Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study

Linear accelerator–magnetic resonance (linac-MR) hybrid systems allow for real-time magnetic resonance imaging (MRI)-guided radiotherapy for more accurate dose delivery to the tumor and improved sparing of the adjacent healthy tissues. However, for real-time tumor detection, it is unfeasible for a h...

Full description

Saved in:
Bibliographic Details
Main Authors: Gawon Han, Keith Wachowicz, Nawaid Usmani, Don Yee, Jordan Wong, Arun Elangovan, Jihyun Yun, B. Gino Fallone
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Algorithms
Subjects:
Online Access:https://www.mdpi.com/1999-4893/18/4/233
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850156008807071744
author Gawon Han
Keith Wachowicz
Nawaid Usmani
Don Yee
Jordan Wong
Arun Elangovan
Jihyun Yun
B. Gino Fallone
author_facet Gawon Han
Keith Wachowicz
Nawaid Usmani
Don Yee
Jordan Wong
Arun Elangovan
Jihyun Yun
B. Gino Fallone
author_sort Gawon Han
collection DOAJ
description Linear accelerator–magnetic resonance (linac-MR) hybrid systems allow for real-time magnetic resonance imaging (MRI)-guided radiotherapy for more accurate dose delivery to the tumor and improved sparing of the adjacent healthy tissues. However, for real-time tumor detection, it is unfeasible for a human expert to manually contour (gold standard) the tumor at the fast imaging rate of a linac-MR. This study aims to develop a neural network-based tumor autocontouring algorithm with patient-specific hyperparameter optimization (HPO) and to validate its contouring accuracy using in vivo MR images of cancer patients. Two-dimensional (2D) intrafractional MR images were acquired at 4 frames/s using 3 tesla (T) MRI from 11 liver, 24 prostate, and 12 lung cancer patients. A U-Net architecture was applied for tumor autocontouring and was further enhanced by implementing HPO using the Covariance Matrix Adaptation Evolution Strategy. Six hyperparameters were optimized for each patient, for which intrafractional images and experts’ manual contours were input into the algorithm to find the optimal set of hyperparameters. For evaluation, Dice’s coefficient (DC), centroid displacement (CD), and Hausdorff distance (HD) were computed between the manual contours and autocontours. The performance of the algorithm was benchmarked against two standardized autosegmentation methods: non-optimized U-Net and nnU-Net. For the proposed algorithm, the mean (standard deviation) DC, CD, and HD of the 47 patients were 0.92 (0.04), 1.35 (1.03), and 3.63 (2.17) mm, respectively. Compared to the two benchmarking autosegmentation methods, the proposed algorithm achieved the best overall performance in terms of contouring accuracy and speed. This work presents the first tumor autocontouring algorithm applicable to the intrafractional MR images of liver and prostate cancer patients for real-time tumor-tracked radiotherapy. The proposed algorithm performs patient-specific HPO, enabling accurate tumor delineation comparable to that of experts.
format Article
id doaj-art-05ae453f46a844b59f91fa8ff48c1fef
institution OA Journals
issn 1999-4893
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Algorithms
spelling doaj-art-05ae453f46a844b59f91fa8ff48c1fef2025-08-20T02:24:43ZengMDPI AGAlgorithms1999-48932025-04-0118423310.3390/a18040233Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot StudyGawon Han0Keith Wachowicz1Nawaid Usmani2Don Yee3Jordan Wong4Arun Elangovan5Jihyun Yun6B. Gino Fallone7Medical Physics Division, Department of Oncology, University of Alberta, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaMedical Physics Division, Department of Oncology, University of Alberta, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaDepartment of Radiation Oncology, Cross Cancer Institute, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaDepartment of Radiation Oncology, Cross Cancer Institute, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaDepartment of Radiation Oncology, Cross Cancer Institute, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaDepartment of Radiation Oncology, Cross Cancer Institute, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaMedical Physics Division, Department of Oncology, University of Alberta, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaMedical Physics Division, Department of Oncology, University of Alberta, 11560 University Avenue, Edmonton, AB T6G 1Z2, CanadaLinear accelerator–magnetic resonance (linac-MR) hybrid systems allow for real-time magnetic resonance imaging (MRI)-guided radiotherapy for more accurate dose delivery to the tumor and improved sparing of the adjacent healthy tissues. However, for real-time tumor detection, it is unfeasible for a human expert to manually contour (gold standard) the tumor at the fast imaging rate of a linac-MR. This study aims to develop a neural network-based tumor autocontouring algorithm with patient-specific hyperparameter optimization (HPO) and to validate its contouring accuracy using in vivo MR images of cancer patients. Two-dimensional (2D) intrafractional MR images were acquired at 4 frames/s using 3 tesla (T) MRI from 11 liver, 24 prostate, and 12 lung cancer patients. A U-Net architecture was applied for tumor autocontouring and was further enhanced by implementing HPO using the Covariance Matrix Adaptation Evolution Strategy. Six hyperparameters were optimized for each patient, for which intrafractional images and experts’ manual contours were input into the algorithm to find the optimal set of hyperparameters. For evaluation, Dice’s coefficient (DC), centroid displacement (CD), and Hausdorff distance (HD) were computed between the manual contours and autocontours. The performance of the algorithm was benchmarked against two standardized autosegmentation methods: non-optimized U-Net and nnU-Net. For the proposed algorithm, the mean (standard deviation) DC, CD, and HD of the 47 patients were 0.92 (0.04), 1.35 (1.03), and 3.63 (2.17) mm, respectively. Compared to the two benchmarking autosegmentation methods, the proposed algorithm achieved the best overall performance in terms of contouring accuracy and speed. This work presents the first tumor autocontouring algorithm applicable to the intrafractional MR images of liver and prostate cancer patients for real-time tumor-tracked radiotherapy. The proposed algorithm performs patient-specific HPO, enabling accurate tumor delineation comparable to that of experts.https://www.mdpi.com/1999-4893/18/4/233autocontouringintrafractional motion managementtumor trackinglinac-MR hybridMRI guidanceradiotherapy
spellingShingle Gawon Han
Keith Wachowicz
Nawaid Usmani
Don Yee
Jordan Wong
Arun Elangovan
Jihyun Yun
B. Gino Fallone
Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
Algorithms
autocontouring
intrafractional motion management
tumor tracking
linac-MR hybrid
MRI guidance
radiotherapy
title Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
title_full Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
title_fullStr Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
title_full_unstemmed Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
title_short Patient-Specific Hyperparameter Optimization of a Deep Learning-Based Tumor Autocontouring Algorithm on 2D Liver, Prostate, and Lung Cine MR Images: A Pilot Study
title_sort patient specific hyperparameter optimization of a deep learning based tumor autocontouring algorithm on 2d liver prostate and lung cine mr images a pilot study
topic autocontouring
intrafractional motion management
tumor tracking
linac-MR hybrid
MRI guidance
radiotherapy
url https://www.mdpi.com/1999-4893/18/4/233
work_keys_str_mv AT gawonhan patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT keithwachowicz patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT nawaidusmani patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT donyee patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT jordanwong patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT arunelangovan patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT jihyunyun patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy
AT bginofallone patientspecifichyperparameteroptimizationofadeeplearningbasedtumorautocontouringalgorithmon2dliverprostateandlungcinemrimagesapilotstudy