Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification
Super-resolution (SR) techniques have gained traction in biomedical imaging for their ability to enhance image quality. However, it remains unclear whether these improvements translate into better performance in clinical tasks. In this study, we provide a comprehensive evaluation of state-of-the-art...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-03-01
|
| Series: | Journal of Imaging |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-433X/11/4/104 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849714209729806336 |
|---|---|
| author | Mario Amoros Manuel Curado Jose F. Vicent |
| author_facet | Mario Amoros Manuel Curado Jose F. Vicent |
| author_sort | Mario Amoros |
| collection | DOAJ |
| description | Super-resolution (SR) techniques have gained traction in biomedical imaging for their ability to enhance image quality. However, it remains unclear whether these improvements translate into better performance in clinical tasks. In this study, we provide a comprehensive evaluation of state-of-the-art SR models—including CNN- and Transformer-based architectures—by assessing not only visual quality metrics (PSNR and SSIM) but also their downstream impact on segmentation and classification performance for lung CT scans. Using U-Net and ResNet architectures, we quantify how SR influences diagnostic tasks across different datasets, and we evaluate model generalization in cross-domain settings. Our findings show that advanced SR models such as SwinIR preserve diagnostic features effectively and, when appropriately applied, can enhance or maintain clinical performance even in low-resolution contexts. This work bridges the gap between image quality enhancement and practical clinical utility, providing actionable insights for integrating SR into real-world biomedical imaging workflows. |
| format | Article |
| id | doaj-art-32b5bda80c954497a981ec2967c837f6 |
| institution | DOAJ |
| issn | 2313-433X |
| language | English |
| publishDate | 2025-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Imaging |
| spelling | doaj-art-32b5bda80c954497a981ec2967c837f62025-08-20T03:13:45ZengMDPI AGJournal of Imaging2313-433X2025-03-0111410410.3390/jimaging11040104Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and ClassificationMario Amoros0Manuel Curado1Jose F. Vicent2Department of Computer Science and Artificial Intelligence, University of Alicante, Campus de San Vicente del Raspeig, Ap. Correos 99, E-03080 Alicante, SpainDepartment of Computer Science and Artificial Intelligence, University of Alicante, Campus de San Vicente del Raspeig, Ap. Correos 99, E-03080 Alicante, SpainDepartment of Computer Science and Artificial Intelligence, University of Alicante, Campus de San Vicente del Raspeig, Ap. Correos 99, E-03080 Alicante, SpainSuper-resolution (SR) techniques have gained traction in biomedical imaging for their ability to enhance image quality. However, it remains unclear whether these improvements translate into better performance in clinical tasks. In this study, we provide a comprehensive evaluation of state-of-the-art SR models—including CNN- and Transformer-based architectures—by assessing not only visual quality metrics (PSNR and SSIM) but also their downstream impact on segmentation and classification performance for lung CT scans. Using U-Net and ResNet architectures, we quantify how SR influences diagnostic tasks across different datasets, and we evaluate model generalization in cross-domain settings. Our findings show that advanced SR models such as SwinIR preserve diagnostic features effectively and, when appropriately applied, can enhance or maintain clinical performance even in low-resolution contexts. This work bridges the gap between image quality enhancement and practical clinical utility, providing actionable insights for integrating SR into real-world biomedical imaging workflows.https://www.mdpi.com/2313-433X/11/4/104super-resolutionsegmentationclassification |
| spellingShingle | Mario Amoros Manuel Curado Jose F. Vicent Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification Journal of Imaging super-resolution segmentation classification |
| title | Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification |
| title_full | Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification |
| title_fullStr | Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification |
| title_full_unstemmed | Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification |
| title_short | Evaluating Super-Resolution Models in Biomedical Imaging: Applications and Performance in Segmentation and Classification |
| title_sort | evaluating super resolution models in biomedical imaging applications and performance in segmentation and classification |
| topic | super-resolution segmentation classification |
| url | https://www.mdpi.com/2313-433X/11/4/104 |
| work_keys_str_mv | AT marioamoros evaluatingsuperresolutionmodelsinbiomedicalimagingapplicationsandperformanceinsegmentationandclassification AT manuelcurado evaluatingsuperresolutionmodelsinbiomedicalimagingapplicationsandperformanceinsegmentationandclassification AT josefvicent evaluatingsuperresolutionmodelsinbiomedicalimagingapplicationsandperformanceinsegmentationandclassification |