Deep transfer learning with fuzzy ensemble approach for the early detection of breast cancer.

Breast Cancer is a significant global health challenge, particularly affecting women with higher mortality compared with other cancer types. Timely detection of such cancer types is crucial, and recent research, employing deep learning techniques, shows promise in earlier detection. The research focuses on the early detection of such tumors using mammogram images with deep-learning models. The paper utilized four public databases where a similar amount of 986 mammograms each for three classes (normal, benign, malignant) are taken for evaluation. Herein, three deep CNN models such as VGG-11, Inception v3, and ResNet50 are employed as base classifiers. The research adopts an ensemble method where the proposed approach makes use of the modified Gompertz function for building a fuzzy ranking of the base classification models and their decision scores are integrated in an adaptive manner for constructing the final prediction of results. The classification results of the proposed fuzzy ensemble approach outperform transfer learning models and other ensemble approaches such as weighted average and Sugeno integral techniques. The proposed ResNet50 ensemble network using the modified Gompertz function-based fuzzy ranking approach provides a superior classification accuracy of 98.986%.

Challenges in management of phyllodes tumors of the breast: a retrospective analysis of 150 patients.

INTRODUCTION: Phyllodes tumors (PT) of the breast seem to get pre-operatively misdiagnosed as fibroadenomas resulting in inadequate resections and high local recurrence rates. MATERIALS AND METHODS: Data of 150 patients with PT of the breast managed from January, 2003 to February, 2013 were retrospectively analyzed. Statistical analysis performed using SPSS version 17 (Pearson Chi-square test and analysis of variance test for analysis). AIM: The aim of this study is to compare clinico-pathological profile and recurrence rates in patients with benign (B), borderline malignant (BL) and malignant (M) PT. RESULTS: In a total of 150 patients with PT (n = 77 B, n = 24 BL, n = 49 M), mean age was 36.92, 44.04 and 40.46 years respectively (P 0.015) and mean tumor size being 8.15 cm, 14.7 cm and 12.9 cm respectively (P 0.000). Pre-operatively cytology suggestive of PT in 24% patients with B PT and 63% in M PT; core tissue biopsy suggestive of PT in 85.4% patients with B PT and 100% in M PT. Recurrence seen in 34.7% out of which 32.7% were post-lumpectomy performed elsewhere. Majority of B PT had lumpectomy (49.3%)/wide local excision (WLE, 31.2%) compared with M PT where 55.1% had simple mastectomy (SM) due to large tumor size. Local recurrence was more in M PT (53%) compared with B PT (20%). We found recurrence rates in L (39.3%) compared with WLE (27.3%) and SM (33.9%) (P 0.049). CONCLUSIONS: Larger tumor size, incomplete resection and M/BL histology predicted higher recurrence in PT. Core biopsy is much more accurate than fine needle cytology in the diagnosis.

Phaeospirillum tilakii sp. nov., a phototrophic alphaproteobacterium isolated from aquatic sediments.

Two strains (JA492(T) and JA590) of spiral-shaped, anaerobic, Gram-stain-negative, motile, purple non-sulfur bacteria were isolated from aquatic sediments from a bird sanctuary and a stream, respectively, and were characterized by a polyphasic taxonomic approach. Bacteriochlorophyll a and carotenoids (rhodopin, lycopene, hydroxylycopene glucoside and dihydroxylycopene diglucoside) were present as photosynthetic pigments. Intracellular photosynthetic membranes were of the stacked type. The major fatty acids were C(18 : 1)ω7c, C(16 : 0) and summed feature 3 (C(16 : 1)ω6c and/or C(16 : 1)ω7c) in both strains. Ubiquinones and menaquinones were present as major quinone components. The genomic DNA G+C contents of strains JA492(T) and JA590 were 63.8 and 61.5 mol%, respectively. Both strains were closely related (mean DNA-DNA hybridization >70 %). Phylogenetic analysis showed that the strains clustered with species of the genus Phaeospirillum of the family Rhodospirillaceae, class Alphaproteobacteria. Based on 16S rRNA gene sequence analysis, both strains showed highest sequence similarity with Phaeospirillum oryzae JA317(T) (97.2-97.4 %), Phaeospirillum molischianum DSM 120(T) (96.5-96.7 %), Phaeospirillum fulvum DSM 113(T) (96.7-96.9 %) and Phaeospirillum chandramohanii JA145(T) (96.5-96.7 %). DNA-DNA relatedness between strain JA492(T) and its closest relative in the genus Phaeospirillum was less than 42 %. It is evident from phenotypic, chemotaxonomic and molecular genetic data that strain JA492(T) represents a novel species of the genus Phaeospirillum, for which the name Phaeospirillum tilakii sp. nov. is proposed; the type strain is JA492(T) ( = NBRC 107650(T) = KCTC 15012(T)).

Small bowel trans-vaginal evisceration following vault biopsy: general surgeons beware!

A case-report of vaginal evisceration following vault biopsy is described. This case highlights the importance of good surgical technique when performing a vaginal biopsy in order to avoid this rare, but life-threatening, complication. General surgeons may well be faced with this acute presentation and prompt management is vital in order to preserve the involved small bowel.

The antifouling and fouling-release performance of hyperbranched fluoropolymer (HBFP)-poly(ethylene glycol) (PEG) composite coatings evaluated by adsorption of biomacromolecules and the green fouling alga Ulva.

Cross-linked hyperbranched fluoropolymer (HBFP) and poly(ethylene glycol) (PEG) amphiphilic networks with PEG weight percentages of 14% (HBFP-PEG14), 29% (HBFP-PEG29), 45% (HBFP-PEG45), and 55% (HBFP-PEG55) were prepared on 3-aminopropyl)triethoxysilane (3-APS) functionalized microscope glass slides for marine antifouling and fouling-release applications. The surface-free energies (gamma(s)), polar (gamma(s)(p) and gamma(s)(AB)), and dispersion (gamma(s)(d) and gamma(s)(LW)) components were evaluated using advancing contact angles by two-liquid geometric-mean and three-liquid Lifshitz-van der Waals acid-base approaches. The HBFP coating exhibited a low surface energy of 22 mJ/m(2), while the gamma(s) and gamma(s)(p) of the cross-linked HBFP-PEG coatings increased proportionally with the PEG weight percentages in the networks. The adsorption of bovine serum albumin (BSA), lectin from Codium fragile (CFL), lipopolysaccharides from Escherichia coli (LPSE) and Salmonella minnesota (LPSS) upon glass, APS-glass, HBFP, PEG, and the cross-linked HBFP-PEG network coatings were investigated by fluorescence microscopy. The marine antifouling and fouling-release properties of the cross-linked HBFP-PEG coatings were evaluated by settlement and release assays involving zoospores of green fouling alga Ulva (syn. Enteromorpha; Hayden, H. S.; Blomster, J.; Maggs, C. A.; Silva, P. C.; Stanhope, M. J.; Waaland, J. R. Eur. J. Phycol. 2003, 38, 277). The growth and release of Ulva sporelings were also investigated upon the HBFP-PEG45 coating in comparison to a poly(dimethylsiloxane) elastomer (PDMSE) standard material. Of the heterogeneous cross-linked network coatings, the maximum resistances to protein, lipopolysaccharide, and Ulva zoospore adhesion, as well as the best zoospore and sporeling release properties, were recorded for the HBFP-PEG45 coating. This material also exhibited better performance than did a standard PDMSE coating, suggesting its unique applicability in fouling-resistance applications.

XNEOr: development and evaluation of an expert system to improve the quality and cost of decision-making in neuro-oncology.

The treatment of brain tumors requires a large team of medical experts. However, the process of medical decision-making for these patients is hampered by the frequent inaccessibility of the experts because of conflicting scheduling, inconsistencies in the management of different patients, and the fact that multiple experts often yield multiple opinions. The goals of this work were (1) to develop and validate an expert system to assist the medical team deliver efficient, quality care to children with recurrent medulloblastoma, a common type of pediatric brain tumor, and (2) to determine if the expert system can be used as an educational tool. The results of our study indicate that residents enjoy learning by using XNEOr, the brain tumor expert system. XNEOr enabled residents to order appropriate ancillary tests for patients and to make fewer incorrect treatment decisions. The potential net effect of residents using XNEOr may be increased patient and family satisfaction and decreased probability of medical liability. At a time of important changes in our health care system, novel expert systems hold promise as tools to reduce medical costs, improve the quality of multi-expert medical care, and advance health care education.