Abstract
When applying a Deep Learning model to medical images, it is crucial to estimate the model uncertainty. Voxel-wise uncertainty is a useful visual marker for human experts and could be used to improve the model’s voxel-wise output, such as segmentation. Moreover, uncertainty provides a solid foundation for out-of-distribution (OOD) detection, improving the model performance on the image-wise level. However, one of the frequent tasks in medical imaging is the segmentation of distinct, local structures such as tumors or lesions. Here, the structure-wise uncertainty allows more precise operations than image-wise and more semantic-aware than voxel-wise. The way to produce uncertainty for individual structures remains poorly explored. We propose a framework to measure the structure-wise uncertainty and evaluate the impact of OOD data on the model performance. Thus, we identify the best UE method to improve the segmentation quality. The proposed framework is tested on three datasets with the tumor segmentation task: LIDC-IDRI, LiTS, and a private one with multiple brain metastases cases.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Lee, J.G., Jun, S., Cho, Y.W., Lee, H., Kim, G.B., Seo, J.B., Kim, N.: Deep learning in medical imaging: general overview. Korean journal of radiology 18(4), 570–584 (2017)
Kompa, B., Snoek, J., Beam, A.L.: Second opinion needed: communicating uncertainty in medical machine learning. NPJ Digital Medicine 4(1), 1–6 (2021)
Iwamoto, S., Raytchev, B., Tamaki, T., Kaneda, K.: Improving the reliability of semantic segmentation of medical images by uncertainty modeling with Bayesian deep networks and curriculum learning. In: Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, and Perinatal Imaging, Placental and Preterm Image Analysis, pp. 34–43. Springer (2021)
Linmans, J., van der Laak, J., Litjens, G.: Efficient out-of-distribution detection in digital pathology using multi-head convolutional neural networks. In: MIDL. pp. 465–478 (2020)
Sahiner, B., Pezeshk, A., Hadjiiski, L.M., Wang, X., Drukker, K., Cha, K.H., Summers, R.M., Giger, M.L.: Deep learning in medical imaging and radiation therapy. Medical physics 46(1), e1–e36 (2019)
Leibig, C., Allken, V., Ayhan, M.S., Berens, P., Wahl, S.: Leveraging uncertainty information from deep neural networks for disease detection. Scientific reports 7(1), 1–14 (2017)
Roy, A.G., Conjeti, S., Navab, N., Wachinger, C., Initiative, A.D.N., et al.: Bayesian quicknat: Model uncertainty in deep whole-brain segmentation for structure-wise quality control. NeuroImage 195, 11–22 (2019)
Nair, T., Precup, D., Arnold, D.L., Arbel, T.: Exploring uncertainty measures in deep networks for multiple sclerosis lesion detection and segmentation. Medical Image Analysis 59, 101557 (2020), https://www.sciencedirect.com/science/article/pii/S1361841519300994
Ozdemir, O., Woodward, B., Berlin, A.A.: Propagating uncertainty in multi-stage Bayesian convolutional neural networks with application to pulmonary nodule detection. CoRR abs/1712.00497 (2017), http://arxiv.org/abs/1712.00497
Bhat, I., Kuijf, H.J., Cheplygina, V., Pluim, J.P.: Using uncertainty estimation to reduce false positives in liver lesion detection. In: 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI). pp. 663–667 (2021)
Mehrtash, A., Wells, W., Tempany, C., Abolmaesumi, P., Kapur, T.: Confidence calibration and predictive uncertainty estimation for deep medical image segmentation. IEEE Transactions on Medical Imaging PP, 1–1 (07 2020)
Hoebel, K., Andrearczyk, V., Beers, A., Patel, J., Chang, K., Depeursinge, A., Müller, H., Kalpathy-Cramer, J.: An exploration of uncertainty information for segmentation quality assessment. In: Išgum, I., Landman, B.A. (eds.) Medical Imaging 2020: Image Processing. vol. 11313, pp. 381–390. International Society for Optics and Photonics, SPIE (2020), https://doi.org/10.1117/12.2548722
Devries, T., Taylor, G.W.: Leveraging uncertainty estimates for predicting segmentation quality. ArXiv abs/1807.00502 (2018)
Seeböck, P., Orlando, J., Schlegl, T., Waldstein, S., Bogunović, H., Riedl, S., Langs, G., Schmidt-Erfurth, U.: Exploiting epistemic uncertainty of anatomy segmentation for anomaly detection in retinal oct. IEEE Transactions on Medical Imaging PP, 1–1 (05 2019)
Hiasa, Y., Otake, Y., Takao, M., Ogawa, T., Sugano, N., Sato, Y.: Automated muscle segmentation from clinical ct using Bayesian u-net for personalized musculoskeletal modeling. IEEE Transactions on Medical Imaging 39(4), 1030–1040 (2020)
Lakshminarayanan, B., Pritzel, A., Blundell, C.: Simple and scalable predictive uncertainty estimation using deep ensembles. Advances in neural information processing systems 30 (2017)
Jungo, A., Reyes, M.: Assessing reliability and challenges of uncertainty estimations for medical image segmentation. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 48–56. Springer (2019)
Houlsby, N., Huszár, F., Ghahramani, Z., Lengyel, M.: Bayesian active learning for classification and preference learning. arXiv preprint arXiv:1112.5745 (2011)
Smith, L., Gal, Y.: Understanding measures of uncertainty for adversarial example detection. arXiv preprint arXiv:1803.08533 (2018)
Lu, S.L., Liao, H.C., Hsu, F.M., Liao, C.C., Lai, F., Xiao, F.: The intracranial tumor segmentation challenge: Contour tumors on brain mri for radiosurgery. NeuroImage 244, 118585 (2021)
van der Voort, S.R., Incekara, F., Wijnenga, M.M., Kapsas, G., Gahrmann, R., Schouten, J.W., Dubbink, H.J., Vincent, A.J., van den Bent, M.J., French, P.J., et al.: The erasmus glioma database (egd): Structural mri scans, who 2016 subtypes, and segmentations of 774 patients with glioma. Data in brief 37, 107191 (2021)
Armato III, S.G., McLennan, G., Bidaut, L., McNitt-Gray, M.F., Meyer, C.R., Reeves, A.P., Zhao, B., Aberle, D.R., Henschke, C.I., Hoffman, E.A., et al.: The lung image database consortium (lidc) and image database resource initiative (idri): a completed reference database of lung nodules on ct scans. Medical physics 38(2), 915–931 (2011)
Tsai, E.B., Simpson, S., Lungren, M.P., Hershman, M., Roshkovan, L., Colak, E., Erickson, B.J., Shih, G., Stein, A., Kalpathy-Cramer, J., et al.: The rsna international covid-19 open radiology database (ricord). Radiology 299(1), E204–E213 (2021)
Bilic, P., Christ, P.F., Vorontsov, E., Chlebus, G., Chen, H., Dou, Q., Fu, C.W., Han, X., Heng, P.A., Hesser, J., et al.: The liver tumor segmentation benchmark (lits). arXiv preprint arXiv:1901.04056 (2019)
Pimkin, A., Samoylenko, A., Antipina, N., Ovechkina, A., Golanov, A., Dalechina, A., Belyaev, M.: Multidomain ct metal artifacts reduction using partial convolution based inpainting. In: 2020 International Joint Conference on Neural Networks (IJCNN). pp. 1–6. IEEE (2020)
Saparov, T., Kurmukov, A., Shirokikh, B., Belyaev, M.: Zero-shot domain adaptation in ct segmentation by filtered back projection augmentation. In: Deep Generative Models, and Data Augmentation, Labelling, and Imperfections, pp. 243–250. Springer (2021)
Isensee, F., Jaeger, P.F., Kohl, S.A., Petersen, J., Maier-Hein, K.H.: nnu-net: a self-configuring method for deep learning-based biomedical image segmentation. Nature methods 18(2), 203–211 (2021)
Milletari, F., Navab, N., Ahmadi, S.A.: V-net: Fully convolutional neural networks for volumetric medical image segmentation. In: 2016 fourth international conference on 3D vision (3DV). pp. 565–571. IEEE (2016)
Acknowledgements
The authors acknowledge the National Cancer Institute and the Foundation for the National Institutes of Health, and their critical role in the creation of the free publicly available LIDC/IDRI Database used in this study. This research was funded by Russian Science Foundation grant number 20-71-10134.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Experimental Setup
Experimental Setup
1.1 Preprocessing
Here, we describe data preparation steps including datasets splits, normalization, and interpolation.
Mets data is randomly split into train (1140 images) and test (414 images) sets. We interpolate the images to have \(1\,\text {mm} \times 1\,\text {mm} \times 1\) mm spacing.
LIDC data is randomly split into train (812 images) and test (204 images) sets. We clip image intensities between −1350 and 350 Hounsfield units (HU)—the standard lung window. We interpolate images to have 1 mm \(\times \) 1 mm \(\times \) 1.5 mm spacing.
LiTS is presented as two subsets, so we use the first as a test (28 images) and the second, excluding cases with empty tumor masks, as a train (90 images) set. The images are cropped to the provided liver masks. The intensities are clipped to the \([-150, 250]\) HU interval—the standard liver window. Finally, we interpolate images to have 0.77 mm \(\times \) 0.77 mm \(\times \) 1 mm spacing.
LiTS-mod is obtained by random changes of the reconstruction kernel to be extremely soft (\(a=-0.7, b=0.5\)) or sharp (\(a=30, b=3\)) using the implementation and notations of [26], and addition of “metal” artifacts (ball of radius 5 and 3000 HU) by substituting the parts of sinogram projection, as in [25].
Before passing through the network, we scale image intensities in [0, 1].
1.2 Training Setup
Although using cross-entropy loss has theoretical justifications of encouraging better calibrated predictions [16], models trained with this loss function fail in our segmentation task. For that reason we use Dice Loss [28] and its modifications in our experiments. Thus, uncertainty estimates might be shifted in such tasks, and experimental evaluation, as in our study, becomes even more relevant. All models are trained in a patch-based manner: patches are sampled randomly so that they contain structures. We use SGD optimizer with Nesterov momentum of 0.9 and \(10^{-3}\) initial learning rate, which is decreased to \(10^{-4}\) after \(80\%\) of epochs. For LiTS and Mets segmentation the model is trained for 100 epochs (100 iterations per epoch, batch size 20), while for LIDC segmentation there are 30 epochs (1000 iterations per epoch, batch size 2).
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Vasiliuk, A., Frolova, D., Belyaev, M., Shirokikh, B. (2023). Exploring Structure-Wise Uncertainty for 3D Medical Image Segmentation. In: Su, R., Zhang, Y., Liu, H., F Frangi, A. (eds) Medical Imaging and Computer-Aided Diagnosis. MICAD 2022. Lecture Notes in Electrical Engineering, vol 810. Springer, Singapore. https://doi.org/10.1007/978-981-16-6775-6_2
Download citation
DOI: https://doi.org/10.1007/978-981-16-6775-6_2
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-6774-9
Online ISBN: 978-981-16-6775-6
eBook Packages: MedicineMedicine (R0)