Search Results (2)

In the context of minimally invasive spine surgery, accurately estimating the 3D coordinates of the vertebrae from intraoperative 2D X-ray images is crucial for aligning preoperative data with the patient’s real-time posture. However, existing methods are hindered by the ill-posed nature of 2D-to-3D localization and the distinctive anatomical features of the spinal column, leading to ambiguities and reduced accuracy. In this paper, we introduce X2P-net, a novel prompt-guided and semantic context-enhanced 2D/3D vertebra detection framework. To achieve this, we design a novel Transformer architecture, referred to as BrickFormer, which can automatically extract the refined vertebral foreground context at low computational cost using a dual-attention mechanism. Comprehensive experiments were conducted to validate the proposed approach on two datasets: a large-scale synthetic dataset (BiSpineX) and a sheep spine dataset (SheepSpineX). Results obtained from these experiments demonstrate superior landmark localization performance of the proposed method compared to other state-of-the-art methods. Specifically, on the BiSpineX dataset, X2P-Net achieves percentages of 96.9% and 98.8% at 10 mm and 20 mm thresholds, respectively, a mean position error of 2.99 mm, and an $AUC$ of 0.9923. Similar superior performance was also observed when the proposed method was applied to the SheepSpineX dataset, with percentages of 98.4% and 100.0% at 10 mm and 20 mm thresholds, respectively, a mean position error of 1.08 mm, and an $AUC$ of 0.9972. Full article

The real-time monitoring of animal postures through computer vision techniques has become essential for modern precision livestock management. To overcome the limitations of current behavioral analysis systems in balancing computational efficiency and detection accuracy, this study develops an optimized deep learning framework named YOLOv8-BCD specifically designed for ovine posture recognition. The proposed architecture employs a multi-level lightweight design incorporating enhanced feature fusion mechanisms and spatial-channel attention modules, effectively improving detection performance in complex farm environments with occlusions and variable lighting. Our methodology introduces three technical innovations: (1) Adaptive multi-scale feature aggregation through bidirectional cross-layer connections. (2) Context-aware attention weighting for critical region emphasis. (3) Streamlined detection head optimization for resource-constrained devices. The experimental dataset comprises 1476 annotated images capturing three characteristic postures (standing, lying, and side lying) under practical farming conditions. Comparative evaluations demonstrate significant improvements over baseline models, achieving 91.7% recognition accuracy with 389 FPS processing speed while maintaining 19.2% parameter reduction and 32.1% lower computational load compared to standard YOLOv8. This efficient solution provides technical support for automated health monitoring in intensive livestock production systems, showing practical potential for large-scale agricultural applications requiring real-time behavioral analysis. Full article