A probabilistic atlas of the human thalamic nuclei combining ex vivo MRI and histology (1806.08634v1)

Abstract: The human thalamus is a brain structure that comprises numerous, highly specific nuclei. Since these nuclei are known to have different functions and to be connected to different areas of the cerebral cortex, it is of great interest for the neuroimaging community to study their volume, shape and connectivity in vivo with MRI. In this study, we present a probabilistic atlas of the thalamic nuclei built using ex vivo brain MRI scans and histological data, as well as the application of the atlas to in vivo MRI segmentation. The atlas was built using manual delineation of 26 thalamic nuclei on the serial histology of 12 whole thalami from six autopsy samples, combined with manual segmentations of the whole thalamus and surrounding structures (caudate, putamen, hippocampus, etc.) made on in vivo brain MR data from 39 subjects. The 3D structure of the histological data and corresponding manual segmentations was recovered using the ex vivo MRI as reference frame, and stacks of blockface photographs acquired during the sectioning as intermediate target. The atlas, which was encoded as an adaptive tetrahedral mesh, shows a good agreement with with previous histological studies of the thalamus in terms of volumes of representative nuclei. When applied to segmentation of in vivo scans using Bayesian inference, the atlas shows excellent test-retest reliability, robustness to changes in input MRI contrast, and ability to detect differential thalamic effects in subjects with Alzheimer's disease. The probabilistic atlas and companion segmentation tool are publicly available as part of the neuroimaging package FreeSurfer.

• The paper introduces a probabilistic atlas combining ex vivo MRI and histology to improve segmentation of the human thalamus.
• It employs innovative blockface photograph registration with Bayesian inference to ensure accurate 3D reconstruction of 26 thalamic nuclei.
• Results show high test-retest reliability and effective differentiation of Alzheimer’s pathology through detailed volumetric analysis.

A Probabilistic Atlas of the Human Thalamic Nuclei

The paper presents a comprehensive paper on the development and application of a probabilistic atlas of the human thalamic nuclei. This atlas is constructed from ex vivo MRI and histological data and represents an important advancement for neuroimaging studies interested in in-depth analysis of the thalamic structure. The focus is on providing a tool that can improve the segmentation of the thalamus in in vivo MRI scans, which is vital for exploring its connectivity, volume, and shape.

Methodology and Atlas Construction

The atlas was constructed using manual delineations of 26 thalamic nuclei derived from the histology of 12 thalami from six autopsy samples, combined with manual segmentations of the whole thalamus and surrounding structures from in vivo MRI data of 39 subjects. A key methodological highlight is utilizing stacks of blockface photographs during sectioning as intermediate targets, assisting in the registration and 3D reconstruction of histology guided by MRI data. This innovative approach facilitated overcoming challenges related to the alignment and registration of histological sections.

The probabilistic atlas is encoded as an adaptive tetrahedral mesh, and the paper emphasizes its consistency with previous histological studies. The atlasing approach incorporates Bayesian inference, which notably enhances the segmentation process by providing robust test-retest reliability and adaptability to various MRI contrasts. These characteristics are vital for applying the atlas to MRI scans of diverse qualities and specifications.

Validation and Results

The paper reports a methodical validation process encompassing multiple aspects:

1. Volumetric Comparison: The atlas was compared to Krauth’s atlas, demonstrating general volumetric agreement with slight discrepancies which did not significantly impact overall perceptions of anatomical accuracy.
2. Test-Retest Reliability: High intraclass correlation coefficients for the whole thalamus and individual nuclei highlight the segmentation method’s reliability.
3. Robustness Across MRI Contrasts: Tested across several MRI contrasts, the atlas proved robust, ensuring utility across a range of scanning protocols.
4. Alzheimer's Disease Study: Significant differences in thalamic volumes were found between Alzheimer’s patients and controls. An LDA-based classifier using thalamic sub-region volumes achieved a higher AUC compared to whole thalamus volume alone, offering better discrimination between groups.

Implications and Future Directions

The implications of this paper are multifaceted. Practically, this atlas is available as part of the FreeSurfer neuroimaging package, providing widespread access to researchers. Theoretically, it enhances our understanding of thalamic structure and function, enabling detailed morphometric and connectivity studies.

Future work could focus on integrating diffusion MRI data into the segmentation process to further improve accuracy, given the high connectivity signature of individual nuclei across the cortical landscape. Additionally, the detailed analysis of thalamic structures for neurodegenerative diseases beyond Alzheimer's can lead to new insights and better therapeutic targeting.

Overall, this work represents a significant tool for the neuroimaging community, addressing longstanding challenges in thalamic segmentation and paving the way for more refined and diagnostically useful brain imaging protocols.