Application of Simultaneous Uncertainty Quantification and Segmentation for Oropharyngeal Cancer Use-case with Bayesian Deep Learning

Overview

Journal Commun Med (Lond)

Publisher Nature Portfolio

Specialty General Medicine

Date 2024 Jun 8

PMID 38851837

Authors

Jaakko Sahlsten

Joel Jaskari

Kareem A Wahid

Sara Ahmed

Enrico Glerean

Renjie He

Benjamin H Kann

Antti Makitie

Clifton D Fuller

Mohamed A Naser

Kimmo Kaski

Affiliations

Soon will be listed here.

Abstract

Background: Radiotherapy is a core treatment modality for oropharyngeal cancer (OPC), where the primary gross tumor volume (GTVp) is manually segmented with high interobserver variability. This calls for reliable and trustworthy automated tools in clinician workflow. Therefore, accurate uncertainty quantification and its downstream utilization is critical.

Methods: Here we propose uncertainty-aware deep learning for OPC GTVp segmentation, and illustrate the utility of uncertainty in multiple applications. We examine two Bayesian deep learning (BDL) models and eight uncertainty measures, and utilize a large multi-institute dataset of 292 PET/CT scans to systematically analyze our approach.

Results: We show that our uncertainty-based approach accurately predicts the quality of the deep learning segmentation in 86.6% of cases, identifies low performance cases for semi-automated correction, and visualizes regions of the scans where the segmentations likely fail.

Conclusions: Our BDL-based analysis provides a first-step towards more widespread implementation of uncertainty quantification in OPC GTVp segmentation.

References

Nikolov S, Blackwell S, Zverovitch A, Mendes R, Livne M, De Fauw J . Clinically Applicable Segmentation of Head and Neck Anatomy for Radiotherapy: Deep Learning Algorithm Development and Validation Study. J Med Internet Res. 2021; 23(7):e26151. PMC: 8314151. DOI: 10.2196/26151. View

Andrearczyk V, Oreiller V, Abobakr M, Akhavanallaf A, Balermpas P, Boughdad S . Overview of the HECKTOR Challenge at MICCAI 2022: Automatic Head and Neck Tumor Segmentation and Outcome Prediction in PET/CT. Head Neck Tumor Chall (2022). 2023; 13626:1-30. PMC: 10171217. DOI: 10.1007/978-3-031-27420-6_1. View

Rasch C, Steenbakkers R, Van Herk M . Target definition in prostate, head, and neck. Semin Radiat Oncol. 2005; 15(3):136-45. DOI: 10.1016/j.semradonc.2005.01.005. View

Wahid K, Ahmed S, He R, van Dijk L, Teuwen J, McDonald B . Evaluation of deep learning-based multiparametric MRI oropharyngeal primary tumor auto-segmentation and investigation of input channel effects: Results from a prospective imaging registry. Clin Transl Radiat Oncol. 2021; 32:6-14. PMC: 8570930. DOI: 10.1016/j.ctro.2021.10.003. View

McDonald B, Cardenas C, OConnell N, Ahmed S, Naser M, Wahid K . Investigation of autosegmentation techniques on T2-weighted MRI for off-line dose reconstruction in MR-linac workflow for head and neck cancers. Med Phys. 2023; 51(1):278-291. PMC: 10799175. DOI: 10.1002/mp.16582. View

Naser M, Wahid K, van Dijk L, He R, Abdelaal M, Dede C . Head and Neck Cancer Primary Tumor Auto Segmentation Using Model Ensembling of Deep Learning in PET/CT Images. Head Neck Tumor Segm Chall (2021). 2022; 13209:121-132. PMC: 8991449. DOI: 10.1007/978-3-030-98253-9_11. View

van den Berg C, Meliado E . Uncertainty Assessment for Deep Learning Radiotherapy Applications. Semin Radiat Oncol. 2022; 32(4):304-318. DOI: 10.1016/j.semradonc.2022.06.001. View

Dolezal J, Srisuwananukorn A, Karpeyev D, Ramesh S, Kochanny S, Cody B . Uncertainty-informed deep learning models enable high-confidence predictions for digital histopathology. Nat Commun. 2022; 13(1):6572. PMC: 9630455. DOI: 10.1038/s41467-022-34025-x. View

Mehrtash A, Wells W, Tempany C, Abolmaesumi P, Kapur T . Confidence Calibration and Predictive Uncertainty Estimation for Deep Medical Image Segmentation. IEEE Trans Med Imaging. 2020; 39(12):3868-3878. PMC: 7704933. DOI: 10.1109/TMI.2020.3006437. View

10.

Dohopolski M, Chen L, Sher D, Wang J . Predicting lymph node metastasis in patients with oropharyngeal cancer by using a convolutional neural network with associated epistemic and aleatoric uncertainty. Phys Med Biol. 2020; 65(22):225002. DOI: 10.1088/1361-6560/abb71c. View

11.

De Biase A, Sijtsema N, van Dijk L, Langendijk J, van Ooijen P . Deep learning aided oropharyngeal cancer segmentation with adaptive thresholding for predicted tumor probability in FDG PET and CT images. Phys Med Biol. 2023; 68(5). DOI: 10.1088/1361-6560/acb9cf. View

12.

Sherer M, Lin D, Elguindi S, Duke S, Tan L, Cacicedo J . Metrics to evaluate the performance of auto-segmentation for radiation treatment planning: A critical review. Radiother Oncol. 2021; 160:185-191. PMC: 9444281. DOI: 10.1016/j.radonc.2021.05.003. View

13.

Njeh C . Tumor delineation: The weakest link in the search for accuracy in radiotherapy. J Med Phys. 2009; 33(4):136-40. PMC: 2772050. DOI: 10.4103/0971-6203.44472. View

14.

Andrearczyk V, Oreiller V, Boughdad S, Rest C, Tankyevych O, Elhalawani H . Automatic Head and Neck Tumor segmentation and outcome prediction relying on FDG-PET/CT images: Findings from the second edition of the HECKTOR challenge. Med Image Anal. 2023; 90:102972. DOI: 10.1016/j.media.2023.102972. View

15.

Wahid K, Glerean E, Sahlsten J, Jaskari J, Kaski K, Naser M . Artificial Intelligence for Radiation Oncology Applications Using Public Datasets. Semin Radiat Oncol. 2022; 32(4):400-414. PMC: 9587532. DOI: 10.1016/j.semradonc.2022.06.009. View

16.

Nguyen D, Barkousaraie A, Bohara G, Balagopal A, McBeth R, Lin M . A comparison of Monte Carlo dropout and bootstrap aggregation on the performance and uncertainty estimation in radiation therapy dose prediction with deep learning neural networks. Phys Med Biol. 2021; 66(5):054002. PMC: 8837265. DOI: 10.1088/1361-6560/abe04f. View

17.

Segedin B, Petric P . Uncertainties in target volume delineation in radiotherapy - are they relevant and what can we do about them?. Radiol Oncol. 2016; 50(3):254-62. PMC: 5024655. DOI: 10.1515/raon-2016-0023. View

18.

Lin D, Wahid K, Nelms B, He R, Naser M, Duke S . prospective acceptability benchmarking from the Contouring Collaborative for Consensus in Radiation Oncology crowdsourced initiative for multiobserver segmentation. J Med Imaging (Bellingham). 2023; 10(Suppl 1):S11903. PMC: 9907021. DOI: 10.1117/1.JMI.10.S1.S11903. View

19.

Shelhamer E, Long J, Darrell T . Fully Convolutional Networks for Semantic Segmentation. IEEE Trans Pattern Anal Mach Intell. 2016; 39(4):640-651. DOI: 10.1109/TPAMI.2016.2572683. View

20.

Virtanen P, Gommers R, Oliphant T, Haberland M, Reddy T, Cournapeau D . SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods. 2020; 17(3):261-272. PMC: 7056644. DOI: 10.1038/s41592-019-0686-2. View