QuMinS: fast and scalable querying, mining and summarizing multi-modal databases

Given a large image set, in which very few images have labels, how to guess labels for the remaining majority? How to spot images that need brand new labels different from the predefined ones? How to summarize these data to route the user’s attention to what really matters? Here we answer all these questions. Specifically, we propose QuMinS, a fast, scalable solution to two problems: (i) Low-labor labeling (LLL) – given an image set, very few images have labels, find the most appropriate labels for the rest; and (ii) Mining and attention routing – in the same setting, find clusters, the top-'N IND.O' outlier images, and the 'N IND.R' images that best represent the data. Experiments on satellite images spanning up to 2.25 GB show that, contrasting to the state-of-the-art labeling techniques, QuMinS scales linearly on the data size, being up to 40 times faster than top competitors (GCap), still achieving better or equal accuracy, it spots images that potentially require unpredicted labels, and it works even with tiny initial label sets, i.e., nearly five examples. We also report a case study of our method’s practical usage to show that QuMinS is a viable tool for automatic coffee crop detection from remote sensing images.

Quantitative analysis of photodynamic therapy effects in rat mammary tumor vascular density using image-pro plus software

Photodynamic therapy (PDT) is a treatment modality that has advanced rapidly in recent years. It causes tissue and vascular damage with the interaction of a photosensitizing agent (PS), light of a proper wavelength, and molecular oxygen. Evaluation of vessel damage usually relies on histopathology evaluation. Results are often qualitative or at best semi-quantitative based on a subjective system. The aim of this study was to evaluate, using CD31 immunohistochem- istry and image analysis software, the vascular damage after PDT in a well-established rodent model of chemically induced mammary tumor. Fourteen Sprague-Dawley rats received a single dose of 7,12-dimethylbenz(a)anthraxcene (80 mg/kg by gavage), treatment efficacy was evaluated by comparing the vascular density of tumors after treatment with Photogem® as a PS, intraperitoneally, followed by interstitial fiber optic lighting, from a diode laser, at 200 mW/cm and light dose of 100 J/cm directed against his tumor (7 animals), with a control group (6 animals, no PDT). The animals were euthanized 30 hours after the lighting and mammary tumors were removed and samples from each lesion were formalin-fixed. Immunostained blood vessels were quantified by Image Pro-Plus version 7.0. The control group had an average of 3368.6 ± 4027.1 pixels per picture and the treated group had an average of 779 ± 1242.6 pixels per area (P < 0.01), indicating that PDT caused a significant decrease in vascular density of mammary tumors. The combined immu- nohistochemistry using CD31, with selection of representative areas by a trained pathology, followed by quantification of staining using Image Pro-Plus version 7.0 system was a practical and robust methodology for vessel damage evalua- tion, which probably could be used to assess other antiangiogenic treatments.

Influence of light, temperature and metallic ions on biodiesel degradation and corrosiveness to copper and brass

Corrosion is a relevant issue regarding the problem of biodiesel compatibility with polymers and metals. This work aims to evaluate the influence of the natural light incidence and temperature in the corrosion rate of brass and copper immersed in commercial biodiesel as well as biodiesel degradation after the contact with metallic ions. The characterization of corrosion behavior was performed by weight loss measurements according to ASTM G1 and ASTM G31. The experiments according to ASTM G1 were performed at room temperature in light presence and absence. Experiments were also conducted at 55 degrees C in order to compare with ASTM G31 that is also performed at that temperature. The biodiesel degradation was characterized by water content, oxidation stability, viscosity as well as XRF, IR and Raman spectroscopies. The results of ASTM G1 tests showed that the thickness loss for both metals determined at room temperature is slightly higher when there is light incidence and these values significantly decrease for the highest temperature. The results of ASTM G31 tests indicated that air bubbling along with higher temperature affects mostly immersed samples. Biodiesel in contact with metals shows significant degradation in its properties as evidenced by increasing water content, higher viscosity and lower oxidation stability. (C) 2012 Elsevier Ltd. All rights reserved.

Relationship between eggplant seed morphology and germination

Structural differences such as abnormalities, damage and free spaces in seeds may affect germination. The aim of this study was to study the relationship between eggplant seed morphology and seed germination. Ten seed lots of the eggplant cultivar Embu were evaluated by X-ray image analysis and the germination test. Seed image analysis was performed by Image Pro Plus® software and the whole seed area and free space between the embryo and endosperm were measured. The internal seed area filled by the embryo and endosperm was calculated from the difference between the whole seed and free space areas. Based on these results and visual seed analysis, seeds were classified into three categories and information on germination was obtained for each one. X-ray image analysis provides a perfect view of the internal seed parts and for seed morphology studies. An increase in seed area filled by the endosperm and embryo does not improve seed germination. Mechanical seed damage and deteriorated tissues can adversely affect seed germination.

Cooking Effects on Iron and Proteins Content of Beans (Phaseolus Vulgaris L.) by GF AAS and MALDI-TOF MS

The effects of domestic cooking on proteins, organic compounds and Fe distribution in beans (Phaseolus vulgaris L.) were investigated. Sequential extraction with different extractant solutions (mixture of methanol and chloroform 1:2 v/v, water, 0.5 mol L-1 NaCl, 70% v/v ethanol and 0.5 mol L-1 NaOH) were used for extracting lipids, albumins, globulins, prolamins and glutelins, respectively. Iron determination by graphite furnace atomic absorption spectrometry (GF AAS), proteins by Bradford method and organic compounds by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) were carried out in this work. High concentration of albumins, globulins and glutelins were found in raw beans, while in the cooked beans, albumins and glutelins are main proteins types. The MALDI-TOF MS spectra of raw and cooked beans revealed that the domestic cooking altered the molecular weight of the organic compounds, since that in the cooked beans were found compounds between 2 and 3.5 kDa, which were not presented in the raw beans. Besides this, in cooked beans were also observed the presence of four compounds of high molecular weight (12-16 kDa), being that in the raw grains there is only one (ca. 15.2 kDa). In raw grains is possible to observe that Fe is mainly associated to albumins, globulins and glutelins. For cooked grains, Fe is associated to albumins and globulins.

Fibers from fruit by-products enhance probiotic viability and fatty acid profile and increase CLA content in yoghurts

This study evaluated the effect of the supplementation of total dietary fiber from apple, banana or passion fruit processing by-products on the post-acidification, total titratable acidity, bacteria counts and fatty acid profiles in skim milk yoghurts co-fermented by four different probiotics strains: Lactobacillus acidophilus L10 and Bifidobacterium animalis subsp. lactis BL04, HN019 and B94. Apple and banana fibers increased the probiotic viability during shelf-life. All the fibers were able to increase the short chain and polyunsaturated fatty acid contents of yoghurts compared to their respective controls. A synergistic effect between the type of fiber and the probiotic strain on the conjugated linoleic acid content was observed, and the amount of alpha-linolenic acid was increased by banana fiber. The results of this study demonstrate, for the first time, that fruit fibers can improve the fatty acid profile of probiotic yoghurts and point out the suitability of using fibers from fruit processing the by-products to develop new high value-added fermented dairy products. (C) 2012 Elsevier By. All rights reserved.

Cooking effects on iron and proteins content of beans (Phaseolus Vulgaris L.) by GF AAS and MALDI-TOF MS

The effects of domestic cooking on proteins, organic compounds and Fe distribution in beans (Phaseolus vulgaris L.) were investigated. Sequential extraction with different extractant solutions (mixture of methanol and chloroform 1:2 v/v, water, 0.5 mol L-1 NaCl, 70% v/v ethanol and 0.5 mol L-1 NaOH) were used for extracting lipids, albumins, globulins, prolamins and glutelins, respectively. Iron determination by graphite furnace atomic absorption spectrometry (GF AAS), proteins by Bradford method and organic compounds by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) were carried out in this work. High concentration of albumins, globulins and glutelins were found in raw beans, while in the cooked beans, albumins and glutelins are main proteins types. The MALDI-TOF MS spectra of raw and cooked beans revealed that the domestic cooking altered the molecular weight of the organic compounds, since that in the cooked beans were found compounds between 2 and 3.5 kDa, which were not presented in the raw beans. Besides this, in cooked beans were also observed the presence of four compounds of high molecular weight (12-16 kDa), being that in the raw grains there is only one (ca. 15.2 kDa). In raw grains is possible to observe that Fe is mainly associated to albumins, globulins and glutelins. For cooked grains, Fe is associated to albumins and globulins.

PRoSPer: perceptual similarity queries in medical CBIR systems through user profiles

In this paper, we present a novel approach to perform similarity queries over medical images, maintaining the semantics of a given query posted by the user. Content-based image retrieval systems relying on relevance feedback techniques usually request the users to label relevant/irrelevant images. Thus, we present a highly effective strategy to survey user profiles, taking advantage of such labeling to implicitly gather the user perceptual similarity. The profiles maintain the settings desired for each user, allowing tuning of the similarity assessment, which encompasses the dynamic change of the distance function employed through an interactive process. Experiments on medical images show that the method is effective and can improve the decision making process during analysis.