Automated segmentation of brain metastases in T1-weighted contrast-enhanced MR images pre and post Stereotactic Radiosurgery

Hemalatha Kanakarajan; Wouter De Baene; Patrick Hanssens; Margriet Sitskoorn

doi:10.1101/2023.07.07.23292387

Abstract

Background and purpose Accurate segmentation of brain metastases on Magnetic Resonance Imaging (MRI) is a tedious and time-consuming task for radiologists that could be optimized with deep learning (DL) methods. Previous studies that evaluated the performance of several DL algorithms focused on training and testing the models on the planning MRI only. The purpose of this study is to evaluate well-known DL approaches (nnU-Net and MedNeXt) for their performance on both planning and follow-up MRI.

Materials and methods Pre-treatment brain MRIs were collected retrospectively for 263 patients at the Gamma Knife Center of Elisabeth-TweeSteden Hospital (ETZ). The patients were split into 203 patients for training and 60 patients for testing. For these 60 patients, the follow-up MRIs were also retrospectively collected. To increase heterogeneity, we added the publicly available MRI from the Mathematical oncology laboratory for 75 patients to the training data. The performance was compared between the two models, with and without addition of the public data.

Results All models obtained a good Dice Similarity Coefficient (all DSC >= 0.928) for planning MRI. MedNeXt trained with combined data was the best performing model for follow-ups at 3, 15 and 21 months (DSC of 0.796, 0.725, and 0.720 respectively) while nnU-Net trained with combined data was the best performing model for follow-ups at 6, 9 and 12 months (DSC of 0.738, 0.740 and 0.713 respectively).

Conclusion The models achieved a good performance score for planning MRI. Though the models performed worse for follow-ups, addition of public data enhanced their performance, providing a viable solution to improve their efficacy for the follow-ups. There was only limited difference in performance of the two algorithms. These algorithms hold promise as a valuable tool for clinicians for automated segmentation of pre- and postsurgical MRI during treatment planning and response evaluations, respectively.

Competing Interest Statement

The authors have declared no competing interest.

Funding Statement

This research is supported by KWF Kankerbestrijding and NWO Domain AES, as part of their joint strategic research programme: Technology for Oncology IL. The collaboration project is co-funded by the PPP Allowance made available by Health Holland, Top Sector Life Sciences & Health, to stimulate public-private partnerships.

Author Declarations

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

The Ethics Review Board (ERB) of Tilburg University gave ethical approval for this work.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes