Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks
Background: in magnetic resonance imaging (MRI), automated detection of brain metastases with convolutional neural networks (CNN) represents an extraordinary challenge due to small lesions sometimes posing as brain vessels as well as other confounders. Literature reporting high false positive rates...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-06-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/11/6/1016 |
| _version_ | 1797531636991000576 |
|---|---|
| author | Jonathan Kottlors Simon Geissen Hannah Jendreizik Nils Große Hokamp Philipp Fervers Lenhard Pennig Kai Laukamp Christoph Kabbasch David Maintz Marc Schlamann Jan Borggrefe |
| author_facet | Jonathan Kottlors Simon Geissen Hannah Jendreizik Nils Große Hokamp Philipp Fervers Lenhard Pennig Kai Laukamp Christoph Kabbasch David Maintz Marc Schlamann Jan Borggrefe |
| author_sort | Jonathan Kottlors |
| collection | DOAJ |
| description | Background: in magnetic resonance imaging (MRI), automated detection of brain metastases with convolutional neural networks (CNN) represents an extraordinary challenge due to small lesions sometimes posing as brain vessels as well as other confounders. Literature reporting high false positive rates when using conventional contrast enhanced (CE) T1 sequences questions their usefulness in clinical routine. CE black blood (BB) sequences may overcome these limitations by suppressing contrast-enhanced structures, thus facilitating lesion detection. This study compared CNN performance in conventional CE T1 and BB sequences and tested for objective improvement of brain lesion detection. Methods: we included a subgroup of 127 consecutive patients, receiving both CE T1 and BB sequences, referred for MRI concerning metastatic spread to the brain. A pretrained CNN was retrained with a customized monolayer classifier using either T1 or BB scans of brain lesions. Results: CE T1 imaging-based training resulted in an internal validation accuracy of 85.5% vs. 92.3% in BB imaging (<i>p</i> < 0.01). In holdout validation analysis, T1 image-based prediction presented poor specificity and sensitivity with an AUC of 0.53 compared to 0.87 in BB-imaging-based prediction. Conclusions: detection of brain lesions with CNN, BB-MRI imaging represents a highly effective input type when compared to conventional CE T1-MRI imaging. Use of BB-MRI can overcome the current limitations for automated brain lesion detection and the objectively excellent performance of our CNN suggests routine usage of BB sequences for radiological analysis. |
| first_indexed | 2024-03-10T10:47:32Z |
| format | Article |
| id | doaj.art-180bf0c1325a4f08bdc8b36e5a8e022c |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-10T10:47:32Z |
| publishDate | 2021-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-180bf0c1325a4f08bdc8b36e5a8e022c2023-11-21T22:27:54ZengMDPI AGDiagnostics2075-44182021-06-01116101610.3390/diagnostics11061016Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural NetworksJonathan Kottlors0Simon Geissen1Hannah Jendreizik2Nils Große Hokamp3Philipp Fervers4Lenhard Pennig5Kai Laukamp6Christoph Kabbasch7David Maintz8Marc Schlamann9Jan Borggrefe10Institute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyDivision of Cardiology, Pneumology, Angiology and Intensive Care, University of Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitute for Diagnostic and Interventional Radiology, University Hospital Cologne, Kerpener Strasse 62, 50935 Cologne, GermanyInstitut für Radiologie Neuroradiologie und Nuklearmedizin, Ruhr-Universität Bochum, Minden University Hospital, Hans-Nolte-Strasse 1, 32429 Minden, GermanyBackground: in magnetic resonance imaging (MRI), automated detection of brain metastases with convolutional neural networks (CNN) represents an extraordinary challenge due to small lesions sometimes posing as brain vessels as well as other confounders. Literature reporting high false positive rates when using conventional contrast enhanced (CE) T1 sequences questions their usefulness in clinical routine. CE black blood (BB) sequences may overcome these limitations by suppressing contrast-enhanced structures, thus facilitating lesion detection. This study compared CNN performance in conventional CE T1 and BB sequences and tested for objective improvement of brain lesion detection. Methods: we included a subgroup of 127 consecutive patients, receiving both CE T1 and BB sequences, referred for MRI concerning metastatic spread to the brain. A pretrained CNN was retrained with a customized monolayer classifier using either T1 or BB scans of brain lesions. Results: CE T1 imaging-based training resulted in an internal validation accuracy of 85.5% vs. 92.3% in BB imaging (<i>p</i> < 0.01). In holdout validation analysis, T1 image-based prediction presented poor specificity and sensitivity with an AUC of 0.53 compared to 0.87 in BB-imaging-based prediction. Conclusions: detection of brain lesions with CNN, BB-MRI imaging represents a highly effective input type when compared to conventional CE T1-MRI imaging. Use of BB-MRI can overcome the current limitations for automated brain lesion detection and the objectively excellent performance of our CNN suggests routine usage of BB sequences for radiological analysis.https://www.mdpi.com/2075-4418/11/6/1016magnetic resonance imagingconvolutional neural networksautomated detection of brain metastases |
| spellingShingle | Jonathan Kottlors Simon Geissen Hannah Jendreizik Nils Große Hokamp Philipp Fervers Lenhard Pennig Kai Laukamp Christoph Kabbasch David Maintz Marc Schlamann Jan Borggrefe Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks Diagnostics magnetic resonance imaging convolutional neural networks automated detection of brain metastases |
| title | Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks |
| title_full | Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks |
| title_fullStr | Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks |
| title_full_unstemmed | Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks |
| title_short | Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks |
| title_sort | contrast enhanced black blood mri sequence is superior to conventional t1 sequence in automated detection of brain metastases by convolutional neural networks |
| topic | magnetic resonance imaging convolutional neural networks automated detection of brain metastases |
| url | https://www.mdpi.com/2075-4418/11/6/1016 |
| work_keys_str_mv | AT jonathankottlors contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT simongeissen contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT hannahjendreizik contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT nilsgroßehokamp contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT philippfervers contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT lenhardpennig contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT kailaukamp contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT christophkabbasch contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT davidmaintz contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT marcschlamann contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks AT janborggrefe contrastenhancedblackbloodmrisequenceissuperiortoconventionalt1sequenceinautomateddetectionofbrainmetastasesbyconvolutionalneuralnetworks |