A Deep Learning-Based Decision Support Tool for Precision Risk Assessment of Breast Cancer.

PURPOSE: The Breast Imaging Reporting and Data System (BI-RADS) lexicon was developed to standardize mammographic reporting to assess cancer risk and facilitate the decision to biopsy. Because of substantial interobserver variability in the application of the BI-RADS lexicon, the decision to biopsy varies greatly and results in overdiagnosis and excessive biopsies. The false-positive rate from mammograms is estimated to be 7% to approximately 10% overall, but within the BI-RADS 4 category, it is greater than 70%. Therefore, we developed the Breast Cancer Risk Calculator (BRISK) to target a well-characterized and specific patient subgroup (BI-RADS 4) rather than a broad heterogeneous group in assessing breast cancer risk. METHODS: BRISK provides a novel precise risk assessment model to reduce overdiagnosis and unnecessary biopsies. It was developed by applying natural language processing and deep learning methods on 5,147 patient records archived in the Houston Methodist systemwide data warehouse from 2006 to May 2015, including imaging and pathology reports, mammographic images, and patient demographics. Key characteristics for BI-RADS 4 patients were collected and computed to output an index measure for biopsy recommendation that is clinically relevant and informative and improves upon the traditional BI-RADS 4 scores. RESULTS: For the validation set, we assessed data from 1,247 BI-RADS 4 patients, including mammographic images and medical reports. The BRISK model sensitivity to predict malignancy was 100%, whereas the specificity was 74%. The total accuracy of our implemented model in BRISK was 81%. Overall area under the curve was 0.93. CONCLUSION: BRISK for abnormal mammogram uses integrative artificial intelligence technology and has demonstrated high sensitivity in the prediction of malignancy. Prospective evaluation is under way and can lead to improvement in patient-physician engagement in making informed decisions with regard to biopsy.

Assessment on reticuloendotheliosis virus infection in specific-pathogen-free chickens based on detection of yolk antibody.

Reticuloendotheliosis virus (REV) is the most frequent exogenous virus that contaminates attenuated vaccines. Therefore, it is extremely important to select REV-free specific-pathogen-free (SPF) chicken embryos. Generally, REV infection is assessed by detecting REV antibodies in SPF chickens. This present study seeks to evaluate REV infection by replacing serum antibody detection with yolk antibody detection. A cohort of 40 nineteen-week-old SPF chickens were artificially inoculated with REV, with 32 SPF chickens raised in another isolation environment served as a blank control. Eggs and serum from 23-week-old chickens were sampled, and yolks were diluted separately to ratios of 1:150, 1:200, 1:300 and 1:400, which were detected together with serum. We found that the yolk antibody detection findings at a dilution of 1:300 had the highest coincidence rate compared with that based on serum antibody measurements. At a dilution ratio of 1:300 for yolk antibody, 72 chickens were continuously observed for 10 weeks from 25- to 34-weeks-old. Our findings were based on serum antibody or yolk antibody detection, and the evaluation results were completely consistent. Therefore, all serum antibody-positive chickens were yolk antibody-positive, and vice versa. Accordingly, vaccine producers can estimate REV cleanliness in a poultry farm by sampling yolk antibody titers.

PM2.5 Cooperative Control with Fuzzy Cost and Fuzzy Coalitions.

Haze control cost is hard to value by a crisp number because it is often affected by various factors such as regional uncertain meteorological conditions and topographical features. Furthermore, regions may be involved in different coalitions for haze control with different levels of effort. In this paper, we propose a PM2.5 cooperative control model with fuzzy cost and crisp coalitions or fuzzy coalitions based on the uncertain cross-border transmission factor. We focus on the Beijing­Tianjin­Hebei regions of China and obtain the following major findings. In the case of haze control in the Beijing­Tianjin­Hebei regions of China, local governments in the global crisp coalition can achieve their emission reduction targets with the lowest aggregated cost. However, Hebei fails to satisfy its individual rationality if there is no cost sharing. Therefore, the Hukuhara­Shapley value is used to allocate the aggregated cost among these regions so that the grand coalition is stable. However, the Beijing­Tianjin­Hebei regions cannot achieve their emission reduction targets in the global fuzzy coalition without government subsidies.

Grand Canonical Monte Carlo Simulations on Phase Equilibria of Methane, Carbon Dioxide, and Their Mixture Hydrates.

The cage occupancy plays a crucial role in the thermodynamic stability of clathrate hydrates and is an important quantity for understanding the CO2-CH4 replacement phenomenon. In this work, the occupancy isotherms of pure CH4, pure CO2, and their mixture in sI and sII hydrates are studied by GCMC + MD simulations. The adsorption of CH4 and CO2 + CH4 in the sI and sII hydrates can be categorized as the one-site Langmuir type. The calculated occupancy ratio θL/θS and the abundance ratio of CO2 to CH4 vary with the temperature and pressure, which provide the prerequisite information for the prediction of CH4 recovery yield at different conditions in the CO2-CH4 gas exchange process. The phase equilibria of clathrate hydrates of pure gases and mixtures are explored and the corresponding heat of dissociation and hydration numbers are determined. The current investigation provides new perspectives to understand the mechanism behind the gas adsorption behavior of clathrate hydrates.