Jonathan D. Olson

Hyperspectral data processing improves PpIX contrast during fluorescence guided surgery of human brain tumors

Fluorescence guided surgery (FGS) using aminolevulinic-acid (ALA) induced protoporphyrin IX (PpIX) provides intraoperative visual contrast between normal and malignant tissue during resection of high grade gliomas. However, maps of the PpIX biodistribution within the surgical field based on either visual perception or the raw fluorescence emissions can be masked by background signals or distorted by variations in tissue optical properties. This study evaluates the impact of algorithmic processing of hyperspectral imaging acquisitions on the sensitivity and contrast of PpIX maps. Measurements in tissue-simulating phantoms showed that (I) spectral fitting enhanced PpIX sensitivity compared with visible or integrated fluorescence, (II) confidence-filtering automatically determined the lower limit of detection based on the strength of the PpIX spectral signature in the collected emission spectrum (0.014–0.041 μg/ml in phantoms), and (III) optical-property corrected PpIX estimates were more highly correlated with independent probe measurements (r = 0.98) than with spectral fitting alone (r = 0.91) or integrated fluorescence (r = 0.82). Application to in vivo case examples from clinical neurosurgeries revealed changes to the localization and contrast of PpIX maps, making concentrations accessible that were not visually apparent. Adoption of these methods has the potential to maintain sensitive and accurate visualization of PpIX contrast over the course of surgery.

5-Aminolevulinic Acid-Induced Fluorescence in Focal Cortical Dysplasia: Report of 3 Cases

BACKGROUND Three patients enrolled in a clinical trial of 5-aminolevulinic-acid (5-ALA)-induced fluorescence-guidance, which has been demonstrated to facilitate intracranial tumor resection, were found on neuropathological examination to have focal cortical dysplasia (FCD). OBJECTIVE To evaluate in this case series visible fluorescence and quantitative levels of protoporphyrin IX (PpIX) during surgery and correlate these findings with preoperative magnetic resonance imaging (MRI) and histopathology. METHODS Patients were administered 5-ALA (20 mg/kg) approximately 3 h prior to surgery and underwent image-guided, microsurgical resection of their MRI- and electrophysiologically identified lesions. Intraoperative visible fluorescence was evaluated using an operating microscope adapted with a commercially available blue light module. Quantitative PpIX levels were assessed using a handheld fiber-optic probe and a wide-field imaging spectrometer. Sites of fluorescence measurements were co-registered with both preoperative MRI and histopathological analysis. RESULTS Three patients with a pathologically confirmed diagnosis of FCD (Types 1b, 2a, and 2b) underwent surgery. All patients demonstrated some degree of visible fluorescence (faint or moderate), and all patients had quantitatively elevated concentrations of PpIX. No evidence of neoplasia was identified on histopathology, and in 1 patient, the highest concentrations of PpIX were found at a tissue site with marked gliosis but no typical histological features of FCD. CONCLUSION FCD has been found to be associated with intraoperative 5-ALA-induced visible fluorescence and quantitatively confirmed elevated concentrations of the fluorophore PpIX in 3 patients. This finding suggests that there may be a role for fluorescence-guidance during surgical intervention for epilepsy-associated FCD.

Image Updating for Brain Shift Compensation During Resection

BACKGROUND In open-cranial neurosurgery, preoperative magnetic resonance (pMR) images are typically coregistered for intraoperative guidance. Their accuracy can be significantly degraded by intraoperative brain deformation, especially when resection is involved. OBJECTIVE To produce model updated MR (uMR) images to compensate for brain shift that occurred during resection, and evaluate the performance of the image-updating process in terms of accuracy and computational efficiency. METHODS In 14 resection cases, intraoperative stereovision image pairs were acquired after dural opening and during resection to generate displacement maps of the surgical field. These data were assimilated by a biomechanical model to create uMR volumes of the evolving surgical field. A tracked stylus provided independent measurements of feature locations to quantify target registration errors (TREs) in the original coregistered pMR and uMR as surgery progressed. RESULTS Updated MR TREs were 1.66 ± 0.27 and 1.92 ± 0.49 mm in the 14 cases after dural opening and after partial resection, respectively, compared to 8.48 ± 3.74 and 8.77 ± 4.61 mm for pMR, respectively. The overall computational time for generating uMRs after partial resection was less than 10 min. CONCLUSION We have developed an image-updating system to compensate for brain deformation during resection using a computational model with data assimilation of displacements measured with intraoperative stereovision imaging that maintains TREs less than 2 mm on average.

Stereovision Co-Registration in Image-Guided Spinal Surgery: Accuracy Assessment Using Explanted Porcine Spines

BACKGROUND Current methods of spine registration for image guidance have a variety of limitations related to accuracy, efficiency, and cost. OBJECTIVE To define the accuracy of stereovision-mediated co-registration of a spinal surgical field. METHODS A total of 10 explanted porcine spines were used. Dorsal soft tissue was removed to a variable degree. Bone screw fiducials were placed in each spine and high-resolution computed tomography (CT) scanning performed. Stereoscopic images were then obtained using a tracked, calibrated stereoscopic camera system; images were processed, reconstructed, and segmented in a semi-automated manner. A multistart registration of the reconstructed spinal surface with preoperative CT was performed. Target registration error (TRE) in the region of the laminae and facets was then determined, using bone screw fiducials not included in the original registration process. Each spine also underwent multilevel laminectomy, and TRE was then recalculated for varying amounts of bone removal. RESULTS The mean TRE of stereovision registration was 2.19 ± 0.69 mm when all soft tissue was removed and 2.49 ± 0.74 mm when limited soft tissue removal was performed. Accuracy of the registration process was not adversely affected by laminectomy. CONCLUSION Stereovision offers a promising means of registering an open, dorsal spinal surgical field. In this study, overall mean accuracy of the registration was 2.21 mm, even when bony anatomy was partially obscured by soft tissue or when partial midline laminectomy had been performed.

Automatic intraoperative fiducial-less patient registration using cortical surface

In image-guided neurosurgery, patient registration is typically performed in the operating room (OR) at the beginning of the procedure to establish the patient-to-image transformation. The accuracy and efficiency of patient registration are crucial as they are associated with surgical outcome, workflow, and healthcare costs. In this paper, we present an automatic fiducial-less patient registration (FLR) by directly registering cortical surface acquired from intraoperative stereovision (iSV) with preoperative MR (pMR) images without incorporating any prior information, and illustrate the method using one patient example. T1-weighted MR images were acquired prior to surgery and the brain was segmented. After dural opening, an image pair of the exposed cortical surface was acquired using an intraoperative stereovision (iSV) system, and a three-dimensional (3D) texture-encoded profile of the cortical surface was reconstructed. The 3D surface was registered with pMR using a multi-start binary registration method to determine the location and orientation of the iSV patch with respect to the segmented brain. A final transformation was calculated to establish the patient-to-MR relationship. The total computational time was ~30 min, and can be significantly improved through code optimization, parallel computing, and/or graphical processing unit (GPU) acceleration. The results show that the iSV texture map aligned well with pMR using the FLR transformation, while misalignment was evident with fiducial-based registration (FBR). The difference between FLR and FBR was calculated at the center of craniotomy and the resulting distance was 4.34 mm. The results presented in this paper suggest potential for clinical application in the future.

Ongoing advances in quantitative PpIX fluorescence guided intracranial tumor resection(Conference Presentation)

Aminolevulinc-acid induced protoporphyrin IX (ALA-PpIX) is being investigated as a biomarker to guide neurosurgical resection of brain tumors. ALA-PpIX fluorescence can be observed visually in the surgical field; however, raw fluorescence emissions can be distorted by factors other than the fluorophore concentration. Specifically, fluorescence emissions are mixed with autofluorescence and attenuated by background absorption and scattering properties of the tissue. Recent work at Dartmouth has developed advanced fluorescence detection approaches that return quantitative assessments of PpIX concentration, which are independent of background optical properties. The quantitative fluorescence imaging (qFI) approach has increased sensitivity to residual disease within the resection cavity at the end of surgery that was not visible to the naked eye through the operating microscope. This presentation outlines clinical observations made during an ongoing investigation of ALA-PpIX based guidance of tumor resection. PpIX fluorescence measurements made in a wide-field hyperspectral imaging approach are co-registered with point-assessment using a fiber optic probe. Data show variations in the measured PpIX accumulation among different clinical tumor grades (i.e. high grade glioma, low grade glioma), types (i.e. primary tumors. metastases) and normal structures of interest (e.g. normal cortex, hippocampus). These results highlight the contrast enhancement and underscore the potential clinical benefit offered from quantitative measurements of PpIX concentration during resection of intracranial tumors.

Patient Registration via Topologically Encoded Depth Projection Images in Spine Surgery

Accurate and efficient patient registration is essential for surgical image-guidance. Here, we present a registration pipeline to establish spatial correspondence between tracked intraoperative stereovision (iSV) and preoperative computed tomography (pCT) for spine surgery. First, depth projection images encoding the common vertebral dorsal surface “height” were generated from pCT and iSV. For pCT, vertebral pose was adjusted when necessary based on anatomic landmarks. For iSV, multiple reconstructed surfaces were combined to generate a unified projection image with accounting of overlapped regions to maximize the sampling of the surgical scene. Rigid registration between the resulting projection images produced an initial alignment for refined registration using an improved iterative closest point algorithm. The technique was applied to four explanted porcine spines in a total of eight poses. Registration accuracy was assessed using bone-implanted mini screws. The average fiducial registration error and target registration error (TRE) for ground-truth probe registration was \(0.50\,{\pm }\,0.08\) and \(0.63\,{\pm }\,0.08\), respectively. The accuracy for iSV registration was \(1.77\,{\pm }\,0.31\,\text {mm}\) in TRE and was \(2.01\,{\pm }\,0.44\,\text {mm}\) for surface reconstruction. The entire registration completed within 2 min. These results suggest potential for application of the method in human patients.