Discovering Mammography-based Machine Learning Classifiers for Breast Cancer Diagnosis

This work explores the design of mammography-based machine learning classifiers (MLC) and proposes a new method to build MLC for breast cancer diagnosis. We massively evaluated MLC configurations to classify features vectors extracted from segmented regions (pathological lesion or normal tissue) on craniocaudal (CC) and/or mediolateral oblique (MLO) mammography image views, providing BI-RADS diagnosis. Previously, appropriate combinations of image processing and normalization techniques were applied to reduce image artifacts and increase mammograms details. The method can be used under different data acquisition circumstances and exploits computer clusters to select well performing MLC configurations. We evaluated 286 cases extracted from the repository owned by HSJ-FMUP, where specialized radiologists segmented regions on CC and/or MLO images (biopsies provided the golden standard). Around 20,000 MLC configurations were evaluated, obtaining classifiers achieving an area under the ROC curve of 0.996 when combining features vectors extracted from CC and MLO views of the same case.     

The role of supportive supervision using mobile technology in monitoring and guiding program performance: a case study in Nigeria, 2015–2016

Background

Supportive supervision is one of the interventions that fosters program improvement by way of imparting knowledge and skills to health workers. The basic challenge in supportive supervision is the availability of data in real time for timely and effective feedback. Thus, the main objective of this study was to determine the contribution of real-time data collection during supportive supervision for timely feedback and generation of evidence for health intervention planning.

Methods

We analyzed supportive supervision records collected through handheld devices employing the open data kit (ODK) platform from July 2015 to June 2016. Supervision was conducted across the country by 592 World Health Organization (WHO) officers. The availability of real-time data and the distance of health facilities to the community were analyzed.

Results

During the study period, 90,396 health facilities were supervised. The average time spent during supervision varied from 1.53 to 3.78 h across the six geopolitical zones of the country. The average interval between completion of the supervisory checklist and synchronization with the server varied from 3.9 h to 7.5 h. The average distance between the health facility and a ward varied from 5 to 24 km.

Conclusion

The use of handheld devices for supportive supervision provided real-time data from health facilities to state and zonal levels for analysis and feedback. Program officers used the findings to rectify process indicators in time for a better outcome.

     

EFFECTS OF CONCURRENT OZONE EXPOSURE ON THE PATHOGENESIS OF CIGARETTE SMOKE-INDUCED EMPHYSEMA IN B6C3F 1 MICE

Episodic elevation of air pollutants may exacerbate respiratory distress associated with chronic obstructive pulmonary disease (COPD), yet few experiments have been performed to determine how continuously polluted atmospheres may contribute to the etiology of COPD, in general and pulmonary emphysema in particular. This study describes the effects of concurrent exposure to ozone (O 3) in the pathogenesis of cigarette smoke (CS)-induced emphysema in the mouse. Female B6C3F 1 mice were whole-body exposed either to filtered air (FA) or to mainstream CS at a concentration of 250 mg total particulate material/m 3 for 6 h/day, 5 days/wk for 15 or 32 wk. Concurrently, mice were exposed either to FA or to O 3 at 0.3 ppm for 8 h/night, 5 nights/wk for the same time periods. At necropsy, mouse lungs were lavaged, and bronchoalveolar lavage fluid (BALF) was analyzed for inflammatory cell numbers, total protein, lactate dehydrogenase (LDH) and alkaline phosphatase (AP) activities, superoxide production by isolated alveolar macrophages, glutathione content, inflammatory cytokines, and proteolytic activity. Other lungs were inflated at constant pressure for 6 h with formalin for fixation, routine histopathology, and stereology. After 32 wk of exposure, CS with or without concurrent O 3 exposure produced stereologic evidence of emphysema as previously described. Concurrent O 3 exposure did not worsen any of these parameters, nor did O 3 by itself cause stereologic changes that were consistent with emphysema. The O 3 exposure caused only slight elevations of BALF macrophages, while CS exposure caused marked increases in the numbers of both BALF macrophages and neutrophils. Neutrophils in the BALF in response to CS exposure were also more numerous at 32 wk than at 15 wk. Exposure to CS caused an increase in BALF total protein, LDH, AP, and interleukin (IL)-1β. After 32 wk, CS exposure was associated with decreased superoxide production from isolated alveolar macrophages. The CS exposure elevated BALF total glutathione primarily at 15 wk. Overall, O 3 had little effect on endpoints that were significantly affected by CS exposure. We conclude that concurrent O 3 exposure has no effect on the induction of emphysema by CS in this animal model.     

Different Assay Conditions for Detecting the Production and Release of Heat-Labile and Heat-Stable Toxins in Enterotoxigenic Escherichia coli Isolates

Enterotoxigenic Escherichia coli (ETEC) produce heat-labile (LT) and/or heat-stable enterotoxins (ST). Despite that, the mechanism of action of both toxins are well known, there is great controversy in the literature concerning the in vitro production and release of LT and, for ST, no major concerns have been discussed. Furthermore, the majority of published papers describe the use of only one or a few ETEC isolates to define the production and release of these toxins, which hinders the detection of ETEC by phenotypic approaches. Thus, the present study was undertaken to obtain a better understanding of ST and LT toxin production and release under laboratory conditions. Accordingly, a collection of 90 LT-, ST-, and ST/LT-producing ETEC isolates was used to determine a protocol for toxin production and release aimed at ETEC detection. For this, we used previously raised anti-LT antibodies and the anti-ST monoclonal and polyclonal antibodies described herein. The presence of bile salts and the use of certain antibiotics improved ETEC toxin production/release. Triton X-100, as chemical treatment, proved to be an alternative method for toxin release. Consequently, a common protocol that can increase the production and release of LT and ST toxins could facilitate and enhance the sensitivity of diagnostic tests for ETEC using the raised and described antibodies in the present work.