Although microscopic analysis of tissue slides has been the basis for disease diagnosis for decades, intra- and inter-observer variabilities remain issues to be resolved. The recent introduction of digital scanners has allowed for using deep learning in the analysis of tissue images because many whole slide images (WSIs) are accessible to researchers. In the present study, we investigated the possibility of a deep learning-based, fully automated, computer-aided diagnosis system with WSIs from a stomach adenocarcinoma dataset. Three different convolutional neural network architectures were tested to determine the better architecture for tissue classifier. Each network was trained to classify small tissue patches into normal or tumor. Based on the patch-level classification, tumor probability heatmaps can be overlaid on tissue images. We observed three different tissue patterns, including clear normal, clear tumor and ambiguous cases. We suggest that longer inspection time can be assigned to ambiguous cases compared to clear normal cases, increasing the accuracy and efficiency of histopathologic diagnosis by pre-evaluating the status of the WSIs. When the classifier was tested with completely different WSI dataset, the performance was not optimal because of the different tissue preparation quality. By including a small amount of data from the new dataset for training, the performance for the new dataset was much enhanced. These results indicated that WSI dataset should include tissues prepared from many different preparation conditions to construct a generalized tissue classifier. Thus, multi-national/multi-center dataset should be built for the application of deep learning in the real world medical practice.
Adenocarcinoma ; Classification ; Dataset ; Diagnosis ; Learning ; Observer Variation ; Stomach

Adenocarcinoma ; classification ; diagnosis ; metabolism ; Humans ; Immunohistochemistry ; International Agencies ; Lung Neoplasms ; classification ; diagnosis ; metabolism ; Neoplasm Invasiveness ; Societies, Medical ; United States

BACKGROUND: The usefulness of quantitative nuclear image analysis in the classification of lung carcinoma is widely investigated and published. In this study, we tried to measure the nuclear characteristics of primary lung carcinomas by image analysis and to find the possibility of differential diagnoses. METHODS: Seventeen cases of adenocarcinomas (not including bronchioloalveolar carcinoma), seven of bronchioloalveolar carcinomas, eight of large cell neuroendocrine carcinomas, five of small cell carcinamas, and 26 of squamous cell carcinomas were analysed. Three different images of each case were captured by digital camera, and we analyzed the nuclear area, perimeter, circularity, and density using the Optimas 6.5 Image Analyser software package. Statistical analyses were done using the statistical program STATISTICA kernel release 5.5. RESULTS: The mean nuclear area was 0.52+/-0.25micrometer2 in the adenocarcinomas, 0.50+/-1.82micrometer2 in the squamous cell carcinomas, 0.45+/-0.20micrometer2 in the large cell neuroendocrine carcinomas, 0.42+/-0.16micrometer2 in the bronchioloalveolar carcinomas, and 0.31+/-0.12micrometer2 in the small cell carcinamas. The nuclear area was significantly different between the small cell carcinomas and the non-small cell carcinomas (p<0.01) and between the adenocarcinomas and the bronchioloalveolar carcinomas (p=0.02). The mean nuclear perimeter was 3.36+/-0.92micrometer2 in the adenocarcinomas, 3.24+/-0.67micrometer2 in the squamous cell carcinomas, 3.16+/-0.82micrometer2 in the large cell neuroendocrine carcinomas, 3.05+/-0.80micrometer2 in the bronchioloalveolar carcinomas, and 2.54+/-0.62micrometer2 in the small cell carcinamas. The nuclear perimeter was significantly different between the small cell carcinomas and the non-small cell carcinomas (p<0.04). The nuclear circularity showed no statistical difference. Nuclear density was the highest in the squamous cell carcinomas, and the lowest in the small cell carcinomas. The large cell neuroendocrine carcinomas showed the lowest standard deviation in nuclear density. CONCLUSION: The analysis of nuclear characteristics using an image analyser can be used as an objective method in the classification of lung carcinoma.
Adenocarcinoma ; Adenocarcinoma, Bronchiolo-Alveolar ; Carcinoma, Neuroendocrine ; Carcinoma, Small Cell ; Carcinoma, Squamous Cell ; Classification ; Diagnosis, Differential ; Lung Neoplasms* ; Lung*

OBJECTIVETo compare the old classification and 2004 WHO histological classification of renal cell carcinoma, summarize the differences and possible reasons, and correct the traditional pathological concepts of kidney cancer.

METHODSSpecimens of 79 cases histopathologically diagnosed as non-clear cell renal cell carcinomas after radical nephrectomy during 1998 to 2005 in Tianjin Medical University Cancer Hospital were reclassified according to the 2004 WHO renal cell carcinoma histological classification system.

RESULTSAfter reclassification, there were 14 cases of clear cell renal cell carcinoma (CCRCC), 23 cases of papillary renal cell carcinoma (PRCC), 34 cases of chromophobe renal cell carcinoma (ChRCC), one collecting duct renal cell carcinoma, one unclassified renal cell carcinoma, 5 cases of mixed cell renal cell carcinoma (CCRCC + PRCC 2 cases, CCRCC + ChRCC 2 cases, PRCC + ChRCC 1 case), and one oncocytoma diagnosed.

CONCLUSIONSSome chromophobe renal cell carcinomas and papillary renal cell carcinomas were easier to be diagnosed as granular cell renal cell carcinoma in the past. The eosinophilic cytoplasm similar to that in the granular cells, and some confusion between PRCC and ChRCC are the main reasons. The cellular characteristic features of granular renal cell carcinoma can be found in many types of renal tumors. Granular cell renal cell carcinoma is not an independent entity, therefore, it should be removed from the histological classification of renal cell carcinoma. The diagnosis standard of mixed renal cell carcinoma (MRCC) need to be determined and consummated.

Adenocarcinoma ; pathology ; Carcinoma, Renal Cell ; classification ; pathology ; Diagnosis, Differential ; Humans ; Kidney Neoplasms ; classification ; pathology ; World Health Organization

PURPOSE: Mucinous gastric carcinoma (MGC) is a histopathologic subtype of gastric adenocarcinoma with a poor prognosis. It comprises about 3~10% of gastric carcinomas. The purpose of this study was to compare the disease course of MGC with non-MGC (NMGC) and study the clinicopathologic features that influence the prognosis of MGC patients. METHODS: We reviewed the records of 2,383 patients with a confirmed histologic diagnosis of gastric carcinoma who underwent surgery at the Department of Surgery, Chonnam National University Hospital. There were 157 patients with MGC compared to 2,226 with NMGC. Patients were evaluated on the basis of gender, age, tumor size, tumor location, depth of invasion, region and number of lymph nodes with metastasis, hepatic or peritoneal metastasis, stage at presentation, estimate of surgical curability, and TNM stage based on the UICC classification. Multivariate analysis was performed to test the hypothesis that the histologic mucin contents themselves in MGC are an independent prognostic factor. RESULTS: There was no gender or age-at-diagnosis distinction between these two groups. The mean tumor size of MGC was larger than that of NMGC, but the difference was not statistically significant. Most carcinomas of both types were located in the antrum with no statistical difference in location between MGC and NMGC. However, a depth of invasion greater than T3 was more frequently found in MGC than in NMGC, not to a statistically significant degree. The mean number of lymph node with metastases was 2.78 in MGC and 2.28 in NMGC (P<0.001). There were more MGC patients with TNM stages II through IV(UICC classification). The overall survival rate was lower for the MGC group(46.5%) than for the NMGC group (64.0%). Depth of invasion, lymph node metastasis, and stage at diagnosis were significant factors affecting the outcome. Mucinous histologic type itself was not an independent predictive factor in survival. CONCLUSION: The factors that influence the poorer prognosis(lower 5-year survival rate) of MGC are advanced stage at the time of diagnosis, lymph node metastases, and a higher TNM status. The histologic subtype itself was not an independent prognostic factor.
Adenocarcinoma ; Classification ; Diagnosis ; Humans ; Jeollanam-do ; Lymph Nodes ; Mucins* ; Multivariate Analysis ; Neoplasm Metastasis ; Prognosis* ; Survival Rate

The diagnosis of gastric cancer in young age group was sometimes missed .The operative risk in senile age group was high because of combined other organ diseases. We tried to determine the difference in their correct clinicopathologic features and the prognosis of young and senile patients with gastric cancer. Clinicopathologic characteristics and surgical results were compared in 40 senile gastric cancer patients who were aged 65 years or above, and in 48 young gastric cancer patients who were aged 40 years or less. In clinicopathologic features, the senile group was characterized by a high incidence of well differentiated adenocarcinoma and intestinal type by Lauren classification. The young age group was characterized by high incidence of undifferentiated type adenocarcinoma and diffuse type by Lauren classification . The others were unremarkable. When the survival rate was compared in all cases between young and senile group, the young age group showed a more favorable prognosis than the senile group, but without statistical difference ( P=0.0058 ) . Also, the survival rate according to UICC stage showed no statistical difference when the same stage of the two different groups were compared . Both group showed better survival in curative surgery cases than in noncurative surgery cases.Our findings ruled out any relationships between age and length of survival time in patients undergoing curative gastrectomy for gastric cancer.
Adenocarcinoma ; Classification ; Diagnosis ; Gastrectomy ; Humans ; Incidence ; Prognosis ; Stomach Neoplasms* ; Survival Rate

Ceruminomas are rare tumors arising from the ceruminous gland, which is a modified apocrine gland in the skin of the external ear canal. There is controversy about these rare tumors regarding their histological classification, their origin and the importance of wide excision, and there is also terminological confusion for making the diagnosis. Ceruminous adenocarcinoma is a malignant subtype of ceruminoma. We report here on a case of adenocarcinoma of the ceruminous gland in a 71-year-old male patient.
Adenocarcinoma* ; Aged ; Apocrine Glands ; Classification ; Diagnosis ; Ear Canal* ; Ear Neoplasms ; Ear, External* ; Humans ; Male ; Skin

PURPOSE: There have been several comparative studies that have focused on elderly groups of patients with gastric cancer. However, new criteria are needed for this elderly group because of the longer life span of Korean people. The diagnosis of gastric cancer has sometimes been missed in the young age group. The perioperative risk is high in the elderly age group because of their combined diseases. This study was designed to determine the differences of the clinicopathologic features and the prognosis between young and elderly patients with gastric cancer. MATERIALS AND METHODS: Eighty patients were divided in two groups and these patients were selected for making comparison between young and elderly groups of patients with gastric cancer. The young age group consisted of 31 patients who were aged 35 years old or less. The elderly age group was made up of 49 patients who were aged 75 years old or above. RESULTS: For the clinicopathologic features, the young age group was characterized by a high incidence of the poorly differentiated type of adenocarcinoma and the diffuse type too, according to the Lauren classification. On the other hand, the elderly group was characterized by a high incidence of poorly to moderate differentiated adenocarcinoma and also the intestinal type according to the Lauren classification. The other clinical differences were unremarkable. Additionally, there was no survival advantage in the young age group compared to the elderly group. CONCLUSION: There were no clinicopathologic and prognostic differences between both extreme age groups. So, active surgical treatment is recommended even for the elderly patients group.
Adenocarcinoma ; Adult ; Aged ; Classification ; Diagnosis ; Hand ; Humans ; Incidence ; Prognosis* ; Stomach Neoplasms*

Adenocarcinoma ; Adenoma ; Adult ; Cysts ; Female ; Humans ; Male ; Middle Aged ; Thyroid Neoplasms ; classification ; diagnosis ; pathology

Adenocarcinoma, Follicular ; classification ; diagnosis ; pathology ; Carcinoma, Papillary ; diagnosis ; pathology ; surgery ; Humans ; Lymphatic Metastasis ; Thyroid Neoplasms ; diagnosis ; pathology ; surgery