A Bayesian approach to fiber orientation estimation guided by volumetric tract segmentation

The mean kurtosis (MK) is more sensitive diagnostic biomarker than fractional anisotropy (FA) for Parkinson's disease: A diagnostic performance study and meta-analysis

BACKGROUND

The mean kurtosis (MK) and fractional anisotropy (FA) in patients of Parkinson's disease (PD) are usually measured by diffusion kurtosis imaging (DKI) and diffusion tensor imaging (DTI), separately.

METHODS

In this study we perform a meta-analysis to discuss which noninvasive biomarker is more advantageous for PD, MK, or FA. Databases including Medline via PubMed, the Cochrane Central Register of Controlled Trials, Embase via OVID and China National Knowledge Infrastructure. Databases are searched up to December 31st, 2019. Four brain regions are identified for analysis based on data extracted from articles.

RESULTS

The articles contain 5 trials with 274 total PD patients and 189 healthy controls (HCs). The results show not only significantly higher MK values of putamen, caudate, globus pallidus in PD compared to that of HCs (weighted mean difference [WMD] = 0.06, 95% CI = 0.02-0.09, P = .002, WMD = 0.03, 95% CI = 0.01-0.067, P = .01, WMD = 0.18, 95% CI = 0.11-0.24, P < .00001), but also a significantly higher FA in caudate of PD compared to HCs (WMD = 0.02, 95% CI = 0.00-0.03, P = .006).

CONCLUSION

This indicates that the sharp difference detected between PD patients and HCs can be detected by DKI and DTI. By further discussing results, we found that MK could be more sensitive diagnostic biomarker than FA toward PD diagnosis.

Volumetric white matter tract segmentation with nested self-supervised learning using sequential pretext tasks

White matter (WM) tract segmentation based on diffusion magnetic resonance imaging (dMRI) provides an important tool for the analysis of brain development, function, and disease. Deep learning based methods of WM tract segmentation have been proposed, which greatly improve the accuracy of the segmentation. However, the training of the deep networks usually requires a large number of manual delineations of WM tracts, which can be especially difficult to obtain and unavailable in many scenarios. Therefore, in this work, we explore how to perform deep learning based WM tract segmentation when annotated training data is scarce. To this end, we seek to exploit the abundant unannotated dMRI data in the self-supervised learning framework. From the unannotated data, knowledge about image context can be learned with pretext tasks that do not require manual annotations. Specifically, a deep network can be pretrained for the pretext task, and the knowledge learned from the pretext task is then transferred to the subsequent WM tract segmentation task with only a small number of annotated scans via fine-tuning. We explore two designs of pretext tasks that are related to WM tracts. The first pretext task predicts the density map of fiber streamlines, which are representations of generic WM pathways, and the training data can be obtained automatically with tractography. The second pretext task learns to mimic the results of registration-based WM tract segmentation, which, although inaccurate, is more relevant to WM tract segmentation and provides a good target for learning context knowledge. Then, we combine the two pretext tasks and develop a nested self-supervised learning strategy. In the nested self-supervised learning strategy, the first pretext task provides initial knowledge for the second pretext task, and the knowledge learned from the second pretext task with the initial knowledge is transferred to the target WM tract segmentation task via fine-tuning. To evaluate the proposed method, experiments were performed on brain dMRI scans from the Human Connectome Project dataset with various experimental settings. The results show that the proposed method improves the performance of WM tract segmentation when tract annotations are scarce.