A novel transfer learning approach for the classification of histological images of colorectal cancer

Colorectal cancer (CRC) is the second most diagnosed cancer in the United States. It is identified by histopathological evaluations of microscopic images of the cancerous region, relying on a subjective interpretation. The Colorectal Histology dataset used in this study contains 5000 images, made available by the University Medical Center Mannheim. This approach proposes the automatic identification of eight types of tissues found in CRC histopathological evaluation. We apply Transfer Learning from architectures of Convolutional Neural Networks (CNNs). We modify the structures of CNNs to extract features from the images and input them to well-known machine learning methods: Naive Bayes, Multilayer Perceptron, k-Nearest Neighbors, Random Forest, and Support Vector Machine (SVM). We evaluated 108 extractor–classifier combinations. The one that achieved the best results is DenseNet169 with SVM (RBF), reaching an Accuracy of 92.083% and F1-Score of 92.117%. Therefore, our approach is capable of distinguishing tissues found in CRC histopathological evaluation.

     

Indoor air quality assessment in dwellings with different ventilation strategies in Nunavik and impacts on bacterial and fungal microbiota

Indoor air quality is a major issue for public health, particularly in northern communities. In this extreme environment, adequate ventilation is crucial to provide a healthier indoor environment, especially in airtight dwellings. The main objective of the study is to assess the impact of ventilation systems and their optimization on microbial communities in bioaerosols and dust in 54 dwellings in Nunavik. Dwellings with three ventilation strategies (without mechanical ventilators, with heat recovery ventilators, and with energy recovery ventilators) were investigated before and after optimization of the ventilation systems. Indoor environmental conditions (temperature, relative humidity) and microbiological parameters (total bacteria, Aspergillus/Penicillium, endotoxin, and microbial biodiversity) were measured. Dust samples were collected in closed face cassettes with a polycarbonate filter using a micro-vacuum while a volume of 20 m³ of bioaerosols were collected on filters using a SASS3100 (airflow of 300 L/min). In bioaerosols, the median number of copies was 4.01 × 10³ copies/m³ of air for total bacteria and 1.45 × 10¹ copies/m³ for Aspergillus/Penicillium. Median concentrations were 5.13 × 10⁴ copies/mg of dust, 5.07 × 10¹ copies/mg, 9.98 EU/mg for total bacteria, Aspergillus/Penicillium and endotoxin concentrations, respectively. The main microorganisms were associated with human occupancy such as skin-related bacteria or yeasts, regardless of the type of ventilation.     

Effects of Electrospinning Parameters on Gelatin/Poly(ϵ‐Caprolactone) Nanofiber Diameter

An electrospinning procedure was carried out to fabricate gelatin/poly(?‐caprolactone) (Gt/PCL) nanofibers. Response surface methodology based on a three‐level, four‐variable Box‐Behnken design technique was used to model the resultant diameter of the as‐spun nanofibers. A second‐order model was obtained to describe the relationship between the fiber diameter and the electrospinning parameters, namely Gt concentration, PCL concentration, content of acetic acid in the overall solvent, and content of Gt solution in the blend solution. The individual and the interactive effects of these parameters on the fiber diameter were determined. Validation experiments verified the accuracy of the model which provided a simple and effective method for fabricating nanofibers with a controllable and predictable fiber diameter.     

A pipeline structure for the block QR update in digital signal processing

The Journal of Supercomputing - There exist problems in the field of digital signal processing, such as filtering of acoustic signals that require processing a large amount of data in real time....