arXiv subjects

Christoph Graeni

Publications and source records attributed to Christoph Graeni.

3 recordsLinked to original sources

arXiv2026

Fluid dynamics informed CCTA-derived geometric parameters in right coronary artery anomalies predict abnormal invasive Adenosine FFR and Dobutamine FFR

Background: Right anomalous aortic origin of coronary arteries (R-AAOCA) involves fixed compression, assessable with adenosine-derived fractional flow reserve (FFRAdnosine), and additional stress-induced dynamic compression captured by dobutamine-derived FFR (FFRDobutamine). We hypothesized that coronary CT angiography (CCTA)-derived fluid dynamics-informed parameters outperform conventional metrics in predicting both FFR types. Methods: We retrospectively analyzed CCTA data from R-AAOCA patients who underwent invasive FFRAdnosine and FFRDobutamine assessment. Parameters were categorized as: (1) conventional metrics (cross-sectional area, perimeter, minor/major axis, intramural lumen area [ILA], effective diameter, area and diameter stenosis ratios) and (2) fluid dynamics-informed metrics (hydraulic diameter, elliptic ratio, circularity, hydraulic diameter stenosis ratio, resistance index [RI], ostial angulation penalty [OAP], and comprehensive stenosis score [CSS]). Hemodynamic relevance was defined as FFR$\leq$0.80. Results: 81 patients were included. FFRAdnosine$\leq$0.80 occurred in 5 (6.2%) and FFRDobutamine$\leq$0.80 in 16 (19.8%) patients. For FFRAdnosine, top discriminators were RI (AUC=0.97), OAP (AUC=0.97), ostial area (AUC=0.96), and CSS (AUC=0.95). For FFRDobutamine, ostial minor diameter led (AUC=0.85), followed by RI (AUC=0.83) and ILA minor diameter (AUC=0.81). RI explained 45% and 43% of FFRAdnosine and FFRDobutamine variance, respectively. At optimal thresholds, RI achieved 100% sensitivity and 95% specificity for FFRAdnosine; ostial minor diameter achieved 100% sensitivity and 57% specificity for FFRDobutamine. Conclusions: In R-AAOCA, CCTA-derived fluid dynamics-informed metrics provide excellent and superior performance compared with conventional geometric parameters in predicting hemodynamic relevance of fixed compression.

physics.med-ph↗

arXiv2025

multimodars: A Rust-powered toolkit for multi-modality cardiac image fusion and registration

Combining complementary imaging modalities is critical to build reliable 3D coronary models: intravascular imaging gives sub-millimetre resolution but limited whole-vessel context, while CCTA supplies 3D geometry but suffers from limited spatial resolution and artefacts (e.g., blooming). Prior work demonstrated intravascular/CCTA fusion, yet no open, flexible toolkit is tailored for multi-state analysis (rest/stress, pre-/post-stenting) while offering deterministic behaviour, high performance, and easy pipeline integration. multimodars addresses this gap with deterministic alignment algorithms, a compact NumPy-centred data model, and an optimised Rust backend suitable for scalable, reproducible experiments. The package accepts CSV/NumPy inputs including data formats produced by the AIVUS-CAA software

cs.CV↗

arXiv2025

From Claims to Evidence: A Unified Framework and Critical Analysis of CNN vs. Transformer vs. Mamba in Medical Image Segmentation

While numerous architectures for medical image segmentation have been proposed, achieving competitive performance with state-of-the-art models networks such as nnUNet, still leave room for further innovation. In this work, we introduce nnUZoo, an open source benchmarking framework built upon nnUNet, which incorporates various deep learning architectures, including CNNs, Transformers, and Mamba-based models. Using this framework, we provide a fair comparison to demystify performance claims across different medical image segmentation tasks. Additionally, in an effort to enrich the benchmarking, we explored five new architectures based on Mamba and Transformers, collectively named X2Net, and integrated them into nnUZoo for further evaluation. The proposed models combine the features of conventional U2Net, nnUNet, CNN, Transformer, and Mamba layers and architectures, called X2Net (UNETR2Net (UNETR), SwT2Net (SwinTransformer), SS2D2Net (SwinUMamba), Alt1DM2Net (LightUMamba), and MambaND2Net (MambaND)). We extensively evaluate the performance of different models on six diverse medical image segmentation datasets, including microscopy, ultrasound, CT, MRI, and PET, covering various body parts, organs, and labels. We compare their performance, in terms of dice score and computational efficiency, against their baseline models, U2Net, and nnUNet. CNN models like nnUNet and U2Net demonstrated both speed and accuracy, making them effective choices for medical image segmentation tasks. Transformer-based models, while promising for certain imaging modalities, exhibited high computational costs. Proposed Mamba-based X2Net architecture (SS2D2Net) achieved competitive accuracy with no significantly difference from nnUNet and U2Net, while using fewer parameters. However, they required significantly longer training time, highlighting a trade-off between model efficiency and computational cost.

eess.IV↗