Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation
Medical image segmentation is a crucial element of computer-aided diagnosis (CAD) systems. Segmentation maps are used to calculate imaging features, such as quantitative disease distribution and radiomic features. Since their introduction in 2015, UNets have become the state-of-the-art segmentation...
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10381702/ |
| _version_ | 1797296354308915200 |
|---|---|
| author | M. Fontanellaz A. Christe S. Christodoulidis E. Dack J. Roos D. Drakopoulos D. Sieron A. Peters T. Geiser M. Funke-Chambour J. Heverhagen H. Hoppe A. K. Exadaktylos L. Ebner S. Mougiakakou |
| author_facet | M. Fontanellaz A. Christe S. Christodoulidis E. Dack J. Roos D. Drakopoulos D. Sieron A. Peters T. Geiser M. Funke-Chambour J. Heverhagen H. Hoppe A. K. Exadaktylos L. Ebner S. Mougiakakou |
| author_sort | M. Fontanellaz |
| collection | DOAJ |
| description | Medical image segmentation is a crucial element of computer-aided diagnosis (CAD) systems. Segmentation maps are used to calculate imaging features, such as quantitative disease distribution and radiomic features. Since their introduction in 2015, UNets have become the state-of-the-art segmentation tools. However, since that time, many new methods for image processing have been introduced, such as vision transformers and multi-layer-perceptron-mixers (MLP-Mixers). Alongside baseline UNets, we have now investigated the application of such MLP-Mixers for medical image segmentation, as part of a CAD system for the diagnosis of interstitial lung diseases (ILDs). Furthermore, we have investigated the effect of 2D and 3D data representations on segmentation and the final CAD results. We have evaluated the performance of the baseline segmentation methods and the MLP-Mixer primary on the overall diagnostic performance of the CAD system - as well as on the accuracy of segmentation as an intermediate step. In addition to network and data representation variations, we have investigated two different techniques for selecting features, an agnostic method and an alternative approach which selects features tailored to a specific segmentation map and diagnosis task. Finally, the CAD’s performance was compared with that of four independent specialists in chest radiology. Among the 105 test cases, the diagnostic accuracy was 77.2±1.6% for the AI-approaches and 79.0±6.9% for the radiologists, indicating that the proposed systems perform comparably well to human readers in most of the cases. For the task of ILD pattern segmentation, similar results were obtained with 3D data and 2D tomography slices. |
| first_indexed | 2024-03-07T22:03:26Z |
| format | Article |
| id | doaj.art-a0f65a8d8101414780b1d2a877e3e9c0 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-03-07T22:03:26Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-a0f65a8d8101414780b1d2a877e3e9c02024-02-24T00:00:44ZengIEEEIEEE Access2169-35362024-01-0112256422565610.1109/ACCESS.2024.335043010381702Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical EvaluationM. Fontanellaz0https://orcid.org/0000-0001-9347-4851A. Christe1https://orcid.org/0000-0002-2355-2591S. Christodoulidis2E. Dack3J. Roos4https://orcid.org/0000-0003-3122-6611D. Drakopoulos5D. Sieron6A. Peters7https://orcid.org/0000-0002-9864-6381T. Geiser8M. Funke-Chambour9https://orcid.org/0000-0003-3417-5872J. Heverhagen10H. Hoppe11A. K. Exadaktylos12L. Ebner13S. Mougiakakou14https://orcid.org/0000-0002-6355-9982AI in Health and Nutrition, ARTORG Center for Biomedical Engineering Research, University of Bern, Bern, SwitzerlandDepartment of Diagnostic, Interventional, and Pediatric Radiology, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandLaboratory of Mathematics and Informatics (MICS), CentraleSupélec, Université Paris Saclay, Gif-sur-Yvette, FranceAI in Health and Nutrition, ARTORG Center for Biomedical Engineering Research, University of Bern, Bern, SwitzerlandInstitute of Radiology and Nuclear Medicine, Luzerner Kantonsspital, Lucerne, Canton of Lucerne, SwitzerlandDepartment of Diagnostic, Interventional, and Pediatric Radiology, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandDivision of Magnetic Resonance Imaging, Silesian Center for Heart Diseases, Zabrze, PolandDepartment of Diagnostic, Interventional, and Pediatric Radiology, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandDepartment of Pulmonary Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandDepartment of Pulmonary Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandDepartment of Diagnostic, Interventional, and Pediatric Radiology, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandDepartment of Radiology, Lindenhofspital, Bern, SwitzerlandDepartment of Emergency Medicine, Inselspital, University Hospital Bern, University of Bern, Bern, SwitzerlandDepartment of Diagnostic, Interventional, and Pediatric Radiology, Inselspital, Bern University Hospital, University of Bern, Bern, SwitzerlandAI in Health and Nutrition, ARTORG Center for Biomedical Engineering Research, University of Bern, Bern, SwitzerlandMedical image segmentation is a crucial element of computer-aided diagnosis (CAD) systems. Segmentation maps are used to calculate imaging features, such as quantitative disease distribution and radiomic features. Since their introduction in 2015, UNets have become the state-of-the-art segmentation tools. However, since that time, many new methods for image processing have been introduced, such as vision transformers and multi-layer-perceptron-mixers (MLP-Mixers). Alongside baseline UNets, we have now investigated the application of such MLP-Mixers for medical image segmentation, as part of a CAD system for the diagnosis of interstitial lung diseases (ILDs). Furthermore, we have investigated the effect of 2D and 3D data representations on segmentation and the final CAD results. We have evaluated the performance of the baseline segmentation methods and the MLP-Mixer primary on the overall diagnostic performance of the CAD system - as well as on the accuracy of segmentation as an intermediate step. In addition to network and data representation variations, we have investigated two different techniques for selecting features, an agnostic method and an alternative approach which selects features tailored to a specific segmentation map and diagnosis task. Finally, the CAD’s performance was compared with that of four independent specialists in chest radiology. Among the 105 test cases, the diagnostic accuracy was 77.2±1.6% for the AI-approaches and 79.0±6.9% for the radiologists, indicating that the proposed systems perform comparably well to human readers in most of the cases. For the task of ILD pattern segmentation, similar results were obtained with 3D data and 2D tomography slices.https://ieeexplore.ieee.org/document/10381702/Chest X-raychest computed tomographyinterstitial lung diseasesradiomicscomputer-aided diagnosissegmentation |
| spellingShingle | M. Fontanellaz A. Christe S. Christodoulidis E. Dack J. Roos D. Drakopoulos D. Sieron A. Peters T. Geiser M. Funke-Chambour J. Heverhagen H. Hoppe A. K. Exadaktylos L. Ebner S. Mougiakakou Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation IEEE Access Chest X-ray chest computed tomography interstitial lung diseases radiomics computer-aided diagnosis segmentation |
| title | Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation |
| title_full | Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation |
| title_fullStr | Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation |
| title_full_unstemmed | Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation |
| title_short | Computer-Aided Diagnosis System for Lung Fibrosis: From the Effect of Radiomic Features and Multi-Layer-Perceptron Mixers to Pre-Clinical Evaluation |
| title_sort | computer aided diagnosis system for lung fibrosis from the effect of radiomic features and multi layer perceptron mixers to pre clinical evaluation |
| topic | Chest X-ray chest computed tomography interstitial lung diseases radiomics computer-aided diagnosis segmentation |
| url | https://ieeexplore.ieee.org/document/10381702/ |
| work_keys_str_mv | AT mfontanellaz computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT achriste computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT schristodoulidis computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT edack computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT jroos computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT ddrakopoulos computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT dsieron computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT apeters computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT tgeiser computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT mfunkechambour computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT jheverhagen computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT hhoppe computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT akexadaktylos computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT lebner computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation AT smougiakakou computeraideddiagnosissystemforlungfibrosisfromtheeffectofradiomicfeaturesandmultilayerperceptronmixerstopreclinicalevaluation |