From Barlow Twins to Triplet Training: Differentiating Dementia with Limited Data

Yitong Li, Tom Nuno Wolf, Sebastian Pölsterl, Igor Yakushev, Dennis M. Hedderich, Christian Wachinger
Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, PMLR 250:888-902, 2024.

Abstract

Differential diagnosis of dementia is challenging due to overlapping symptoms, with structural magnetic resonance imaging (MRI) being the primary method for diagnosis. Despite the clinical value of computer-aided differential diagnosis, research has been limited, mainly due to the absence of public datasets that contain diverse types of dementia. This leaves researchers with small in-house datasets that are insufficient for training deep neural networks (DNNs). Self-supervised learning shows promise for utilizing unlabeled MRI scans in training, but small batch sizes for volumetric brain scans make its application challenging. To address these issues, we propose Triplet Training for differential diagnosis with limited target data. It consists of three key stages: (i) self-supervised pre-training on unlabeled data with Barlow Twins, (ii) self-distillation on task-related data, and (iii) fine-tuning on the target dataset. Our approach significantly outperforms traditional training strategies, achieving a balanced accuracy of 75.6%. We further provide insights into the training process by visualizing changes in the latent space after each step. Finally, we validate the robustness of Triplet Training in terms of its individual components in a comprehensive ablation study. Our code is available at https://github.com/ai-med/TripletTraining.

Cite this Paper

BibTeX
@InProceedings{pmlr-v250-li24a, title = {From Barlow Twins to Triplet Training: Differentiating Dementia with Limited Data}, author = {Li, Yitong and Wolf, Tom Nuno and P\"olsterl, Sebastian and Yakushev, Igor and Hedderich, Dennis M. and Wachinger, Christian}, booktitle = {Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning}, pages = {888--902}, year = {2024}, editor = {Burgos, Ninon and Petitjean, Caroline and Vakalopoulou, Maria and Christodoulidis, Stergios and Coupe, Pierrick and Delingette, Hervé and Lartizien, Carole and Mateus, Diana}, volume = {250}, series = {Proceedings of Machine Learning Research}, month = {03--05 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v250/main/assets/li24a/li24a.pdf}, url = {https://proceedings.mlr.press/v250/li24a.html}, abstract = {Differential diagnosis of dementia is challenging due to overlapping symptoms, with structural magnetic resonance imaging (MRI) being the primary method for diagnosis. Despite the clinical value of computer-aided differential diagnosis, research has been limited, mainly due to the absence of public datasets that contain diverse types of dementia. This leaves researchers with small in-house datasets that are insufficient for training deep neural networks (DNNs). Self-supervised learning shows promise for utilizing unlabeled MRI scans in training, but small batch sizes for volumetric brain scans make its application challenging. To address these issues, we propose Triplet Training for differential diagnosis with limited target data. It consists of three key stages: (i) self-supervised pre-training on unlabeled data with Barlow Twins, (ii) self-distillation on task-related data, and (iii) fine-tuning on the target dataset. Our approach significantly outperforms traditional training strategies, achieving a balanced accuracy of 75.6%. We further provide insights into the training process by visualizing changes in the latent space after each step. Finally, we validate the robustness of Triplet Training in terms of its individual components in a comprehensive ablation study. Our code is available at https://github.com/ai-med/TripletTraining.} }
Endnote
%0 Conference Paper %T From Barlow Twins to Triplet Training: Differentiating Dementia with Limited Data %A Yitong Li %A Tom Nuno Wolf %A Sebastian Pölsterl %A Igor Yakushev %A Dennis M. Hedderich %A Christian Wachinger %B Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2024 %E Ninon Burgos %E Caroline Petitjean %E Maria Vakalopoulou %E Stergios Christodoulidis %E Pierrick Coupe %E Hervé Delingette %E Carole Lartizien %E Diana Mateus %F pmlr-v250-li24a %I PMLR %P 888--902 %U https://proceedings.mlr.press/v250/li24a.html %V 250 %X Differential diagnosis of dementia is challenging due to overlapping symptoms, with structural magnetic resonance imaging (MRI) being the primary method for diagnosis. Despite the clinical value of computer-aided differential diagnosis, research has been limited, mainly due to the absence of public datasets that contain diverse types of dementia. This leaves researchers with small in-house datasets that are insufficient for training deep neural networks (DNNs). Self-supervised learning shows promise for utilizing unlabeled MRI scans in training, but small batch sizes for volumetric brain scans make its application challenging. To address these issues, we propose Triplet Training for differential diagnosis with limited target data. It consists of three key stages: (i) self-supervised pre-training on unlabeled data with Barlow Twins, (ii) self-distillation on task-related data, and (iii) fine-tuning on the target dataset. Our approach significantly outperforms traditional training strategies, achieving a balanced accuracy of 75.6%. We further provide insights into the training process by visualizing changes in the latent space after each step. Finally, we validate the robustness of Triplet Training in terms of its individual components in a comprehensive ablation study. Our code is available at https://github.com/ai-med/TripletTraining.
APA
Li, Y., Wolf, T.N., Pölsterl, S., Yakushev, I., Hedderich, D.M. & Wachinger, C.. (2024). From Barlow Twins to Triplet Training: Differentiating Dementia with Limited Data. Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 250:888-902 Available from https://proceedings.mlr.press/v250/li24a.html.

Related Material