New Mathematical Computing Techniques for Analysis of Pre-Operation vs Post-Operation Changes in Brainstem White Matter Tracts using Diffusion Tensor Imaging (TÜBİTAK 1001, 2013-2016)
Brainstem lesions are an important and challenging problem in neurosurgery because main motor and sensory pathways of the human body pass through the brainstem region. Brainstem lesions are diagnosed with biopsy, and treated with resection in neurosurgical practice. Brainstem surgery has progressed successfully in the last 10 years due to advances in microsurgery particularly brainstem dissection techniques (Türe et al., 2000). Similarly, within the last decade, radiological advances provided  patients’ pre-operational (Diffusion) Magnetic Resonance Images (D-MRI) and Diffusion Tensor Images (DTI), which highly aid the surgeon in his/her surgical planning (Kovanlıkaya et al., 2011, Chen et., al 2007a). White matter fiber pathways (tracts) surrounding a lesion can be observed, as they can be created as 3-dimensional (3D) curves in digital form via Diffusion Tensor Tractography (DTT) obtained from DTI. As brainstem is an important region for brain tracts, understanding the interaction of its main pathways, corticospinal tract (CST) and medial lemniscus (ML), with lesions is clinically important within the context of surgical planning of brainstem lesions, and their post-operative evaluation.

The reliability and robustness of pre-operative and post-operative DTI and DTT of brainstem lesion regions were studied by Prof. Dr. Türe and his colleagues (Kovanlıkaya et al., 2011), who are part of our project team. High sensitivity and negative predictive values between CST’s visual evaluation from DTT and neurological findings were reported. However, a high number of false positive values was also reported particularly for pre-operative images, which is due to artefacts in DTT that are caused by both the lesion itself and the imaging affected by the lesion (for instance bleeding in the lesion). Nevertheless, it was suggested that new studies should be carried out as information from DTT together with neurological examination of a patient can be used in understanding and evaluating post-operative results, in addition to neurosurgical planning. Related studies in the literature include geometrical and morphological tract information only visually (qualitatively) in analysis, as it is not possible for the experts to determine pre-op and post-op differences manually. Mathematical computer techniques are required for analyzing post-op structural fiber information and its comparison to pre-op fiber structures.

Our aim in this project is quantification of brain white matter fiber structure from Diffusion MR images; design and computer-based implementation of mathematical techniques for quantification of pre-operation versus post-operation structural and orientation changes of fiber tracts. These techniques will provide comparison of structural fiber information to neurological findings as well as quantitatively carrying out pre-op post-op change analysis. In addition, simultaneous visualization of fiber structures and lesion surfaces will aid surgical planning and navigation. The desired techniques in this project will lead to enhancement of patient treatment and follow-up of the treatment.