[Automatic Quantification of Breast Density from Mammography Using Deep Learning].

BACKGROUND: In the field of breast screening using mammography, announcing to the examinees whether they are dense or not has not been deprecated in Japan. One of the reasons is a shortage of objectivity estimating their dense breast. Our aim is to build a system with deep learning algorithm to calculate and quantify objective breast density automatically. MATERIAL AND METHOD: Mammography images taken in our institute that were diagnosed as category 1 were collected. Each processed image was transformed into eight-bit grayscale, with the size of 2294 pixels by 1914 pixels. The "base pixel value" was calculated from the fatty area within the breast for each image. The "relative density" was calculated by dividing each pixel value by the base pixel value. Semantic segmentation algorithm was used to automatically segment the area of breast tissue within the mammography image, which was resized to 144 pixels by 120 pixels. By aggregating the relative density within the breast tissue area, the "breast density" was obtained automatically. RESULT: From each but one mammography image, the breast density was successfully calculated automatically. By defining a dense breast as the breast density being greater than or equal to 30%, the evaluation of the dense breast was consistent with that by a computer and human (76.6%). CONCLUSION: Deep learning provides an excellent estimation of quantification of breast density. This system could contribute to improve the efficiency of mammography screening system.

Comparative Analysis of Bone Marrow-derived Mast Cell Differentiation in C57BL/6 and BALB/c Mice.

Allergic diseases have increased in the last three decades. Mast cells play a critical role in allergic diseases along with allergen-specific immunoglobulin E (IgE). Following mast cell degranulation elicited by ligation of the IgE-FcεRI receptor complex with allergen, allergic reactions are followed by various symptoms such as vascular hyperpermeability, mucous secretion, itching, sneezing, wheezing, rashes, fever, and anaphylactic shock. Susceptibility or inclination to allergy varies depending on individual genetic traits and living environment, and it has long been believed that such an inclination is determined by an immunologic balance of T helper cell types. Mouse strains also have different susceptibilities to allergy. Similar to T helper cells and macrophages, it is not known whether mast cells can also be divided into two different types between mouse strains. In this study, we prepared bone marrow-derived mast cells from BALB/c and C57BL/6 mice and examined their cellular properties. Cellular response to IL-3 and the process of mast cell differentiation from bone marrow cells were different on the basis of cell surface marker molecules. BALB/c-derived cells more efficiently exhibited degranulation than did C57BL/6-derived cells following both calcium ionophore and receptor crosslinking. These functional differences persisted even after a longer cell culture for 8 weeks, suggesting a difference in cell-autonomous characteristics. These results support the concept that mast cells also have different cell types dependent on their genetic background. Abbreviations: Ab: antibody; BMMC: bone marrow-derived mast cell; DNP: dinitrophenyl; FACS: fluorescence-activated cell sorter; FCS: fetal calf serum; FITC: fluorescein isothiocyanate; FSC: forward scatter; HRP: horseradish peroxidase; HSA: human serum albumin; Ig: immunoglobulin; IL: interleukin; MIP-2: macrophage inflammatory protein-2; MCP: mast cell protease; PE: phycoerythrin; PerCP: Peridinin chlorophyll protein complex; SNP: single nucleotide polymorphisms; SSC: side scatter; Th: T helper; TNF-α: tumor necrosis factor alpha.

A novel biomarker TERTmRNA is applicable for early detection of hepatoma.

BACKGROUNDS: We previously reported a highly sensitive method for serum human telomerase reverse transcriptase (hTERT) mRNA for hepatocellular carcinoma (HCC). alpha-fetoprotein (AFP) and des-gamma-carboxy prothrombin (DCP) are good markers for HCC. In this study, we verified the significance of hTERTmRNA in a large scale multi-centered trial, collating quantified values with clinical course. METHODS: In 638 subjects including 303 patients with HCC, 89 with chronic hepatitis (CH), 45 with liver cirrhosis (LC) and 201 healthy individuals, we quantified serum hTERTmRNA using the real-time RT-PCR. We examined its sensitivity and specificity in HCC diagnosis, clinical significance, ROC curve analysis in comparison with other tumor markers, and its correlations with the clinical parameters using Pearson relative test and multivariate analyses. Furthermore, we performed a prospective and comparative study to observe the change of biomarkers, including hTERTmRNA in HCC patients receiving anti-cancer therapies. RESULTS: hTERTmRNA was demonstrated to be independently correlated with clinical parameters; tumor size and tumor differentiation (P < 0.001, each). The sensitivity/specificity of hTERTmRNA in HCC diagnosis showed 90.2%/85.4% for hTERT. hTERTmRNA proved to be superior to AFP, AFP-L3, and DCP in the diagnosis and underwent an indisputable change in response to therapy. The detection rate of small HCC by hTERTmRNA was superior to the other markers. CONCLUSIONS: hTERTmRNA is superior to conventional tumor markers in the diagnosis and recurrence of HCC at an early stage.

Elevated serum ALT levels during pegylated interferon monotherapy may be caused by hepatic iron overload.

OBJECTIVE: Persistently elevated serum alanine aminotransferase (ALT) levels have been observed in chronic hepatitis C (CHC) patients during pegylated interferon (PEG-IFN) therapy. We investigated whether elevated serum ALT levels during PEG-IFN therapy are associated with iron overload. METHODS: Sixty-three CHC patients treated with PEG-IFNalpha-2a monotherapy were evaluated. The associations between elevated serum ALT levels (> or =70 IU/l) were investigated before and 24 weeks after therapy. We classified patients as follows: patients with no elevated serum ALT levels (group NE: n = 35), patients with elevated serum ALT levels (group E: n = 28), and patients with no elevated serum ALT level and negative HCV RNA (group NE-: n = 24), and patients with elevated serum ALT level and negative HCV RNA (group E-: n = 19). We also compared total iron score (TIS) and fibrosis stage in liver specimens obtained before and during therapy from 3 patients with elevated serum ALT levels. RESULTS: Serum ferritin levels were significantly increased after 24 weeks compared to baseline levels in group E (218 +/- 273 vs. 438 +/- 308 ng/ml; p < 0.0001) and group E- (146 +/- 152 vs. 410 +/- 291 ng/ml; p < 0.0001). Serum ALT and ferritin levels were significantly correlated after 24 weeks. The liver specimens revealed that TIS and fibrosis progressed during therapy. CONCLUSION: Our findings suggest that the elevation in serum ALT levels during therapy is caused by iron overload which may be induced by PEG-IFNalpha-2a.

Development of a novel assay to quantify serum human telomerase reverse transcriptase messenger RNA and its significance as a tumor marker for hepatocellular carcinoma.

Currently available tumor markers for hepatocellular carcinoma (HCC) are alpha-fetoprotein (AFP), lens culinaris agglutinin-reactive AFP (AFP-L3), and Des-gamma-carboxy prothrombin (DCP). However, their positive rate can not surpass abdominal ultrasonography (US) as modalities to detect small HCC at early stage, resulting in a possible delay of its diagnosis. There is a need to develop an additional sensitive marker to improve the early detection of HCC. We here introduced a newly developed quantitative detection method for serum hTERT mRNA, which has a clinical significance in HCC diagnosis. Briefly, we examined its sensitivity and specificity in HCC diagnosis, clinical significance in comparison with other tumor markers, and its correlations with the clinical parameters. Serum hTERT mRNA showed higher values in patients with HCC than those with chronic liver diseases. hTERT mRNA expression independently correlated with clinical parameters such as differentiation degree (p < 0.001). The sensitivity/specificity of hTERT mRNA in HCC diagnosis showed 88.2/70.0%. hTERT mRNA proved to be expectedly superior to AFP mRNA , AFP and DCP in HCC diagnosis. Importantly, hTERT mRNA in serum correlated with that in HCC tissue. Thus, we report that serum hTERT mRNA is a novel and available marker for HCC diagnosis.

Outcomes of nontransplant potentially curative therapy for early-stage hepatocellular carcinoma in Child-Pugh stage A cirrhosis is comparable with liver transplantation.

BACKGROUND: This study was undertaken to assess the outcome of potentially curative therapy for early-stage hepatocellular carcinoma (HCC) in patients with Child-Pugh stage A cirrhosis as well as to investigate the impact of low-dose interferon (IFN) therapy after curative therapy on survival. METHODS: This study retrospectively evaluated clinical outcomes in a cohort of 224 Child-Pugh stage A cirrhotic patients who received either resection (53 cases) or radiofrequency ablation (RFA: 171 cases) for HCC within Milan criteria. Thirty patients were treated with low-dose maintenance IFN therapy after initial curative therapy. The median follow-up period was 36.7 months. RESULTS: The 5-year survival rate of all patients was 74.9%, with similar rates for the resection and RFA groups (70.4 vs. 76.8%; p = 0.561). The 5-year HCC recurrence rate was higher in the RFA group than the resection group (85.3 vs. 73.2%; p = 0.012). The maintenance IFN-treated group maintained their liver function within Child-Pugh stage A for a significantly longer time (median time 36.9 vs. 32.2 months; p = 0.0025). CONCLUSION: The 5-year outcomes of resection and RFA in patients with Child-Pugh stage A cirrhosis and early stage HCC were comparable with liver transplantation. Low-dose, long-term maintenance IFN therapy after curative therapy was significantly beneficial on survival.

Prognostic factors for portal venous invasion in patients with hepatocellular carcinoma.

BACKGROUND: Factors involved in portal venous invasion (PVI) must be clarified to enable better determination of therapeutic strategies and outcomes in patients with hepatocellular carcinoma (HCC). METHODS: Of 365 patients with HCC who consulted our department between January 1999 and January 2003, 53 with PVI at the initial consultation were excluded, and the other 312 without PVI were included in this study. Of these patients, we compared liver function, tumor markers, and initial treatment between 287 patients without PVI during follow-up (until December 2004) and 25 patients who developed PVI, and investigated prognostic factors. RESULTS: Multivariate analysis using a COX regression model showed that a Lens culinaris A-reactive fraction of alpha-fetoprotein (AFP-L3) rate of 15% or more, a des-gamma-carboxy prothrombin (DCP) level of 100 mAU/ml or more, multiple tumors, and a platelet count of 130 000/mm(3) or more were correlated with PVI. CONCLUSIONS: HCC frequently infiltrated the portal vein in patients with a high rate of AFP-L3, a high level of DCP, or multiple tumors. Furthermore, the incidence of PVI was significantly higher in patients with a platelet count of 130 000/mm(3) or more.