Automated and Reliable Tractography-Based Thalamic Segmenation for DBS SurgeryKeywords: thalamus, diffusion tensor imaging, deep brain stimulation, anatomy, magnetic resonance imagingInteractive Manuscript
Ask Questions of this Manuscript:
What is the background behind your study?
To address the imprecision of indirect thalamic targeting for the stereotactic treatment of tremor, we recently reported a means of identifying a patient-specific thalamic target using probabilistic tractography-based segmentation. In its original description, the method required manual thalamic and cortical delineations and was time-consuming, impractical for clinical implementation, and subject to inter-rater variability.
What is the purpose of your study?
We now describe automation of this methodology and our initial experience with prospective validation.
Describe your patient group.
Two patients were evaluated.
Describe what you did.
Each subject underwent MRI (prior to DBS surgery), including a 1mm isotropic T1-weighted MRI and a 20-direction DTI sequence. A fully automated thalamic segmentation algorithm was implemented in the UCLA Laboratory of NeuroImaging Pipeline environment, integrating functionality from publicly available brain image analysis suites, including FSL, FreeSurfer, and BrainParser. The system was designed in an extensible manner to enable investigation of and application to targets for other diseases. Reliability and automation were evaluated.
Describe your main findings.
Automation and reliability testing was performed in four subjects and preliminary clinical experience was assessed in two patients. In each subject, the automated algorithm was run three times in each hemisphere, consistently identifying the thalamic region with the highest probability of connectivity with the precentral gyrus (PMC thalamus) and producing a color-coded segmentation of the thalamus that resembled previously described histologic thalamic nuclear architecture with no inter-trial variation. In the two surgical subjects, the therapeutic thalamic target for tremor precisely coincided with PMC thalamus.
Describe the main limitation of this study.
This is a retrospective study.
Describe your main conclusion.
Fully automated thalamic segmentation reliably and consistently produces maps that identify the therapeutic target for tremor.
Describe the importance of your findings and how they can be used by others.
This approach may prove advantageous over indirect targeting, providing patient-specific targets that could improve the precision, efficacy, and efficiency of DBS surgery. Further prospective evaluation and extension to targets for other diseases is warranted.
To address the imprecision of indirect thalamic targeting for the stereotactic treatment of tremor, we recently reported a means of identifying a patient-specific thalamic target using probabilistic tractography-based segmentation. In its original description, the method required manual thalamic and cortical delineations and was time-consuming, impractical for clinical implementation, and subject to inter-rater variability.
We now describe automation of this methodology and our initial experience with prospective validation.
Two patients were evaluated.
Each subject underwent MRI (prior to DBS surgery), including a 1mm isotropic T1-weighted MRI and a 20-direction DTI sequence. A fully automated thalamic segmentation algorithm was implemented in the UCLA Laboratory of NeuroImaging Pipeline environment, integrating functionality from publicly available brain image analysis suites, including FSL, FreeSurfer, and BrainParser. The system was designed in an extensible manner to enable investigation of and application to targets for other diseases. Reliability and automation were evaluated.
Automation and reliability testing was performed in four subjects and preliminary clinical experience was assessed in two patients. In each subject, the automated algorithm was run three times in each hemisphere, consistently identifying the thalamic region with the highest probability of connectivity with the precentral gyrus (PMC thalamus) and producing a color-coded segmentation of the thalamus that resembled previously described histologic thalamic nuclear architecture with no inter-trial variation. In the two surgical subjects, the therapeutic thalamic target for tremor precisely coincided with PMC thalamus.
This is a retrospective study.
Fully automated thalamic segmentation reliably and consistently produces maps that identify the therapeutic target for tremor.
This approach may prove advantageous over indirect targeting, providing patient-specific targets that could improve the precision, efficacy, and efficiency of DBS surgery. Further prospective evaluation and extension to targets for other diseases is warranted.
Project Roles:
