Environmental impacts and effects on greenhouse gas emissions in shrimp feed production system for aquaculture - A case study in India.

Attributional life cycle assessment study examines the environmental impact of raw materials, machinery, and unit operations. In the present work, an attributional life cycle assessment (LCA) was employed to assess the environmental and greenhouse gas impacts of a shrimp feed production system. A commercial shrimp feed mill in Tamil Nadu, India, provided inventory data for one-ton shrimp feed (functional unit) for a Cradle-to-Gate evaluation using environmental impact methodologies, specifically Impact 2002+ in SimaPro® (V9.3.0.3) software. The results showed that human health (0.003357 DALY), ecosystem quality (2720.518 PDF × m2 × yr), climate change (2031.696 kg CO2 eq), and resources (71019.42 MJ primary) were the most significantly impacted. The human health category was found to be the most prominent after normalization and weighting (0.47 pt), and strategies were suggested accordingly. The GWP20 and GWP100 measures for long-term climate change were calculated to be 8.7 and 7.33 kg CO2 eq, respectively. Cast iron used in machinery production (GWP 20-15.40%, GWP100-134.5%) and electricity use (GWP 20-6.13%, GWP 100-6.9%) accounted for sizable portions of the burden. Feed production is estimated to contribute 0.2% of global CO2 emissions within the proposed global context. These findings are significant regarding economically and environmentally sustainable shrimp feed production worldwide.

Biceph-Net: A Robust and Lightweight Framework for the Diagnosis of Alzheimer's Disease Using 2D-MRI Scans and Deep Similarity Learning.

Alzheimer's Disease (AD) is a neurodegenerative disease that is one of the significant causes of death in the elderly population. Many deep learning techniques have been proposed to diagnose AD using Magnetic Resonance Imaging (MRI) scans. Predicting AD using 2D slices extracted from 3D MRI scans is challenging as the inter-slice information gets lost. To this end, we propose a novel and lightweight framework termed 'Biceph-net' for AD diagnosis using 2D MRI scans that models both the intra-slice and inter-slice information. 'Biceph-net' has been experimentally shown to perform similar to other Spatio-temporal neural networks while being computationally more efficient. Biceph-net is also superior in performance compared to vanilla 2D convolutional neural networks (CNN) for AD diagnosis using 2D MRI slices. Biceph-net also has an inbuilt neighbourhood-based model interpretation feature that can be exploited to understand the classification decision taken by the network. Biceph-net experimentally achieves a test accuracy of 100% in the classification of Cognitively Normal (CN) vs AD, 98.16% for Mild Cognitive Impairment (MCI) vs AD, and 97.80% for CN vs MCI vs AD.