Mutual information in coupled multi-shape model for medical image segmentation.

This paper presents extensions which improve the performance of the shape-based deformable active contour model presented earlier in [IEEE Conf. Comput. Vision Pattern Recog. 1 (2001) 463] for medical image segmentation. In contrast to that previous work, the segmentation framework that we present in this paper allows multiple shapes to be segmented simultaneously in a seamless fashion. To achieve this, multiple signed distance functions are employed as the implicit representations of the multiple shape classes within the image. A parametric model for this new representation is derived by applying principal component analysis to the collection of these multiple signed distance functions. By deriving a parametric model in this manner, we obtain a coupling between the multiple shapes within the image and hence effectively capture the co-variations among the different shapes. The parameters of the multi-shape model are then calculated to minimize a single mutual information-based cost criterion for image segmentation. The use of a single cost criterion further enhances the coupling between the multiple shapes as the deformation of any given shape depends, at all times, upon every other shape, regardless of their proximity. We found that this resulting algorithm is able to effectively utilize the co-dependencies among the different shapes to aid in the segmentation process. It is able to capture a wide range of shape variability despite being a parametric shape-model. And finally, the algorithm is robust to large amounts of additive noise. We demonstrate the utility of this segmentation framework by applying it to a medical application: the segmentation of the prostate gland, the rectum, and the internal obturator muscles for MR-guided prostate brachytherapy.

Coupled multi-shape model and mutual information for medical image segmentation.

This paper presents extensions which improve the performance of the shape-based deformable active contour model presented earlier in [9]. In contrast to that work, the segmentation framework that we present in this paper allows multiple shapes to be segmented simultaneously in a seamless fashion. To achieve this, multiple signed distance functions are employed as the implicit representations of the multiple shape classes within the image. A parametric model for this new representation is derived by applying principal component analysis to the collection of these multiple signed distance functions. By deriving a parametric model in this manner, we obtain a coupling between the multiple shapes within the image and hence effectively capture the co-variations among the different shapes. The parameters of the multi-shape model are then calculated to minimize a single mutual information-based cost functional for image segmentation. The use of a single cost criterion further enhances the coupling between the multiple shapes as the deformation of any given shape depends, at all times, upon every other shape, regardless of their proximity. We demonstrate the utility of this algorithm to the segmentation of the prostate gland, the rectum, and the internal obturator muscles for MR-guided prostate brachytherapy.

Transcranial magnetic stimulation coregistered with MRI: a comparison of a guided versus blind stimulation technique and its effect on evoked compound muscle action potentials.

INTRODUCTION AND METHODS: Compound muscle action potentials (CMAPs) elicited by transcranial magnetic stimulation (TMS) are characterized by enormous variability, even when attempts are made to stimulate the same scalp location. This report describes the results of a comparison of the spatial errors in coil placement and resulting CMAP characteristics using a guided and blind TMS stimulation technique. The former uses a coregistration system, which displays the intersection of the peak TMS induced electric field with the cortical surface. The latter consists of the conventional placement of the TMS coil on the optimal scalp position for activation of the first dorsal interossei (FDI) muscle. RESULTS: Guided stimulation resulted in significantly improved spatial precision for exciting the corticospinal projection to the FDI compared to blind stimulation. This improved precision of coil placement was associated with a significantly increased probability of eliciting FDI responses. Although these responses tended to have larger amplitudes and areas, the coefficient of variation between guided and blind stimulation induced CMAPs did not significantly differ. CONCLUSION: The results of this study demonstrate that guided stimulation improves the ability to precisely revisit previously stimulated cortical loci as well as increasing the probability of eliciting TMS induced CMAPs. Response variability, however, is due to factors other than coil placement.

Excision of cortical dysplasia in the language area with use of a surgical navigator: a case report.

PURPOSE: We have developed an intraoperative optical tracking-based navigational system that allows localization in the operative space. Using three-dimensional reconstruction, this system has provided precise spatial information for intraoperative cortical mapping in patients with intractable epilepsy in whom the lesion lies close to eloquent cortex. METHODS: A 23-year-old man with intractable complex partial seizures (CPS) presented to our institution. Proton-density magnetic resonance imaging (MRI) showed a 3-cm lesion which lay 2 cm beneath the left frontal operculum. A three-dimensional model of the patient was reconstructed using MR modalities. Intraoperatively, subdural grid and strips were placed over the lesion and their electrodes were registered to the three-dimensional model, which was displayed on a monitor. The navigational system was used to localize each electrode on the three-dimensional model. By the second operation, the sites of seizure activity were established and recorded on the three-dimensional model. A bipolar stimulator was also used to determine the speech area. RESULTS: The lesion, which proved to be cortical dysplasia, was removed completely and the cortical speech area was avoided. During the postoperative period, the patient had no neurological symptoms and no seizure activity. CONCLUSIONS: The localization of a lesion and its correlation with epileptogenic foci is important in optimizing treatment in patients with cortical dysplasia. Our navigational system provided accurate localization of the lesion and correlation with the epileptogenic focus and related eloquent cortex. We believe that the safe removal of the lesion was facilitated by this system.

Volumetric object modeling for surgical simulation.

Surgical simulation has many applications in medical education, surgical training, surgical planning and intra-operative assistance. However, extending current surface-based computer graphics methods to model phenomena such as the deformation, cutting, tearing or repairing of soft tissues poses significant challenges for real-time interactions. This paper discusses the use of volumetric methods for modeling complex anatomy and tissue interactions. New techniques are introduced that use volumetric methods for modeling soft-tissue deformation and tissue cutting at interactive rates. An initial prototype for simulating arthroscopic knee surgery is described which uses volumetric models of the knee derived from 3-D magnetic resonance imaging, visual feedback via real-time volume and polygon rendering, and haptic feedback provided by a force-feedback device.

Three-dimensional image reconstruction for low-grade glioma surgery.

Three-dimensional image reconstruction for preoperative surgical planning and intraoperative navigation for the resection of low-grade gliomas was performed in 20 patients. Thirteen of these surgeries were performed while the patient received a local anesthetic to allow for cortical mapping. Ninety percent of the patients were functionally intact postoperatively. The authors propose that the combination of the three-dimensional image reconstruction and surgical navigation, in conjunction with intraoperative cortical mapping, provides an additional means for surgeons to improve the safety and precision of the procedures.

Three-dimensional reconstruction and surgical navigation in pediatric epilepsy surgery.

We have used MRI-based three-dimensional (3D) reconstruction and a real-time, frameless, stereotactic navigation device to facilitate the removal of seizure foci in children suffering from intractable epilepsy. Using this system, the location of subdural grid and strip electrodes is recorded on the 3D model to facilitate focus localization and resection. Ten operations were performed, including 2 girls and 8 boys ranging in age from 3 to 17, during which 3D reconstruction and surgical instrument tracking navigation was used. In all the cases, the patients tolerated the procedure well and showed no postoperative neurological deficits. We believe this to be a valuable tool for a complete and safe resection of seizure foci, thereby reducing the incidence of postoperative neurological deficits and significantly improving the overall quality of life of the patients.