Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset
Abstract Brain extraction, or skull-stripping, is an essential data preprocessing step for machine learning approaches to brain MRI analysis. Currently, there are limited extraction algorithms for the neonatal brain. We aim to adapt an established deep learning algorithm for the automatic segmentati...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-02-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-024-54436-8 |
| _version_ | 1797275139755212800 |
|---|---|
| author | Joshua V. Chen Yi Li Felicia Tang Gunvant Chaudhari Christopher Lew Amanda Lee Andreas M. Rauschecker Aden P. Haskell-Mendoza Yvonne W. Wu Evan Calabrese |
| author_facet | Joshua V. Chen Yi Li Felicia Tang Gunvant Chaudhari Christopher Lew Amanda Lee Andreas M. Rauschecker Aden P. Haskell-Mendoza Yvonne W. Wu Evan Calabrese |
| author_sort | Joshua V. Chen |
| collection | DOAJ |
| description | Abstract Brain extraction, or skull-stripping, is an essential data preprocessing step for machine learning approaches to brain MRI analysis. Currently, there are limited extraction algorithms for the neonatal brain. We aim to adapt an established deep learning algorithm for the automatic segmentation of neonatal brains from MRI, trained on a large multi-institutional dataset for improved generalizability across image acquisition parameters. Our model, ANUBEX (automated neonatal nnU-Net brain MRI extractor), was designed using nnU-Net and was trained on a subset of participants (N = 433) enrolled in the High-dose Erythropoietin for Asphyxia and Encephalopathy (HEAL) study. We compared the performance of our model to five publicly available models (BET, BSE, CABINET, iBEATv2, ROBEX) across conventional and machine learning methods, tested on two public datasets (NIH and dHCP). We found that our model had a significantly higher Dice score on the aggregate of both data sets and comparable or significantly higher Dice scores on the NIH (low-resolution) and dHCP (high-resolution) datasets independently. ANUBEX performs similarly when trained on sequence-agnostic or motion-degraded MRI, but slightly worse on preterm brains. In conclusion, we created an automatic deep learning-based neonatal brain extraction algorithm that demonstrates accurate performance with both high- and low-resolution MRIs with fast computation time. |
| first_indexed | 2024-03-07T15:09:03Z |
| format | Article |
| id | doaj.art-589caae17a3a4f28953eb0778a03a7b9 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-07T15:09:03Z |
| publishDate | 2024-02-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-589caae17a3a4f28953eb0778a03a7b92024-03-05T18:43:59ZengNature PortfolioScientific Reports2045-23222024-02-0114111010.1038/s41598-024-54436-8Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional datasetJoshua V. Chen0Yi Li1Felicia Tang2Gunvant Chaudhari3Christopher Lew4Amanda Lee5Andreas M. Rauschecker6Aden P. Haskell-Mendoza7Yvonne W. Wu8Evan Calabrese9Department of Radiology, University of California San FranciscoDepartment of Radiology, University of California San FranciscoDepartment of Radiology, University of California San FranciscoDepartment of Radiology, University of California San FranciscoDivision of Neuroradiology, Department of Radiology, Duke University Medical CenterDivision of Neuroradiology, Department of Radiology, Duke University Medical CenterDepartment of Radiology, University of California San FranciscoDuke University School of MedicineUniversity of California San Francisco Weill Institute for NeurosciencesDivision of Neuroradiology, Department of Radiology, Duke University Medical CenterAbstract Brain extraction, or skull-stripping, is an essential data preprocessing step for machine learning approaches to brain MRI analysis. Currently, there are limited extraction algorithms for the neonatal brain. We aim to adapt an established deep learning algorithm for the automatic segmentation of neonatal brains from MRI, trained on a large multi-institutional dataset for improved generalizability across image acquisition parameters. Our model, ANUBEX (automated neonatal nnU-Net brain MRI extractor), was designed using nnU-Net and was trained on a subset of participants (N = 433) enrolled in the High-dose Erythropoietin for Asphyxia and Encephalopathy (HEAL) study. We compared the performance of our model to five publicly available models (BET, BSE, CABINET, iBEATv2, ROBEX) across conventional and machine learning methods, tested on two public datasets (NIH and dHCP). We found that our model had a significantly higher Dice score on the aggregate of both data sets and comparable or significantly higher Dice scores on the NIH (low-resolution) and dHCP (high-resolution) datasets independently. ANUBEX performs similarly when trained on sequence-agnostic or motion-degraded MRI, but slightly worse on preterm brains. In conclusion, we created an automatic deep learning-based neonatal brain extraction algorithm that demonstrates accurate performance with both high- and low-resolution MRIs with fast computation time.https://doi.org/10.1038/s41598-024-54436-8 |
| spellingShingle | Joshua V. Chen Yi Li Felicia Tang Gunvant Chaudhari Christopher Lew Amanda Lee Andreas M. Rauschecker Aden P. Haskell-Mendoza Yvonne W. Wu Evan Calabrese Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset Scientific Reports |
| title | Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset |
| title_full | Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset |
| title_fullStr | Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset |
| title_full_unstemmed | Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset |
| title_short | Automated neonatal nnU-Net brain MRI extractor trained on a large multi-institutional dataset |
| title_sort | automated neonatal nnu net brain mri extractor trained on a large multi institutional dataset |
| url | https://doi.org/10.1038/s41598-024-54436-8 |
| work_keys_str_mv | AT joshuavchen automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT yili automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT feliciatang automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT gunvantchaudhari automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT christopherlew automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT amandalee automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT andreasmrauschecker automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT adenphaskellmendoza automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT yvonnewwu automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset AT evancalabrese automatedneonatalnnunetbrainmriextractortrainedonalargemultiinstitutionaldataset |