Dataset Information

An intensity-based self-supervised domain adaptation method for intervertebral disc segmentation in magnetic resonance imaging.

ABSTRACT:

Background and objective

Accurate IVD segmentation is crucial for diagnosing and treating spinal conditions. Traditional deep learning methods depend on extensive, annotated datasets, which are hard to acquire. This research proposes an intensity-based self-supervised domain adaptation, using unlabeled multi-domain data to reduce reliance on large annotated datasets.

Methods

The study introduces an innovative method using intensity-based self-supervised learning for IVD segmentation in MRI scans. This approach is particularly suited for IVD segmentations due to its ability to effectively capture the subtle intensity variations that are characteristic of spinal structures. The model, a dual-task system, simultaneously segments IVDs and predicts intensity transformations. This intensity-focused method has the advantages of being easy to train and computationally light, making it highly practical in diverse clinical settings. Trained on unlabeled data from multiple domains, the model learns domain-invariant features, adeptly handling intensity variations across different MRI devices and protocols.

Results

Testing on three public datasets showed that this model outperforms baseline models trained on single-domain data. It handles domain shifts and achieves higher accuracy in IVD segmentation.

Conclusions

This study demonstrates the potential of intensity-based self-supervised domain adaptation for IVD segmentation. It suggests new directions for research in enhancing generalizability across datasets with domain shifts, which can be applied to other medical imaging fields.

SUBMITTER: Fiorentino MC

PROVIDER: S-EPMC11365836 | biostudies-literature | 2024 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

An intensity-based self-supervised domain adaptation method for intervertebral disc segmentation in magnetic resonance imaging.

Fiorentino Maria Chiara MC Villani Francesca Pia FP Benito Herce Rafael R González Ballester Miguel Angel MA Mancini Adriano A López-Linares Román Karen K

International journal of computer assisted radiology and surgery 20240708 9

<h4>Background and objective</h4>Accurate IVD segmentation is crucial for diagnosing and treating spinal conditions. Traditional deep learning methods depend on extensive, annotated datasets, which are hard to acquire. This research proposes an intensity-based self-supervised domain adaptation, using unlabeled multi-domain data to reduce reliance on large annotated datasets.<h4>Methods</h4>The study introduces an innovative method using intensity-based self-supervised learning for IVD segmentati ...[more]

PMID: 38976178

Similar Datasets

Project description:IntroductionLarge clinical studies regarding cervical intervertebral disc extrusion (IVDE) in Dachshunds are lacking. This retrospective multicentric study therefore aims to describe the clinical features, magnetic resonance imaging (MRI) findings and outcomes of Dachshunds diagnosed with cervical IVDE.MethodsMedical records of Dachshunds with cervical IVDE were reviewed for signalment, onset of clinical signs, neurological examination, MRI features, treatment and outcome.ResultsEighty Dachshunds were included in the study, mostly ambulatory (55% grade 1 and 33% grade 2) and without nerve root signature (85% of cases) on presentation. Information on coat type was available for 56% of dogs; specifically, 41% were smooth-haired, 9% were long-haired and 6% were wire-haired Dachshunds. There were 29 (36%) neutered female, 27 (34%) male entire, 15 (19%) male neutered and 9 (11%) entire female dogs. The onset of clinical signs was most often >48 h (84%). The most common intervertebral disc space affected was C2-C3 (38%) and foraminal IVDEs were reported in 14% of dogs. A foraminal IVDE was diagnosed in only 25% of dogs presented with nerve root signatures. Most dogs (77.5%) were treated surgically. In this group, a higher body condition score on presentation and a higher mean spinal cord compression ratio calculated on MRI were directly and moderately associated with a longer hospitalization time (r = 0.490 p = 0.005 and r = 0.310 p = 0.012, respectively). The recovery time was longer in dogs with an onset of clinical signs <24 h or 24-48 h compared to those with an onset of clinical signs >48 h (3.1 ± 6.5 days versus 1.6 ± 6.2, p < 0.001) in both medically and surgically treated groups. Data about the outcome was available for 83% of dogs. Eighty percent of the entire population of dogs was considered to have completely returned to normal. There was no association between the therapeutic choice (surgical versus medical management) and the outcome of the dogs included in this study.

Project description:The thalamus is an essential relay station in the cortical-subcortical connections. It is characterized by a complex anatomical architecture composed of numerous small nuclei, which mediate the involvement of the thalamus in a wide range of neurological functions. We present a novel framework for segmenting the thalamic nuclei, which explores the orientation distribution functions (ODFs) from diffusion magnetic resonance images at 3 T. The differentiation of the complex intra-thalamic microstructure is improved by using the spherical harmonic (SH) representation of the ODFs, which provides full angular characterization of the diffusion process in each voxel. The clustering was performed using the k-means algorithm initialized in a data-driven manner. The method was tested on 35 healthy volunteers and our results show a robust, reproducible and accurate segmentation of the thalamus in seven nuclei groups. Six of them closely matched the anatomy and were labeled as anterior, ventral anterior, medio-dorsal, ventral latero-ventral, ventral latero-dorsal and pulvinar, while the seventh cluster included the centro-lateral and the latero-posterior nuclei. Results were evaluated both qualitatively, by comparing the segmented nuclei to the histological atlas of Morel, and quantitatively, by measuring the clusters' extent and the clusters' spatial distribution across subjects and hemispheres. We also showed the robustness of our approach across different sequences and scanners, as well as intra-subject reproducibility of the segmented clusters using additional two scan-rescan datasets. We also observed an overlap between the path of the main long-connection tracts passing through the thalamus and the spatial distribution of the nuclei identified with our clustering algorithm. Our approach, based on SH representations of the ODFs, outperforms the one based on angular differences between the principle diffusion directions, which is considered so far as state-of-the-art method. Our findings show an anatomically reliable segmentation of the main groups of thalamic nuclei that could be of potential use in many clinical applications.

Dataset Information

An intensity-based self-supervised domain adaptation method for intervertebral disc segmentation in magnetic resonance imaging.

Background and objective

Methods

Results

Conclusions

Publications

An intensity-based self-supervised domain adaptation method for intervertebral disc segmentation in magnetic resonance imaging.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets