A Multi-Scale Inception-UNet with Structure-Aware Evaluation for Branch-Preserving Segmentation of Organoids

Sandra H. Andrusca, Christopher D. Kießling, Andreas R. Bausch
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:4071-4084, 2026.

Abstract

Branched organoids exhibit increasingly complex morphologies as they progress from simple spheroid states to highly ramified structures, making topology-preserving segmentation essential for quantitative biological analysis. Capturing thin protrusions and maintaining branch continuity remains challenging for classical UNet-based architectures, particularly in brightfield imaging where fine structures are easily blurred or disconnected. In this work, we present a multi-scale Inception-UNet designed to capture the heterogeneous spatial scales of branched organoids through parallel convolutional paths with complementary receptive fields. As a model system, we analyze brightfield pancreatic ductal adenocarcinoma (PDAC) organoids, a system known for strong morphological heterogeneity and invasive branching behavior randriamanantsoa2022pdacOrganoids, cultured using high-throughput Patternoid assays kurzbach2025patternoid that enable standardized imaging and robust quantitative analysis. To assess segmentation quality beyond region overlap, we combine Dice with the structure-aware clDice metric that directly probes branch integrity and topological continuity. Across deterministic seeds and strictly separated organoid positions, the Inception-UNet achieves the highest region-based Dice ($0.868 \pm 0.062$) and clDice ($0.545 \pm 0.123$), and most importantly, the strongest preservation of branch continuity compared to UNet and UNet++. These improvements become increasingly pronounced with growing morphological complexity. Overall, our results demonstrate that multi-scale feature extraction combined with topology-aware evaluation substantially improves segmentation of branched organoids and provides a robust foundation for downstream morphological and invasion-related analyses.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-andrusca26a, title = {A Multi-Scale Inception-UNet with Structure-Aware Evaluation for Branch-Preserving Segmentation of Organoids}, author = {Andrusca, Sandra H. and Kie{\ss}ling, Christopher D. and Bausch, Andreas R.}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {4071--4084}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/andrusca26a/andrusca26a.pdf}, url = {https://proceedings.mlr.press/v315/andrusca26a.html}, abstract = {Branched organoids exhibit increasingly complex morphologies as they progress from simple spheroid states to highly ramified structures, making topology-preserving segmentation essential for quantitative biological analysis. Capturing thin protrusions and maintaining branch continuity remains challenging for classical UNet-based architectures, particularly in brightfield imaging where fine structures are easily blurred or disconnected. In this work, we present a multi-scale Inception-UNet designed to capture the heterogeneous spatial scales of branched organoids through parallel convolutional paths with complementary receptive fields. As a model system, we analyze brightfield pancreatic ductal adenocarcinoma (PDAC) organoids, a system known for strong morphological heterogeneity and invasive branching behavior randriamanantsoa2022pdacOrganoids, cultured using high-throughput Patternoid assays kurzbach2025patternoid that enable standardized imaging and robust quantitative analysis. To assess segmentation quality beyond region overlap, we combine Dice with the structure-aware clDice metric that directly probes branch integrity and topological continuity. Across deterministic seeds and strictly separated organoid positions, the Inception-UNet achieves the highest region-based Dice ($0.868 \pm 0.062$) and clDice ($0.545 \pm 0.123$), and most importantly, the strongest preservation of branch continuity compared to UNet and UNet++. These improvements become increasingly pronounced with growing morphological complexity. Overall, our results demonstrate that multi-scale feature extraction combined with topology-aware evaluation substantially improves segmentation of branched organoids and provides a robust foundation for downstream morphological and invasion-related analyses.} }
Endnote
%0 Conference Paper %T A Multi-Scale Inception-UNet with Structure-Aware Evaluation for Branch-Preserving Segmentation of Organoids %A Sandra H. Andrusca %A Christopher D. Kießling %A Andreas R. Bausch %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-andrusca26a %I PMLR %P 4071--4084 %U https://proceedings.mlr.press/v315/andrusca26a.html %V 315 %X Branched organoids exhibit increasingly complex morphologies as they progress from simple spheroid states to highly ramified structures, making topology-preserving segmentation essential for quantitative biological analysis. Capturing thin protrusions and maintaining branch continuity remains challenging for classical UNet-based architectures, particularly in brightfield imaging where fine structures are easily blurred or disconnected. In this work, we present a multi-scale Inception-UNet designed to capture the heterogeneous spatial scales of branched organoids through parallel convolutional paths with complementary receptive fields. As a model system, we analyze brightfield pancreatic ductal adenocarcinoma (PDAC) organoids, a system known for strong morphological heterogeneity and invasive branching behavior randriamanantsoa2022pdacOrganoids, cultured using high-throughput Patternoid assays kurzbach2025patternoid that enable standardized imaging and robust quantitative analysis. To assess segmentation quality beyond region overlap, we combine Dice with the structure-aware clDice metric that directly probes branch integrity and topological continuity. Across deterministic seeds and strictly separated organoid positions, the Inception-UNet achieves the highest region-based Dice ($0.868 \pm 0.062$) and clDice ($0.545 \pm 0.123$), and most importantly, the strongest preservation of branch continuity compared to UNet and UNet++. These improvements become increasingly pronounced with growing morphological complexity. Overall, our results demonstrate that multi-scale feature extraction combined with topology-aware evaluation substantially improves segmentation of branched organoids and provides a robust foundation for downstream morphological and invasion-related analyses.
APA
Andrusca, S.H., Kießling, C.D. & Bausch, A.R.. (2026). A Multi-Scale Inception-UNet with Structure-Aware Evaluation for Branch-Preserving Segmentation of Organoids. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:4071-4084 Available from https://proceedings.mlr.press/v315/andrusca26a.html.

Related Material