Efficient breadth first search on multi-GPU systems

Simple algorithms for the execution of a Breadth First Search on large graphs lead, running on clusters of GPUs, to a situation of load unbalance among threads and un-coalesced memory accesses, resulting in pretty low performances. To obtain a significant improvement on a single GPU and to scale by using multiple GPUs, we resort to a suitable combination of operations to rearrange data before processing them. We propose a novel technique for mapping threads to data that achieves a perfect load balance by leveraging prefix-sum and binary search operations. To reduce the communication overhead, we perform a pruning operation on the set of edges that needs to be exchanged at each BFS level. The result is an algorithm that exploits at its best the parallelism available on a single GPU and minimizes communication among GPUs. We show that a cluster of GPUs can efficiently perform a distributed BFS on graphs with billions of nodes.     

Effectiveness of Semi-Supervised Active Learning in Automated Wound Image Segmentation

Appropriate wound management shortens the healing times and reduces the management costs, benefiting the patient in physical terms and potentially reducing the healthcare system’s economic burden. Among the instrumental measurement methods, the image analysis of a wound area is becoming one of the cornerstones of chronic ulcer management. Our study aim is to develop a solid AI method based on a convolutional neural network to segment the wounds efficiently to make the work of the physician more efficient, and subsequently, to lay the foundations for the further development of more in-depth analyses of ulcer characteristics. In this work, we introduce a fully automated model for identifying and segmenting wound areas which can completely automatize the clinical wound severity assessment starting from images acquired from smartphones. This method is based on an active semi-supervised learning training of a convolutional neural network model. In our work, we tested the robustness of our method against a wide range of natural images acquired in different light conditions and image expositions. We collected the images using an ad hoc developed app and saved them in a database which we then used for AI training. We then tested different CNN architectures to develop a balanced model, which we finally validated with a public dataset. We used a dataset of images acquired during clinical practice and built an annotated wound image dataset consisting of 1564 ulcer images from 474 patients. Only a small part of this large amount of data was manually annotated by experts (ground truth). A multi-step, active, semi-supervised training procedure was applied to improve the segmentation performances of the model. The developed training strategy mimics a continuous learning approach and provides a viable alternative for further medical applications. We tested the efficiency of our model against other public datasets, proving its robustness. The efficiency of the transfer learning showed that after less than 50 epochs, the model achieved a stable DSC that was greater than 0.95. The proposed active semi-supervised learning strategy could allow us to obtain an efficient segmentation method, thereby facilitating the work of the clinician by reducing their working times to achieve the measurements. Finally, the robustness of our pipeline confirms its possible usage in clinical practice as a reliable decision support system for clinicians.     

In-House Validation and Comparison of Two Wheat (<Emphasis Type="Italic">Triticum aestivum</Emphasis>) Taxon-Specific Real-Time PCR Methods for GMO Quantification Supported by Droplet Digital PCR

Common wheat is one of the most important staple food crops worldwide. However, unlike other important staple crops such as maize or soybean, genetically modified (GM) wheat is not yet present in the global food market. Nonetheless, in the recent past, the adventitious presence of GM glyphosate-tolerant volunteers was reported in open wheat fields in the USA. The European Union Reference Laboratory for GM Food and Feed (EURL-GMFF) was therefore called to develop a strategy to detect such unauthorised GM wheat in wheat samples by using both taxon-specific and screening tests. Two candidate common wheat taxon-specific real-time PCR methods were suggested, one targeting ssII-D gene coding for starch synthase and the other targeting waxy-D1 gene, coding for granule-bound starch synthase. In the present study, the two above-mentioned real-time PCR taxon-specific methods were in-house verified and compared, proposing droplet digital PCR (ddPCR) as a new tool for supporting the application of the European Network of GMO Laboratories (ENGL) established method performance criteria. Preliminary performance data of waxy-D1 and ssII-D methods in ddPCR format are shown too to give a contribution to the bridging process from the consolidated to the emerging quantitative PCR methodology.     

DNA barcoding as a new tool for food traceability

Food safety and quality are nowadays a major concern. Any case of food alteration, especially when reported by the media, has a great impact on public opinion. There is an increasing demand for the improvement of quality controls, hence addressing scientific research towards the development of reliable molecular tools for food analysis. DNA barcoding is a widely used molecular-based system, which can identify biological specimens, and is used for the identification of both raw materials and processed food. In this review the results of several researches are critically analyzed, in order to exploit the effectiveness of DNA barcoding in food traceability, and to delineate some best practices in the application of DNA barcoding throughout the industrial pipeline. The use of DNA barcoding for food safety and in the identification of commercial fraud is also discussed.     

Functional properties of reduced moisture fruits as ingredients in food systems

Fruit pieces of some species have been processed to be used as ingredients in various food systems, either by partial dehydration alone or by osmosis and partial dehydration, to achieve different levels of water activity and solids contents in the final products.

The relationships existing among processing, phase composition after processing and functional properties of the products have been assessed within a useful range of water content and water activity.

The functional properties are expressed as diagrams relating the phase composition (soluble solids, insoluble solids and water) to the range of consistencies obtainable at various water activities. These diagrams provide a useful tool for preparing fruit ingredients suitable for specific food systems.     

Superior Cerebellar Atrophy: An Imaging Clue to Diagnose ITPR1-Related Disorders

The inositol 1,4,5-triphosphate receptor type 1 (ITPR1) gene encodes an InsP₃-gated calcium channel that modulates intracellular Ca²⁺ release and is particularly expressed in cerebellar Purkinje cells. Pathogenic variants in the ITPR1 gene are associated with different types of autosomal dominant spinocerebellar ataxia: SCA15 (adult onset), SCA29 (early-onset), and Gillespie syndrome. Cerebellar atrophy/hypoplasia is invariably detected, but a recognizable neuroradiological pattern has not been identified yet. With the aim of describing ITPR1-related neuroimaging findings, the brain MRI of 14 patients with ITPR1 variants (11 SCA29, 1 SCA15, and 2 Gillespie) were reviewed by expert neuroradiologists. To further evaluate the role of superior vermian and hemispheric cerebellar atrophy as a clue for the diagnosis of ITPR1-related conditions, the ITPR1 gene was sequenced in 5 patients with similar MRI pattern, detecting pathogenic variants in 4 of them. Considering the whole cohort, a distinctive neuroradiological pattern consisting in superior vermian and hemispheric cerebellar atrophy was identified in 83% patients with causative ITPR1 variants, suggesting this MRI finding could represent a hallmark for ITPR1-related disorders.