Kitware (United States)

The fovea is a unique specialization of the primate retina and is a promising site for obtaining the first complete connectome of a human central nervous system (CNS) structure. Within the fovea, neural cells and circuits have been miniaturized and compressed during evolution to sample the visual image at highest spatial resolution and begin the neural processing that serves human form, color, and motion perception. Here we present a comprehensive analysis of a sample of human foveal retina using deep learning-based segmentation to reconstruct all cells and synaptic connections at nanoscale resolution. We classified ~3,000 cells into 51 distinct morphological types based on their structural features and connectivity patterns. Our observations reveal novel synaptic pathways absent in non-human primates, suggesting specialized circuits contribute uniquely to human trichromatic color vision. A biophysical model of the distinct connectomes made by gap junctions (electrical synapses) between short- (S) and medium-long- (ML) wavelength-sensitive cone photoreceptors, suggests chromatic interactions between S and ML cones prior to the first chemical synapse. Segmentation of retinal ganglion cells (RGCs) suggests the presence of only 11 visual pathways, with 5 high-density RGC pathways accounting for over 95% of foveal output to the brain: a dramatic contrast to the 40+ ganglion cell types recognized in mouse retina. Our connectomic analysis reveals distinctive features of human neural circuitry and demonstrates how AI-based computational approaches can advance understanding of human brain structure and function.

This study simulates and analyzes a dual-band microstrip patch antenna designed to be optimal for Global Navigation Satellite System usage in the L2 (1227.60 MHz) and L5 (1176.45 MHz) bands. This research focuses on designing a single antenna element to receive signals from multiple Global Navigation Satellite System (GNSS) constellations operating over the L2 and L5 frequency bands. The antenna design presented herein utilizes a novel technique that combines a slotted patch configuration with a defective ground plane to achieve dual-band functionality while maintaining the miniaturization and low-profile characteristics of the structure. Full-wave numerical simulations were performed using a finite element-based electromagnetic solver to fine-tune various performance parameters, such as S11 bandwidths, radiation patterns, and realized gain of this proposed microstrip slot-patch antenna structure implemented on a high-frequency electronic material substrate. The effects of different design factors, including substrate selection and characteristics, as well as variation in the length-width ratio of the rectangular patch strip, are also investigated, and their influence on key antenna performance metrics is analyzed. This study contributes to the field of GNSS antenna design by presenting a cost-effective and efficient dual-band solution that can be easily integrated into modern navigation devices. In addition, this study explains how much more gain the antennas in the L2 and L5 bands of the GNSS provided in the antenna array with different feeding techniques.

BackgroundMedial temporal lobe (MTL) atrophy measured on MRI is a sensitive biomarker of AD‐linked neurodegeneration but is not specific to AD. LATE is a common AD co‐pathology that is challenging to distinguish from AD based on available validated biomarkers. LATE and AD are both associated with hippocampus and entorhinal cortex atrophy, but recent studies suggest that there are differences in the severity and spatial patterns of atrophy. We sought to use postmortem MRI and histology to examine how MRI morphometric measures in AD and LATE relate to direct measures of neurodegeneration, including neuron number, size, and density.MethodThionin‐stained 50µm histology sections from 24 brain donors (5 AD‐LATE‐, 14 AD+LATE‐, 5 AD+LATE+) with available 9.4T postmortem MRI were used to delineate subfields CA1, subiculum and entorhinal cortex. Deep learning method StarDist was used to detect star‐shaped objects in thionine slides, combined with weakly‐supervised learning to identify artifact‐free cortical regions and Gaussian mixture modeling to distinguish neurons from glia (Figure 1AB). Validation against stereology measures of neuronal/glial density was performed in 50 separate regions. Neuronal measures were compared to MTL volume and thickness measures extracted from MRI.ResultAutomated neuronal density estimates (r=0.72, Figure 1C) agreed with stereology, but not glial density. Estimated number and size of neurons in CA1 and ERC were higher in individuals with less CA1/ERC atrophy (Figure 2) consistent with greater neuronal loss in advanced AD and LATE. However, CA1 neuronal and glial density, and ERC glial density, were higher in individuals with greater atrophy, suggesting tighter packing of neurons in remaining tissue and significant contribution of neuropil loss to MRI‐based atrophy measures. Pointwise analysis in Figure 3 shows patterns of association between regional MTL thickness and CA1 neuronal count, size, and density measures.ConclusionThese initial feasibility results in two sections in 24 brain donors encourage us to apply this pipeline to a larger dataset of paired postmortem MRI and serial histology sections (Ravikumar et al., 2024) and to study associations between tau pathology, neuronal loss, and MRI‐based measures of atrophy, with the aim of better differentiating atrophy linked to LATE and AD.

Purpose This study aims to investigate the ethical and practical implications of delegating decision-making to AI systems, focusing on the necessity for a robust governance framework. Specifically, it examines how autonomy, transparency and accountability within AI governance influence organizational decision-making. Design/methodology/approach Employing a quantitative survey methodology, this study gathered data from 452 business owners and managers in Indian IT companies. The questionnaire was disseminated using online platforms and departmental communication channels. Structural equation modelling (SEM) was utilized for data analysis, allowing for the examination of relationships among autonomy, transparency, accountability and decision-making. Findings The findings indicate that autonomy, transparency and accountability significantly impact organizational decision-making processes. Specifically, autonomy and accountability were found to directly influence decision-making, while transparency also played a crucial role. Additionally, social innovation was identified as a significant moderating factor, enhancing the relationship between AI governance and decision-making outcomes. Originality/value This research contributes to the existing literature on AI governance by elucidating the critical role of ethical frameworks in organizational decision-making. By incorporating social innovation as a moderating variable, the study offers novel insights into how AI governance can be optimized to enhance decision-making processes. The application of SEM provides a rigorous analytical approach, facilitating a deeper understanding of the interplay between governance dimensions and decision-making outcomes. The findings have practical implications for organizations seeking to implement effective AI governance strategies in their operations.

The tourism sector, which is an important contributor to global economic development, in Kolkata, a vibrant city full of cultural vibrancy, has undergone a deep transformation with the advent of artificial intelligence (AI). This paper discusses the impact of AI on the tourism industry in Kolkata using the strategic management perspective. The research, based on empirical data and a comprehensive literature review, explores how AI technologies are changing consumer experiences and operations in the tourism business, such as planning, bookings, and experiencing travel by tourists in Kolkata’s tourism industry. In this context, findings will be presented to show that AI improves operational efficiency, customer satisfaction, and safety in the post-pandemic era, making it an important tool for sustainable tourism development. However, the integration has its challenges, such as workforce displacement and technology that needs to be sustained. More importantly, AI has a double role in developing community-based tourism enterprises (CBTEs) and large-scale tourism projects in order to balance growth with social equity and to maximize resource use and operations costs. The study contributes to the ongoing debate on AI adoption in tourism and provides strategic insights for policymakers and industry stakeholders in Kolkata to leverage the potential of AI.

This paper introduces an advanced system for Accent Conversion (AC) that enhances the transformation of nonnative English speech to a native-accented equivalent while meticulously preserving the speaker's unique identity. Integrating a Speaker Encoder, an Accent Encoder, and a Voice Conversion model, our system enables real-time accent conversion with minimal processing delays. Unlike traditional AC methods which require native-accented input and suffer from lengthy processing times, our approach operates independently of native utterances, leveraging recent advances in speaker-independent phonetic posteriorgrams for accent conversion. Trained on an extensive dataset encompassing a diverse range of non-native and native English accents, our model proficiently captures intricate phonetic and prosodic nuances, thus facilitating highly accurate native accent synthesis. The proposed system is distinguished by its operational immediacy and reduced computational demands, rendering it a viable solution for immediate application scenarios.

We introduce KAMERA: a comprehensive system for multi-camera, multi-spectral synchronization and real-time detection of seals and polar bears. Utilized in aerial surveys for ice-associated seals in the Bering, Chukchi, and Beaufort seas around Alaska, KAMERA provides up to an 80% reduction in dataset processing time over previous methods. Our rigorous calibration and hardware synchronization enable using multiple spectra for object detection. All collected data are annotated with metadata so they can be easily referenced later. All imagery and animal detections from a survey are mapped onto a world plane for accurate surveyed area estimates and quick assessment of survey results. We hope KAMERA will inspire other mapping and detection efforts in the scientific community, with all software, models, and schematics fully open-sourced.