Few Shot Hematopoietic Cell Classification

Vu Nguyen, Prantik Howlader, Le Hou, Dimitris Samaras, Rajarsi R. Gupta, Joel Saltz
Medical Imaging with Deep Learning, PMLR 227:1085-1103, 2024.

Abstract

We propose a few shot learning approach for the problem of hematopoietic cell classification in digital pathology. In hematopoiesis cell classification, the classes correspond to the different stages of the cellular maturation process. Two consecutive stage categories are considered to have a neighborhood relationship, which implies a visual similarity between the two categories. We propose RelationVAE which incorporates these relationships between hematopoietic cell classes to robustly generate more data for the classes with limited training data. Specifically, we first model these relationships using a graphical model, and propose RelationVAE, a deep generative model which implements the graphical model. RelationVAE is trained to optimize the lower bound of the pairwise data likelihood of the graphical model. In this way, it can identify class level features of a specific class from a small number of input images together with the knowledge transferred from visually similar classes, leading to more robust sample synthesis. The experiments on our collected hematopoietic dataset show the improved results of our proposed RelationVAE over a baseline VAE model and other few shot learning methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v227-nguyen24a, title = {Few Shot Hematopoietic Cell Classification}, author = {Nguyen, Vu and Howlader, Prantik and Hou, Le and Samaras, Dimitris and Gupta, Rajarsi R. and Saltz, Joel}, booktitle = {Medical Imaging with Deep Learning}, pages = {1085--1103}, year = {2024}, editor = {Oguz, Ipek and Noble, Jack and Li, Xiaoxiao and Styner, Martin and Baumgartner, Christian and Rusu, Mirabela and Heinmann, Tobias and Kontos, Despina and Landman, Bennett and Dawant, Benoit}, volume = {227}, series = {Proceedings of Machine Learning Research}, month = {10--12 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v227/nguyen24a/nguyen24a.pdf}, url = {https://proceedings.mlr.press/v227/nguyen24a.html}, abstract = {We propose a few shot learning approach for the problem of hematopoietic cell classification in digital pathology. In hematopoiesis cell classification, the classes correspond to the different stages of the cellular maturation process. Two consecutive stage categories are considered to have a neighborhood relationship, which implies a visual similarity between the two categories. We propose RelationVAE which incorporates these relationships between hematopoietic cell classes to robustly generate more data for the classes with limited training data. Specifically, we first model these relationships using a graphical model, and propose RelationVAE, a deep generative model which implements the graphical model. RelationVAE is trained to optimize the lower bound of the pairwise data likelihood of the graphical model. In this way, it can identify class level features of a specific class from a small number of input images together with the knowledge transferred from visually similar classes, leading to more robust sample synthesis. The experiments on our collected hematopoietic dataset show the improved results of our proposed RelationVAE over a baseline VAE model and other few shot learning methods.} }
Endnote
%0 Conference Paper %T Few Shot Hematopoietic Cell Classification %A Vu Nguyen %A Prantik Howlader %A Le Hou %A Dimitris Samaras %A Rajarsi R. Gupta %A Joel Saltz %B Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2024 %E Ipek Oguz %E Jack Noble %E Xiaoxiao Li %E Martin Styner %E Christian Baumgartner %E Mirabela Rusu %E Tobias Heinmann %E Despina Kontos %E Bennett Landman %E Benoit Dawant %F pmlr-v227-nguyen24a %I PMLR %P 1085--1103 %U https://proceedings.mlr.press/v227/nguyen24a.html %V 227 %X We propose a few shot learning approach for the problem of hematopoietic cell classification in digital pathology. In hematopoiesis cell classification, the classes correspond to the different stages of the cellular maturation process. Two consecutive stage categories are considered to have a neighborhood relationship, which implies a visual similarity between the two categories. We propose RelationVAE which incorporates these relationships between hematopoietic cell classes to robustly generate more data for the classes with limited training data. Specifically, we first model these relationships using a graphical model, and propose RelationVAE, a deep generative model which implements the graphical model. RelationVAE is trained to optimize the lower bound of the pairwise data likelihood of the graphical model. In this way, it can identify class level features of a specific class from a small number of input images together with the knowledge transferred from visually similar classes, leading to more robust sample synthesis. The experiments on our collected hematopoietic dataset show the improved results of our proposed RelationVAE over a baseline VAE model and other few shot learning methods.
APA
Nguyen, V., Howlader, P., Hou, L., Samaras, D., Gupta, R.R. & Saltz, J.. (2024). Few Shot Hematopoietic Cell Classification. Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 227:1085-1103 Available from https://proceedings.mlr.press/v227/nguyen24a.html.

Related Material