Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI
Abstract Manual segmentation of lesions, required for radiotherapy planning and follow-up, is time-consuming and error-prone. Automatic detection and segmentation can assist radiologists in these tasks. This work explores the automated detection and segmentation of brain metastases (BMs) in longitud...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-12-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-024-78865-7 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559461793103872 |
|---|---|
| author | Vincent Andrearczyk Luis Schiappacasse Daniel Abler Marek Wodzinski Andreas Hottinger Matthieu Raccaud Jean Bourhis John O. Prior Vincent Dunet Adrien Depeurnge |
| author_facet | Vincent Andrearczyk Luis Schiappacasse Daniel Abler Marek Wodzinski Andreas Hottinger Matthieu Raccaud Jean Bourhis John O. Prior Vincent Dunet Adrien Depeurnge |
| author_sort | Vincent Andrearczyk |
| collection | DOAJ |
| description | Abstract Manual segmentation of lesions, required for radiotherapy planning and follow-up, is time-consuming and error-prone. Automatic detection and segmentation can assist radiologists in these tasks. This work explores the automated detection and segmentation of brain metastases (BMs) in longitudinal MRIs. It focuses on several important aspects: identifying and segmenting new lesions for screening and treatment planning, re-segmenting lesions in successive images using prior lesion locations as an additional input channel, and performing multi-component segmentation to distinguish between enhancing tissue, edema, and necrosis. The key component of the proposed approach is to propagate the lesion mask from the previous time point to improve the detection performance, which we refer to as “re-segmentation”. The retrospective data includes 518 metastases in 184 contrast-enhanced T1-weighted MRIs originating from 49 patients (63% male, 37% female). 131 time-points (36 patients, 418 BMs) are used for cross-validation, the remaining 53 time-points (13 patients, 100 BMs) are used for testing. The lesions were manually delineated with label 1: enhancing lesion, label 2: edema, and label 3: necrosis. One-tailed t-tests are used to compare model performance including multiple segmentation and detection metrics. Significance is considered as p < 0.05. A Dice Similarity Coefficient (DSC) of 0.79 and $$F_{1}$$ F 1 -score of 0.80 are obtained for the segmentation of new lesions. On follow-up scans, the re-segmentation model significantly outperforms the segmentation model (DSC and $$F_{1}$$ F 1 0.78 and 0.88 vs 0.56 and 0.60). The re-segmentation model also significantly outperforms the simple segmentation model on the enhancing lesion (DSC 0.76 vs 0.53) and edema (0.52 vs 0.47) components, while similar scores are obtained on the necrosis component (0.62 vs 0.63). Additionally, we analyze the correlation between lesion size and segmentation performance, as demonstrated in various studies that highlight the challenges in segmenting small lesions. Our findings indicate that this correlation disappears when utilizing the re-segmentation approach and evaluating with the unbiased normalized DSC. In conclusion, the automated segmentation of new lesions and subsequent re-segmentation in follow-up images was achievable, with high level of performance obtained for single- and multiple-component segmentation tasks. |
| format | Article |
| id | doaj-art-b79b5b47c86647deb6299b5893e8145f |
| institution | Kabale University |
| issn | 2045-2322 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-b79b5b47c86647deb6299b5893e8145f2025-01-05T12:29:40ZengNature PortfolioScientific Reports2045-23222024-12-0114111010.1038/s41598-024-78865-7Automatic detection and multi-component segmentation of brain metastases in longitudinal MRIVincent Andrearczyk0Luis Schiappacasse1Daniel Abler2Marek Wodzinski3Andreas Hottinger4Matthieu Raccaud5Jean Bourhis6John O. Prior7Vincent Dunet8Adrien Depeurnge9Institute of Informatics, HES-SO Valais-Wallis University of Applied Sciences and Arts Western SwitzerlandDepartment of Radiation Oncology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Institute of Informatics, HES-SO Valais-Wallis University of Applied Sciences and Arts Western SwitzerlandInstitute of Informatics, HES-SO Valais-Wallis University of Applied Sciences and Arts Western SwitzerlandDepartment of Oncology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Department of Medical Radiology, Service of Diagnostic and Interventional Radiology, Neuroradiology Unit, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Department of Radiation Oncology, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Department of Nuclear Medicine and Molecular Imaging, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Department of Medical Radiology, Service of Diagnostic and Interventional Radiology, Neuroradiology Unit, Lausanne University Hospital (CHUV) and University of Lausanne (UNIL)Institute of Informatics, HES-SO Valais-Wallis University of Applied Sciences and Arts Western SwitzerlandAbstract Manual segmentation of lesions, required for radiotherapy planning and follow-up, is time-consuming and error-prone. Automatic detection and segmentation can assist radiologists in these tasks. This work explores the automated detection and segmentation of brain metastases (BMs) in longitudinal MRIs. It focuses on several important aspects: identifying and segmenting new lesions for screening and treatment planning, re-segmenting lesions in successive images using prior lesion locations as an additional input channel, and performing multi-component segmentation to distinguish between enhancing tissue, edema, and necrosis. The key component of the proposed approach is to propagate the lesion mask from the previous time point to improve the detection performance, which we refer to as “re-segmentation”. The retrospective data includes 518 metastases in 184 contrast-enhanced T1-weighted MRIs originating from 49 patients (63% male, 37% female). 131 time-points (36 patients, 418 BMs) are used for cross-validation, the remaining 53 time-points (13 patients, 100 BMs) are used for testing. The lesions were manually delineated with label 1: enhancing lesion, label 2: edema, and label 3: necrosis. One-tailed t-tests are used to compare model performance including multiple segmentation and detection metrics. Significance is considered as p < 0.05. A Dice Similarity Coefficient (DSC) of 0.79 and $$F_{1}$$ F 1 -score of 0.80 are obtained for the segmentation of new lesions. On follow-up scans, the re-segmentation model significantly outperforms the segmentation model (DSC and $$F_{1}$$ F 1 0.78 and 0.88 vs 0.56 and 0.60). The re-segmentation model also significantly outperforms the simple segmentation model on the enhancing lesion (DSC 0.76 vs 0.53) and edema (0.52 vs 0.47) components, while similar scores are obtained on the necrosis component (0.62 vs 0.63). Additionally, we analyze the correlation between lesion size and segmentation performance, as demonstrated in various studies that highlight the challenges in segmenting small lesions. Our findings indicate that this correlation disappears when utilizing the re-segmentation approach and evaluating with the unbiased normalized DSC. In conclusion, the automated segmentation of new lesions and subsequent re-segmentation in follow-up images was achievable, with high level of performance obtained for single- and multiple-component segmentation tasks.https://doi.org/10.1038/s41598-024-78865-7SegmentationBrain metastasesMagnetic resonance imagingDeep learning |
| spellingShingle | Vincent Andrearczyk Luis Schiappacasse Daniel Abler Marek Wodzinski Andreas Hottinger Matthieu Raccaud Jean Bourhis John O. Prior Vincent Dunet Adrien Depeurnge Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI Scientific Reports Segmentation Brain metastases Magnetic resonance imaging Deep learning |
| title | Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI |
| title_full | Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI |
| title_fullStr | Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI |
| title_full_unstemmed | Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI |
| title_short | Automatic detection and multi-component segmentation of brain metastases in longitudinal MRI |
| title_sort | automatic detection and multi component segmentation of brain metastases in longitudinal mri |
| topic | Segmentation Brain metastases Magnetic resonance imaging Deep learning |
| url | https://doi.org/10.1038/s41598-024-78865-7 |
| work_keys_str_mv | AT vincentandrearczyk automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT luisschiappacasse automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT danielabler automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT marekwodzinski automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT andreashottinger automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT matthieuraccaud automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT jeanbourhis automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT johnoprior automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT vincentdunet automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri AT adriendepeurnge automaticdetectionandmulticomponentsegmentationofbrainmetastasesinlongitudinalmri |