscribed masses with posterior acoustic enhancement are common in the basal-like subtype, spiculated masses with a poorly circumscribed margin and posterior acoustic shadowing in the luminal subtype, and pleomorphic calcifications in the HER2-enriched subtype. Understanding the clinical implications of the molecular subtypes and imaging phenotypes could help radiologists guide precision medicine, tailoring medical treatment to patients and their tumor characteristics.
Acoustics ; Breast Neoplasms* ; Breast* ; Classification ; Diagnosis ; Gene Expression Profiling ; Humans ; Phenobarbital ; Phenotype* ; Precision Medicine ; Receptor, Epidermal Growth Factor ; Shadowing (Histology) ; Ultrasonography

Breast cancer is a heterogeneous disease. The formation and progression of tumor and the sensitivity to treatment differs from patient to patient. In addition to the widely used molecular subtype, novel markers are needed to better personalize the treatment of breast cancer. Tumor infiltrating lymphocyte (TIL) have been consistently documented in breast cancer lesions especially in triple negative and human epidermal growth factor receptor-2 positive breast cancer. Several clinical trials have revealed that TIL are associated with prognosis and can predict therapeutic efficacy of special therapy. TIL could be divided to different subtypes including CD8 + TIL, CD4 + TIL, cytotoxic T lymphocyte-associated antigen-4 + TIL, programmed death-1 + TIL. They play different roles in the process of anti-tumor immunity and can predict different prognosis. Screening out special TIL subtype which is well associated with prognosis and therapeutic efficacy and developing targeting immunotherapy can help to improve outcomes of breast cancer patients.
Breast Neoplasms ; diagnosis ; immunology ; Humans ; Lymphocytes, Tumor-Infiltrating ; classification ; cytology ; Prognosis

The feature extraction and feature selection are the important issues in pattern recognition. Based on the geometric algebra representation of vector, a new feature extraction method using blade coefficient of geometric algebra was proposed in this study. At the same time, an improved differential evolution (DE) feature selection method was proposed to solve the elevated high dimension issue. The simple linear discriminant analysis was used as the classifier. The result of the 10-fold cross-validation (10 CV) classification of public breast cancer biomedical dataset was more than 96% and proved superior to that of the original features and traditional feature extraction method.
Algorithms ; Artificial Intelligence ; Breast Neoplasms ; classification ; diagnosis ; Discriminant Analysis ; Female ; Humans