Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency
Deep learning methods provide significant assistance in analyzing coronavirus disease (COVID-19) in chest computed tomography (CT) images, including identification, severity assessment, and segmentation. Although the earlier developed methods address the lack of data and specific annotations, the cu...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Journal of Imaging |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-433X/8/9/234 |
| _version_ | 1797486419371884544 |
|---|---|
| author | Stanislav Shimovolos Andrey Shushko Mikhail Belyaev Boris Shirokikh |
| author_facet | Stanislav Shimovolos Andrey Shushko Mikhail Belyaev Boris Shirokikh |
| author_sort | Stanislav Shimovolos |
| collection | DOAJ |
| description | Deep learning methods provide significant assistance in analyzing coronavirus disease (COVID-19) in chest computed tomography (CT) images, including identification, severity assessment, and segmentation. Although the earlier developed methods address the lack of data and specific annotations, the current goal is to build a robust algorithm for clinical use, having a larger pool of available data. With the larger datasets, the domain shift problem arises, affecting the performance of methods on the unseen data. One of the critical sources of domain shift in CT images is the difference in reconstruction kernels used to generate images from the raw data (sinograms). In this paper, we show a decrease in the COVID-19 segmentation quality of the model trained on the smooth and tested on the sharp reconstruction kernels. Furthermore, we compare several domain adaptation approaches to tackle the problem, such as task-specific augmentation and unsupervised adversarial learning. Finally, we propose the unsupervised adaptation method, called F-Consistency, that outperforms the previous approaches. Our method exploits a set of unlabeled CT image pairs which differ only in reconstruction kernels within every pair. It enforces the similarity of the network’s hidden representations (feature maps) by minimizing the mean squared error (MSE) between paired feature maps. We show our method achieving a 0.64 Dice Score on the test dataset with unseen sharp kernels, compared to the 0.56 Dice Score of the baseline model. Moreover, F-Consistency scores 0.80 Dice Score between predictions on the paired images, which almost doubles the baseline score of 0.46 and surpasses the other methods. We also show F-Consistency to better generalize on the unseen kernels and without the presence of the COVID-19 lesions than the other methods trained on unlabeled data. |
| first_indexed | 2024-03-09T23:33:57Z |
| format | Article |
| id | doaj.art-37eac7ab297746fcad789009ba1180ba |
| institution | Directory Open Access Journal |
| issn | 2313-433X |
| language | English |
| last_indexed | 2024-03-09T23:33:57Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Imaging |
| spelling | doaj.art-37eac7ab297746fcad789009ba1180ba2023-11-23T17:04:32ZengMDPI AGJournal of Imaging2313-433X2022-08-018923410.3390/jimaging8090234Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps ConsistencyStanislav Shimovolos0Andrey Shushko1Mikhail Belyaev2Boris Shirokikh3Moscow Institute of Physics and Technology, 141701 Moscow, RussiaMoscow Institute of Physics and Technology, 141701 Moscow, RussiaSkolkovo Institute of Science and Technology, 143026 Moscow, RussiaSkolkovo Institute of Science and Technology, 143026 Moscow, RussiaDeep learning methods provide significant assistance in analyzing coronavirus disease (COVID-19) in chest computed tomography (CT) images, including identification, severity assessment, and segmentation. Although the earlier developed methods address the lack of data and specific annotations, the current goal is to build a robust algorithm for clinical use, having a larger pool of available data. With the larger datasets, the domain shift problem arises, affecting the performance of methods on the unseen data. One of the critical sources of domain shift in CT images is the difference in reconstruction kernels used to generate images from the raw data (sinograms). In this paper, we show a decrease in the COVID-19 segmentation quality of the model trained on the smooth and tested on the sharp reconstruction kernels. Furthermore, we compare several domain adaptation approaches to tackle the problem, such as task-specific augmentation and unsupervised adversarial learning. Finally, we propose the unsupervised adaptation method, called F-Consistency, that outperforms the previous approaches. Our method exploits a set of unlabeled CT image pairs which differ only in reconstruction kernels within every pair. It enforces the similarity of the network’s hidden representations (feature maps) by minimizing the mean squared error (MSE) between paired feature maps. We show our method achieving a 0.64 Dice Score on the test dataset with unseen sharp kernels, compared to the 0.56 Dice Score of the baseline model. Moreover, F-Consistency scores 0.80 Dice Score between predictions on the paired images, which almost doubles the baseline score of 0.46 and surpasses the other methods. We also show F-Consistency to better generalize on the unseen kernels and without the presence of the COVID-19 lesions than the other methods trained on unlabeled data.https://www.mdpi.com/2313-433X/8/9/234chest computed tomographyconvolutional neural networkCOVID-19 segmentationdomain adaptation |
| spellingShingle | Stanislav Shimovolos Andrey Shushko Mikhail Belyaev Boris Shirokikh Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency Journal of Imaging chest computed tomography convolutional neural network COVID-19 segmentation domain adaptation |
| title | Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency |
| title_full | Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency |
| title_fullStr | Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency |
| title_full_unstemmed | Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency |
| title_short | Adaptation to CT Reconstruction Kernels by Enforcing Cross-Domain Feature Maps Consistency |
| title_sort | adaptation to ct reconstruction kernels by enforcing cross domain feature maps consistency |
| topic | chest computed tomography convolutional neural network COVID-19 segmentation domain adaptation |
| url | https://www.mdpi.com/2313-433X/8/9/234 |
| work_keys_str_mv | AT stanislavshimovolos adaptationtoctreconstructionkernelsbyenforcingcrossdomainfeaturemapsconsistency AT andreyshushko adaptationtoctreconstructionkernelsbyenforcingcrossdomainfeaturemapsconsistency AT mikhailbelyaev adaptationtoctreconstructionkernelsbyenforcingcrossdomainfeaturemapsconsistency AT borisshirokikh adaptationtoctreconstructionkernelsbyenforcingcrossdomainfeaturemapsconsistency |