Frontiers in Medicine (Nov 2024)
Automated lung segmentation on chest MRI in children with cystic fibrosis
- Friedemann G. Ringwald,
- Friedemann G. Ringwald,
- Lena Wucherpfennig,
- Lena Wucherpfennig,
- Lena Wucherpfennig,
- Niclas Hagen,
- Niclas Hagen,
- Jonas Mücke,
- Sebastian Kaletta,
- Monika Eichinger,
- Monika Eichinger,
- Monika Eichinger,
- Mirjam Stahl,
- Mirjam Stahl,
- Mirjam Stahl,
- Mirjam Stahl,
- Simon M. F. Triphan,
- Simon M. F. Triphan,
- Simon M. F. Triphan,
- Patricia Leutz-Schmidt,
- Patricia Leutz-Schmidt,
- Patricia Leutz-Schmidt,
- Sonja Gestewitz,
- Sonja Gestewitz,
- Sonja Gestewitz,
- Simon Y. Graeber,
- Simon Y. Graeber,
- Simon Y. Graeber,
- Simon Y. Graeber,
- Hans-Ulrich Kauczor,
- Hans-Ulrich Kauczor,
- Hans-Ulrich Kauczor,
- Abdulsattar Alrajab,
- Jens-Peter Schenk,
- Olaf Sommerburg,
- Olaf Sommerburg,
- Olaf Sommerburg,
- Marcus A. Mall,
- Marcus A. Mall,
- Marcus A. Mall,
- Marcus A. Mall,
- Petra Knaup,
- Petra Knaup,
- Mark O. Wielpütz,
- Mark O. Wielpütz,
- Mark O. Wielpütz,
- Urs Eisenmann,
- Urs Eisenmann
Affiliations
- Friedemann G. Ringwald
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Friedemann G. Ringwald
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Lena Wucherpfennig
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Lena Wucherpfennig
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Lena Wucherpfennig
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Niclas Hagen
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Niclas Hagen
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Jonas Mücke
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Sebastian Kaletta
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Monika Eichinger
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Monika Eichinger
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Monika Eichinger
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Mirjam Stahl
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Mirjam Stahl
- Department of Pediatric Respiratory Medicine, Immunology and Critical Care Medicine, Charité-Universitätsmedizin Berlin, Berlin, Germany
- Mirjam Stahl
- German Center for Lung Research (DZL), Associated Partner Site, Berlin, Germany
- Mirjam Stahl
- Berlin Institute of Health (BIH) at Charité-Universitätsmedizin Berlin, Berlin, Germany
- Simon M. F. Triphan
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Simon M. F. Triphan
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Simon M. F. Triphan
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Patricia Leutz-Schmidt
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Patricia Leutz-Schmidt
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Patricia Leutz-Schmidt
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Sonja Gestewitz
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Sonja Gestewitz
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Sonja Gestewitz
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Simon Y. Graeber
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Simon Y. Graeber
- Department of Pediatric Respiratory Medicine, Immunology and Critical Care Medicine, Charité-Universitätsmedizin Berlin, Berlin, Germany
- Simon Y. Graeber
- German Center for Lung Research (DZL), Associated Partner Site, Berlin, Germany
- Simon Y. Graeber
- Berlin Institute of Health (BIH) at Charité-Universitätsmedizin Berlin, Berlin, Germany
- Hans-Ulrich Kauczor
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Hans-Ulrich Kauczor
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Hans-Ulrich Kauczor
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Abdulsattar Alrajab
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Jens-Peter Schenk
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Olaf Sommerburg
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Olaf Sommerburg
- Division of Pediatric Pulmonology & Allergy and Cystic Fibrosis Center, Department of Pediatrics, University Hospital Heidelberg, Heidelberg, Germany
- Olaf Sommerburg
- Department of Translational Pulmonology, University Hospital Heidelberg, Heidelberg, Germany
- Marcus A. Mall
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Marcus A. Mall
- Department of Pediatric Respiratory Medicine, Immunology and Critical Care Medicine, Charité-Universitätsmedizin Berlin, Berlin, Germany
- Marcus A. Mall
- German Center for Lung Research (DZL), Associated Partner Site, Berlin, Germany
- Marcus A. Mall
- Berlin Institute of Health (BIH) at Charité-Universitätsmedizin Berlin, Berlin, Germany
- Petra Knaup
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Petra Knaup
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Mark O. Wielpütz
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- Mark O. Wielpütz
- Department of Diagnostic and Interventional Radiology, University Hospital Heidelberg, Heidelberg, Germany
- Mark O. Wielpütz
- Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University Hospital Heidelberg, Heidelberg, Germany
- Urs Eisenmann
- Institute of Medical Informatics, Heidelberg University, Heidelberg, Germany
- Urs Eisenmann
- Translational Lung Research Center Heidelberg (TLRC), German Center for Lung Research (DZL), Heidelberg, Germany
- DOI
- https://doi.org/10.3389/fmed.2024.1401473
- Journal volume & issue
-
Vol. 11
Abstract
IntroductionSegmentation of lung structures in medical imaging is crucial for the application of automated post-processing steps on lung diseases like cystic fibrosis (CF). Recently, machine learning methods, particularly neural networks, have demonstrated remarkable improvements, often outperforming conventional segmentation methods. Nonetheless, challenges still remain when attempting to segment various imaging modalities and diseases, especially when the visual characteristics of pathologic findings significantly deviate from healthy tissue.MethodsOur study focuses on imaging of pediatric CF patients [mean age, standard deviation (7.50 ± 4.6)], utilizing deep learning-based methods for automated lung segmentation from chest magnetic resonance imaging (MRI). A total of 165 standardized annual surveillance MRI scans from 84 patients with CF were segmented using the nnU-Net framework. Patient cases represented a range of disease severities and ages. The nnU-Net was trained and evaluated on three MRI sequences (BLADE, VIBE, and HASTE), which are highly relevant for the evaluation of CF induced lung changes. We utilized 40 cases for training per sequence, and tested with 15 cases per sequence, using the Sørensen-Dice-Score, Pearson’s correlation coefficient (r), a segmentation questionnaire, and slice-based analysis.ResultsThe results demonstrated a high level of segmentation performance across all sequences, with only minor differences observed in the mean Dice coefficient: BLADE (0.96 ± 0.05), VIBE (0.96 ± 0.04), and HASTE (0.95 ± 0.05). Additionally, the segmentation quality was consistent across different disease severities, patient ages, and sizes. Manual evaluation identified specific challenges, such as incomplete segmentations near the diaphragm and dorsal regions. Validation on a separate, external dataset of nine toddlers (2–24 months) demonstrated generalizability of the trained model achieving a Dice coefficient of 0.85 ± 0.03.Discussion and conclusionOverall, our study demonstrates the feasibility and effectiveness of using nnU-Net for automated segmentation of lung halves in pediatric CF patients, showing promising directions for advanced image analysis techniques to assist in clinical decision-making and monitoring of CF lung disease progression. Despite these achievements, further improvements are needed to address specific segmentation challenges and enhance generalizability.
Keywords