Connectomics Reconstruction Associated with Temporal Lobe Lesions and Their Surgical Resection

Academic Background

The human brain’s structural organization is increasingly conceptualized and analyzed from a network perspective, greatly enhancing understanding of health and disease. In recent years, thanks to advances in neuroimaging technology and complex data analysis, these methods have been widely applied. Especially with the development of diffusion magnetic resonance imaging (MRI), researchers have been able to approximate the mapping of structural brain connectivity (connectome) in vivo and further reveal the organizational principles of brain functional networks by systematically characterizing brain connectivity.

However, despite some progress in this field, our understanding of how focal lesions regulate brain networks remains limited. Anterior temporal lobectomy is the most effective treatment for drug-resistant temporal lobe epilepsy (TLE), making this syndrome an ideal model for studying the impact of lesions on brain networks.

Source of the Paper

This paper was completed by a research team led by Sara Larivière, consisting of members from several well-known research institutions including the McConnell Brain Imaging Centre, Brigham and Women’s Hospital at Harvard University, the Department of Data Science at Inha University in South Korea, and Jinling Hospital, Nanjing University School of Medicine. The paper was published in the journal “Brain” on May 22, 2024, by Oxford University Press.

Research Methods

Research Process Overview

The research included the following main steps: 1. Subject Identification: - Included 37 patients with drug-resistant TLE who underwent anterior temporal lobectomy and 31 age- and gender-matched healthy controls. - All patients underwent high-resolution 3T MRI imaging before and after surgery, including T1-weighted imaging (T1W) and diffusion MRI imaging (DWI).

1. Data Processing and Analysis:

   • MRI data were first preprocessed, including image de-skew, reorientation, intensity normalization, and skull stripping.
   • Cortical and subcortical structures were segmented using FreeSurfer and FSL FIRST.
   • Fiber tractography was performed on diffusion MRI data using MRtrix3 software to generate individual structural connectomes.

2. Surgery Cavity Mapping:

   • The surgical cavity was automatically segmented by registering preoperative and postoperative T1-weighted images to the MNI152 standard template and calculating the difference map.

3. Generation of Structural Connectomes:

   • Anatomically constrained tractography was used to generate whole-brain structural connectomes for each participant. With spherical deconvolution and intensity normalization, 40 million streamlines were generated and mapped to 400 similarly sized cortical regions and 14 subcortical and hippocampal regions, creating individual-specific structural connectivity matrices.

4. Connectome Gradient Estimation:

   • Brainspace toolbox generated gradients on cortical and subcortical structural connectomes, revealing structural connectivity characteristics through nonlinear dimensionality reduction and aligning and comparing gradients between individuals and groups.

5. Quantification of Pre- and Postoperative Gradient Changes:

   • Linear mixed-effects models were used to assess connectome changes over time, identifying significant multidimensional gradient change regions pre- and postoperatively.

Research Results

1. Preoperative Connectome Gradient Changes in TLE Patients:

   • Compared to the healthy control group, preoperative TLE patients exhibited significant gradient changes in bilateral temporoparietal regions and orbitofrontal cortex, reflecting increased isolation between the ipsilateral anterior temporal lobe and other brain regions.
   • By quantifying positional changes in gradient space, significant deformation was found in the ipsilateral medial temporal lobe region of TLE patients, indicating increased isolation of these regions in the gradient space.

2. Gradient Deformation Tracking Induced by Surgery:

   • Comparing pre- and postoperative structural connectome changes, significant multidimensional gradient changes were found in the posterior temporal region adjacent to the resection site and in the contralateral temporoparietal cortex. These regions tended towards more integrated positions postoperatively, especially on gradient axis 1.

3. Association with Clinical Variables:

   • Partial Least Squares (PLS) analysis revealed correlations between surgery-induced 4D gradient deformation patterns and clinical variables such as hippocampal atrophy, seizure frequency, and secondary generalized convulsions.
   • Increased connectivity integration in the pre- and post-central cortical regions was associated with greater ipsilateral hippocampal atrophy and lower seizure frequency.

4. Cortical Morphological and Microstructural Changes:

   • In gray and white matter assessments, the contralateral temporoparietal and frontoparietal regions showed reduced cortical thinning postoperatively, while the thinning increased around the resection site. White matter disturbances occurred mainly in the contralateral temporoparietal region, consistent with the gradient changes observed after surgery.

Research Conclusions

This study used gradient mapping of structural connectome data from pre- and postoperative drug-resistant TLE patients to reveal the profound impact of focal lesions and their surgical resection on brain connectivity networks. Preoperatively, TLE patients showed increased isolation between the ipsilateral anterior temporal lobe and other brain regions, while postoperatively, not only this region exhibited significant reorganization, but the contralateral temporoparietal cortex also exhibited connectivity changes associated with increased integration. Compared to traditional graph theoretical analysis, gradient mapping offers a low-dimensional perspective that captures brain structural connectivity reorganization in a data-driven manner, thus better elucidating the organizational principles of the nervous system. These results not only deepen our understanding of brain structure but also offer new possibilities for individualized patient care and treatment.

Highlights and Significance of the Study

• Novel Approach and Perspective: Using structural connectome gradient mapping technology provides a low-dimensional and continuous coordinate system to quantify the impact of focal lesions and their surgical resection on the brain’s structural network as a whole.
• Clinical Relevance Analysis Across Individuals: Partial Least Squares (PLS) analysis revealed associations between connectome reorganization patterns and clinical variables, providing new insights for clinical diagnosis and treatment.
• Broad Application Prospects: This study provides a reference for other neurological disease research, particularly in exploring the effects of focal lesions on the overall functional network of the brain, which is of great significance.

This study explores how the brain undergoes large-scale structural reorganization in response to focal lesions and surgical intervention, providing a new analytical framework for subsequent research and offering scientific evidence for personalized medical practice.