Accelerated diffusion tensor imaging with self-supervision and fine-tuning
Abstract Diffusion tensor imaging (DTI) is essential for assessing brain microstructure but requires long acquisition times, limiting clinical use. Recent deep learning (DL) approaches, such as SuperDTI or deepDTI, improve DTI metrics but demand large, high-quality datasets for training. We propose...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2025-04-01
|
| Series: | Scientific Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1038/s41598-025-96459-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850181699430776832 |
|---|---|
| author | Phillip Martin Diego Martin Maria Altbach Ali Bilgin |
| author_facet | Phillip Martin Diego Martin Maria Altbach Ali Bilgin |
| author_sort | Phillip Martin |
| collection | DOAJ |
| description | Abstract Diffusion tensor imaging (DTI) is essential for assessing brain microstructure but requires long acquisition times, limiting clinical use. Recent deep learning (DL) approaches, such as SuperDTI or deepDTI, improve DTI metrics but demand large, high-quality datasets for training. We propose a self-supervised deep learning with fine-tuning (SSDLFT) framework to reduce training data requirements. SSDLFT involves self-supervised pretraining, which denoises data without clean labels, followed by fine-tuning with limited high-quality data. Experiments using Human Connectome Project data show that SSDLFT outperforms traditional methods and other DL approaches in qualitative and quantitative assessments of DWI reconstructions and tensor metrics. SSDLFT’s ability to maintain high performance with fewer training subjects and DWIs presents a significant advancement, enhancing DTI’s practical applications in clinical and research settings. |
| format | Article |
| id | doaj-art-05b3a7e19e4d40f9b12105f5ea88e00e |
| institution | OA Journals |
| issn | 2045-2322 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj-art-05b3a7e19e4d40f9b12105f5ea88e00e2025-08-20T02:17:50ZengNature PortfolioScientific Reports2045-23222025-04-0115111210.1038/s41598-025-96459-9Accelerated diffusion tensor imaging with self-supervision and fine-tuningPhillip Martin0Diego Martin1Maria Altbach2Ali Bilgin3Department of Radiology, Houston Methodist Research InstituteDepartment of Radiology, Houston Methodist Research InstituteDepartment of Biomedical Engineering, University of ArizonaDepartment of Electrical and Computer Engineering, University of ArizonaAbstract Diffusion tensor imaging (DTI) is essential for assessing brain microstructure but requires long acquisition times, limiting clinical use. Recent deep learning (DL) approaches, such as SuperDTI or deepDTI, improve DTI metrics but demand large, high-quality datasets for training. We propose a self-supervised deep learning with fine-tuning (SSDLFT) framework to reduce training data requirements. SSDLFT involves self-supervised pretraining, which denoises data without clean labels, followed by fine-tuning with limited high-quality data. Experiments using Human Connectome Project data show that SSDLFT outperforms traditional methods and other DL approaches in qualitative and quantitative assessments of DWI reconstructions and tensor metrics. SSDLFT’s ability to maintain high performance with fewer training subjects and DWIs presents a significant advancement, enhancing DTI’s practical applications in clinical and research settings.https://doi.org/10.1038/s41598-025-96459-9Diffusion tensor imaging (DTI)Deep learning (DL)Self-supervised learningFractional anisotropy (FA)Mean diffusivity (MD) |
| spellingShingle | Phillip Martin Diego Martin Maria Altbach Ali Bilgin Accelerated diffusion tensor imaging with self-supervision and fine-tuning Scientific Reports Diffusion tensor imaging (DTI) Deep learning (DL) Self-supervised learning Fractional anisotropy (FA) Mean diffusivity (MD) |
| title | Accelerated diffusion tensor imaging with self-supervision and fine-tuning |
| title_full | Accelerated diffusion tensor imaging with self-supervision and fine-tuning |
| title_fullStr | Accelerated diffusion tensor imaging with self-supervision and fine-tuning |
| title_full_unstemmed | Accelerated diffusion tensor imaging with self-supervision and fine-tuning |
| title_short | Accelerated diffusion tensor imaging with self-supervision and fine-tuning |
| title_sort | accelerated diffusion tensor imaging with self supervision and fine tuning |
| topic | Diffusion tensor imaging (DTI) Deep learning (DL) Self-supervised learning Fractional anisotropy (FA) Mean diffusivity (MD) |
| url | https://doi.org/10.1038/s41598-025-96459-9 |
| work_keys_str_mv | AT phillipmartin accelerateddiffusiontensorimagingwithselfsupervisionandfinetuning AT diegomartin accelerateddiffusiontensorimagingwithselfsupervisionandfinetuning AT mariaaltbach accelerateddiffusiontensorimagingwithselfsupervisionandfinetuning AT alibilgin accelerateddiffusiontensorimagingwithselfsupervisionandfinetuning |