Explainable Weakly-Supervised Cell Segmentation by Canonical Shape Learning and Transformation

Pedro Costa, Alex Gaudio, Aurélio Campilho, Jaime S. Cardoso
Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, PMLR 172:250-260, 2022.

Abstract

Microscopy images have been increasingly analyzed quantitatively in biomedical research. Segmenting individual cell nucleus is an important step as many research studies involve counting cell nuclei and analysing their shape. We propose a novel weakly supervised instance segmentation method trained with image segmentation masks only. Our system comprises two models: an implicit shape Multi-Layer Perceptron (MLP) that learns the shape of the nuclei in canonical coordinates; and 2) an encoder that predicts the parameters of the affine transformation to deform the canonical shape into the correct location, scale, and orientation in the image. To further improve the performance of the model, we propose a loss that uses the total number of nuclei in an image as supervision. Our system is explainable, as the implicit shape MLP learns that the canonical shape of the cell nuclei is a circle, and interpretable as the output of the encoder are parameters of affine transformations. We obtain image segmentation performance close to DeepLabV3 and, additionally, obtain an F1-score$_{IoU=0.5}$ of $68.47%$ at the instance segmentation task, even though the system was trained with image segmentations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v172-costa22a, title = {Explainable Weakly-Supervised Cell Segmentation by Canonical Shape Learning and Transformation}, author = {Costa, Pedro and Gaudio, Alex and Campilho, Aur\'elio and Cardoso, Jaime S.}, booktitle = {Proceedings of The 5th International Conference on Medical Imaging with Deep Learning}, pages = {250--260}, year = {2022}, editor = {Konukoglu, Ender and Menze, Bjoern and Venkataraman, Archana and Baumgartner, Christian and Dou, Qi and Albarqouni, Shadi}, volume = {172}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v172/costa22a/costa22a.pdf}, url = {https://proceedings.mlr.press/v172/costa22a.html}, abstract = {Microscopy images have been increasingly analyzed quantitatively in biomedical research. Segmenting individual cell nucleus is an important step as many research studies involve counting cell nuclei and analysing their shape. We propose a novel weakly supervised instance segmentation method trained with image segmentation masks only. Our system comprises two models: an implicit shape Multi-Layer Perceptron (MLP) that learns the shape of the nuclei in canonical coordinates; and 2) an encoder that predicts the parameters of the affine transformation to deform the canonical shape into the correct location, scale, and orientation in the image. To further improve the performance of the model, we propose a loss that uses the total number of nuclei in an image as supervision. Our system is explainable, as the implicit shape MLP learns that the canonical shape of the cell nuclei is a circle, and interpretable as the output of the encoder are parameters of affine transformations. We obtain image segmentation performance close to DeepLabV3 and, additionally, obtain an F1-score$_{IoU=0.5}$ of $68.47%$ at the instance segmentation task, even though the system was trained with image segmentations.} }
Endnote
%0 Conference Paper %T Explainable Weakly-Supervised Cell Segmentation by Canonical Shape Learning and Transformation %A Pedro Costa %A Alex Gaudio %A Aurélio Campilho %A Jaime S. Cardoso %B Proceedings of The 5th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2022 %E Ender Konukoglu %E Bjoern Menze %E Archana Venkataraman %E Christian Baumgartner %E Qi Dou %E Shadi Albarqouni %F pmlr-v172-costa22a %I PMLR %P 250--260 %U https://proceedings.mlr.press/v172/costa22a.html %V 172 %X Microscopy images have been increasingly analyzed quantitatively in biomedical research. Segmenting individual cell nucleus is an important step as many research studies involve counting cell nuclei and analysing their shape. We propose a novel weakly supervised instance segmentation method trained with image segmentation masks only. Our system comprises two models: an implicit shape Multi-Layer Perceptron (MLP) that learns the shape of the nuclei in canonical coordinates; and 2) an encoder that predicts the parameters of the affine transformation to deform the canonical shape into the correct location, scale, and orientation in the image. To further improve the performance of the model, we propose a loss that uses the total number of nuclei in an image as supervision. Our system is explainable, as the implicit shape MLP learns that the canonical shape of the cell nuclei is a circle, and interpretable as the output of the encoder are parameters of affine transformations. We obtain image segmentation performance close to DeepLabV3 and, additionally, obtain an F1-score$_{IoU=0.5}$ of $68.47%$ at the instance segmentation task, even though the system was trained with image segmentations.
APA
Costa, P., Gaudio, A., Campilho, A. & Cardoso, J.S.. (2022). Explainable Weakly-Supervised Cell Segmentation by Canonical Shape Learning and Transformation. Proceedings of The 5th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 172:250-260 Available from https://proceedings.mlr.press/v172/costa22a.html.

Related Material