基於不確定性感知測試時間適應的逆一致性微分同胚肺部影像配準
本文提出了一種基於不確定性感知測試時間適應的框架，用於逆一致性微分同胚肺部影像配準，利用蒙特卡洛 dropout 估計空間不確定性，以提高模型在處理大變形時的配準精度。
逆整合性を保つ微分同相写像を用いた肺画像レジストレーションのための、不確実性認識型テスト時適応
本論文では、吸気時と呼気時の胸部CTスキャンのレジストレーション精度を向上させるため、逆整合性を保つ微分同相写像を用いた、不確実性認識型テスト時適応フレームワークを提案する。
Uncertainty-Aware Test-Time Adaptation Improves Inverse Consistent Diffeomorphic Lung Image Registration, Especially for Large Deformations
Incorporating uncertainty awareness into deep learning-based diffeomorphic lung image registration significantly improves accuracy, particularly in cases with large deformations between inspiratory and expiratory lung volumes, enabling more reliable and robust registration for both forward (TLC to FRC) and inverse (FRC to TLC) transformations.
医療セグメンテーションAIアルゴリズム評価のための新たなベンチマーク：Touchstoneの紹介と評価結果
標準的なベンチマークの限界を克服するため、大規模で多様なデータセットと包括的な評価手法を用いた医療セグメンテーションAIのためのベンチマーク「Touchstone」を開発し、その有効性を示した。
AtlasSeg: Enhancing Fetal Brain MRI Segmentation Using Atlas Priors and Dual-U-Net Architecture with Multi-Scale Attentive Fusion
AtlasSeg, a novel deep learning model, leverages gestational age-specific atlas priors and a dual-U-Net architecture with multi-scale attentive fusion to significantly improve the accuracy of cortical segmentation in fetal brain MRI, outperforming existing state-of-the-art methods.
マルチスケールネスト型 Residual UNetに基づく肺CTにおける多段階気道セグメンテーション
本稿では、複雑な気道構造を正確にセグメント化するため、マルチスケールネスト型 Residual UNet（MNR-UNet）と重み付きBreakage-Aware Loss（wBAL）を組み合わせた新しい3段階セグメンテーション手法を提案する。
Enhancing Pancreatic Tumor Segmentation through Optimized Synthetic Data Generation
Strategically selecting a combination of synthetic tumor sizes and generating synthetic tumors with precise boundaries significantly improves the accuracy of deep learning-based pancreatic tumor segmentation models.
Quantitative Analysis of Cervical Spinal Cord Microstructure and Macrostructure Using Deep Learning-based Segmentation of Magnetic Resonance Images
This research aims to investigate the relationships between microstructural and macrostructural features of the cervical spinal cord in a healthy population using quantitative MRI analysis, and to develop a deep learning-based segmentation framework for accurate measurement of macrostructural characteristics.
Pathological Clue-driven Representation Learning for Accurate Brain CT Report Generation
Leveraging diverse pathological clues, including segmented regions, entities, and report themes, to build fine-grained cross-modal representations and seamlessly transfer them to enhance the quality of generated brain CT reports.
Cascading Refinement CNN for Semantic Segmentation of Myocardial Pathologies from Multi-Sequence Cardiac MRI
A cascading refinement CNN model, MS-CaRe-CNN, can semantically segment the left and right ventricle, healthy and scarred myocardial tissue, as well as edema from multi-sequence cardiac MRI data, enabling accurate assessment of myocardial viability.
© 2024 by Linnk AI