Standard planning for dental implants consists of a prosthesis simulation on diagnostic casts and radiographic examination of anatomical structures. The clinician visually locates the planned trajectory on the surgical site in the patient's mouth without direct correlation between the radiographs and the anatomy. We have developed a computer assisted technique to define the optimal position of the bone implant using computed tomography to accurately place the implant in the planned position using a guide drilled into a resin splint. J Image Guid Surg 1:53-58 (1995). © 1995 Wiley-Liss, Inc.
The computer-assisted spine surgery system presented in this paper follows the basic ideas which have been developed for computer-assisted medical interventions (CAMI) in our lab since 1985. There are three steps to insert a linear tool inside vertebral pedicles. First, the surgeon defines an optimal trajectory on pre-operative computed tomography. Second, this trajectory is reported in the operating room coordinate system using an intra-operative sensor and a registration algorithm. Third, a guiding system helps the surgeon follow the selected trajectory. In this paper, we present an implementation of this method that uses only a 3-dimensional optical localizer. Results on cadaver specimens and on the first seven patients are presented. J Image Guid Surg 1:65-73 (1995). © 1995 Wiley-Liss, Inc.
A variety of medical robots for stereotactic neurosurgery has been developed in recent years. Almost of all these robots use computed tomography (CT) to scan the brain of the patient before and during surgery. Currently, we are developing a needle insertion manipulator for magnetic resonance imaging (MRI)-guided neurosurgery. MRI techniques, including MRI angiography and functional MRI, are attractive for the development of interventional MRI therapies and operations. If a robot were available, these therapies would be minimally invasive, with more accurate guidance than is possible with current CT-guided systems. Actuation of a robot in an MRI environment is difficult because of the presence of strong magnetic fields. Therefore, the robot must be constructed of nonmagnetic materials. The system frame was manufactured using polyethylene terephthalate (PET) and was actuated using ultrasonic motors. Accuracy-evaluation procedures and phantom tests have been performed. The total accuracy of the system was approximately 3.0 mm. No artifacts caused by the manipulator were observed in the images.
The segmentation of MRI scans of patients with white matter lesions (WML) is difficult because the MRI characteristics of WML are similar to those of gray matter. Intensity-based statistical classification techniques misclassify some WML as gray matter and some gray matter as WML. We developed a fast elastic matching algorithm that warps a reference data set containing information about the location of the gray matter into the approximate shape of the patient's brain. The region of white matter was segmented after segmenting the cortex and deep gray matter structures. The cortex was identified by using a three-dimensional, region-growing algorithm that was constrained by anatomical, intensity gradient, and tissue class parameters. White matter and WML were then segmented without interference from gray matter by using a two-class minimum-distance classifier. Analysis of double-echo spin-echo MRI scans of 16 patients with clinically determined multiple sclerosis (MS) was carried out. The segmentation of the cortex and deep gray matter structures provided anatomical context. This was found to improve the segmentation of MS lesions by allowing correct classification of the white matter region despite the overlapping tissue class distributions of gray matter and MS lesion.
Starting in 1952, Prof. Gavril Ilizarov, a Russian orthopedic surgeon, devised a complex method of long bone deformity correction using an external bone fixation system consisting of rings and wires. We have developed a computer-assisted surgery planning system to aid in correction using the Ilizarov method. The hardware for the system is composed of an ultrasonic digitizer as input device, an HI plotter as output device, a 386/486 personal computer with super VGA monitor, and a mouse interfacing between user and computer. The software includes two separate programs. The main program is a menu-driven two-dimensional graphics program, which divides the planning procedure into several major stages, including preoperative planning and postoperative verification. The second program uses the information generated from the first program to create an image of the ideal three-dimensional frame construction to give a general impression of the appearance of the constructed frame.
Because of the high level of accuracy needed in neurosurgery, many computer-assisted surgery (CAS) and augmented reality techniques have been developed in this field. A common issue with all of these techniques is registration between preoperative three-dimensional images (computed tomography and magnetic resonance imaging) and the patient in the operating room. We present, in the first part of this paper, a survey of the latest CAS technologies, using fully automatic registration without fiducial landmarks. All of the registration algorithms described are based on minimization of a cost function. We then describe our approach. Our cost function is simply the mean square error (MSE), minimized by the iterative closest point algorithm (ICP). Because the weak point of the ICP algorithm is the closest point computational cost, we precalculate it by a "closest point map," inspired from classical distance map. We finally perturb the found solution to eliminate local minima close to the global minimum. This paper summarizes the various methods presented. We study the shape of the different cost functions and show that there is no need for a complex cost function. MSE has sufficiently good convergence properties to reach a position very close to the global minimum. We also demonstrate the influence of a final perturbation of the found solution to improve registration. Finally, we test the registration on different regions of the patient's head.
Injuries in trauma affect anatomical structures, indirectly affecting physiological systems through mechanical behavior and physical proximity. This paper describes the theory for and preliminary results from our approach to couple a three-dimensional (3-D) anatomical model of the chest with a physiological model of respiratory mechanics. In particular, we investigated behavior in quiet, normal breathing and in an open, sucking chest wound. We envision that our integrated simulation of respiratory anatomy and respiratory mechanics could assist students in visualizing and predicting relationships between structural-anatomical and functional-physiological changes in an interactive, 3-D environment.
This paper describes a simulated surgical setup based on modern, frameless stereotactic techniques that enable surgeons to visualize the field of view of the surgical microscope, overlaid with the segmented volumetric medical images, of a localized area of the patient's head. Using this “true three-dimensional” navigation system, the surgeon visualizes the surgical site while exploring the inner layers of the patient's anatomy via the surgical microscope. It also allows the surgeon to “fly through,” and around, the site of the surgery to visualize several alternatives and qualitatively choose what he or she believes is the best surgical approach. Moving surgical devices are tracked with stereo vision cameras, allowing determination of their spatial relationship to the target lesion.
An essential component in the execution of image-guided surgery is a hand-held probe whose spatial position is tracked during the procedure and displayed on a three-dimensional operative workstation. This paper describes an experiment performed in order to compare the accuracy of a mechanically linked pointing device (FARO surgical arm) and an optical position tracker (OPTOTRAK) against a "gold standard."
Misalignment of the extensor apparatus is an essential factor in impairment of the patellar-femoral joint. This may be partly or entirely responsible for patellar dislocation or lateral patellar-femoral arthrosis. One surgical method to correct the pathology is medial transposition of the patellar ligament on the tibial tuberosity (anteriorly or posteriorly, distally or ventrally). These interventions correct misalignment of the extensor apparatus relative to statistical norms. We propose a mathematical method based on the use of computed tomography (CT) images to determine the ideal tibial insertion for the patellar tendon. This method is based on biomechanical modeling and the use of equipressure criteria. It is the first step in allowing the use of mathematics to model correctly tibial insertion of the patellar ligament, an entirely new development. This is important because it will allow surgeons greater accuracy in distal correction of extensor apparatus misalignments.
We describe the implementation of a robotic arm connected to a neurosurgical operative microscope. A force feedback sensor drives the motors of the arm in response to the positioning of the microscope by the surgeon. Computer graphic techniques allow tracking of the current position of the microscope within the volumetric reconstruction of the brain. The integration of the prototype into the neurosurgical operating room is currently being evaluated. Preliminary comments on this experimental phase are offered.
This paper describes the first prostatic biopsy on a human patient using a robotic and telerobotic system. This system was designed at the Politecnico di Milano, and the biopsy was performed on April 7, 1995, in the Hospital Policlinico in Milan, Italy.
With increased use of magnetic resonance imaging (MRI), diagnosis of cavernous malformations (CMs) has become straightforward. Surgical excision is the treatment of choice for these lesions. These malformations, though, are often small and can be difficult to localize during surgery. In these cases, stereotactic resection with a frame-based system is recommended to aid in localization of the malformation. However, use of these frame-based systems can be time consuming for the surgeon and onerous for the patient. With the advent of frameless stereotactic systems, these problems can be circumvented. Therefore, stereotactic resection of 17 CMs was performed for 15 patients over the course of 2 years at our institution during an investigative trial of a frameless stereotactic device. Eight patients presented with seizures, five patients with hemorrhage, and two patients with progressive headaches. Twelve of fifteen patients had normal neurological examination results on presentation, whereas three patients had deficits resulting from intracranial hemorrhages. All patients underwent diagnostic MRI preoperatively. Fourteen lesions were found to be cortical and subcortical; the other three lesions were in the basal ganglia, lateral ventricle, and pons. Following resection, 11 of 15 patients improved. Two patients developed postoperative deficits shortly after resection. One patient with a preoperative neurological deficit remained unchanged, and one patient had a recurrence of a deficit several months following resection. Image-guided stereotactic resection provides for easy localization of small malformations without requiring the use of a stereotactic frame or retractor and is well suited for resection of cavernous malformations.
We are seeing the emergence of medical applications for virtual reality (VR). These include telepresence surgery, three-dimensional (3-D) visualization of anatomy for medical education, VR surgical simulators, and virtual prototyping of surgical equipment and operating rooms. Today, approximately 90% of the knowledge a physician requires can be obtained through electronic means, such as diagnostic sensors and imaging modalities, directly seeing the patient with a video camera for medical consultation, or using electronic medical records. In addition, with telepresence, a therapy can be effected electronically, regardless of the physical location of the patient. Therefore, it makes sense to send the electronic information or manipulation, rather than sending the patient or blood samples, to obtain tests or to produce a cure. In that these applications are mediated through the computer interface, they are the embodiment of VR as the major force for change in the field of medicine. The Green Telepresence Surgery System consists of two components, the surgical workstation and the remote worksite. At the remote site are a 3-D camera system and responsive manipulators with sensory input. At the workstation are a 3-D monitor and dexterous handles with force feedback. The next generation in medical education can learn anatomy from a new perspective by "flying" inside and around the organs, using sophisticated computer systems and 3-D visualization. The VR surgical simulator is a stylized recreation of the human abdomen with several essential organs. Using this, students and surgeons can practice surgical procedures with virtual scalpels and clamps. To support these advanced technologies, the operating room and hospital of the future will first be designed and tested in virtual reality, allowing multiple iterations of equipment and surgical rooms before they are actually built. Insofar as all these technologies are based on digital information, they are the building blocks for the digital physician of the 21st century.
The aim of conformal radiotherapy is to deliver precisely a specific dose of radiation to a planning target volume, concurrently radiating as little healthy tissue and organs as possible. This can be accomplished only with the accurate positioning of the patient with respect to the radiotherapy system. In this paper, we describe a system to achieve a higher overall accuracy in the delivery of a prostatic radiation boost for treatment of carcinoma of the prostate. The system is based on the use of ultrasound images for measuring the actual position of the patient's prostate just before the radiation. Since these images are registered with pretreatment computed tomography or magnetic resonance imaging, the position and orientation of the planning target volume are computed with respect to the radiotherapy system and can be corrected as needed. This system is under clinical evaluation.
The introduction of spiral computed tomography (CT) of the thoracic cavity has allowed the development of new visualization tools. These tools provide a three-dimensional (3-D) endoluminal reconstruction of the tracheobronchial tree, as it would be viewed through a fibroscopic instrument. However, 3-D reconstruction techniques cannot replace conventional fibroscopy, which remains indispensable for obtaining histological samples. Furthermore, when CT-detected mediastinal or parenchymal lesions are not seen during fiberoptic bronchoscopy, guiding transbronchial needle biopsy is a major challenge. Computer-guided transbronchial biopsy involves the fusion of image data from both CT slices and bronchoscopic video sequences. This fusion is described in this paper in two parts. First, we present a segmentation process, using mathematical morphology operators, in order to analyze the video sequence and localize the bronchoscopic camera within the tracheobronchial tree. Second, we present tools used to match this localization knowledge with CT data. Finally, we produce images that create a bronchoscopic augmented reality, using elements extracted from the CT examination.
The goal of intrasurgical registration is to establish a common reference frame between presurgical and intrasurgical three-dimensional data sets that correspond to the same anatomy. This paper presents two novel techniques that have application to this problem, high-speed pose tracking and intrasurgical data selection. In the first part of this paper, we describe an approach for tracking the pose of arbitrarily shaped rigid objects at rates up to 10 Hz. Static accuracies on the order of 1 mm in translation and 1 degree in rotation have been achieved. We have demonstrated the technique on a human face using a high-speed VLSI range sensor; however, the technique is independent of the sensor used or the anatomy tracked. In the second part of this paper, we describe a general purpose approach for selecting near-optimal intrasurgical registration data. Because of the high costs of acquisition of intrasurgical data, our goal is to minimize the amount of data acquired while ensuring registration accuracy. We synthesize near-optimal intrasurgical data sets, based on an analysis of differential surface properties of presurgical data. We demonstrate, using data from a human femur, that discrete-point data sets selected using our method are superior to those selected by human experts in terms of the resulting pose-refinement accuracy.
Standard planning for dental implants consists of a prosthesis simulation on diagnostic casts and radiographic examination of anatomical structures. The clinician visually locates the planned trajectory on the surgical site in the patient's mouth without direct correlation between the radiographs and the anatomy. We have developed a computer assisted technique to define the optimal position of the bone implant using computed tomography to accurately place the implant in the planned position using a guide drilled into a resin splint.
In this preliminary study, the use of real-time ultrasonography to visualize the effects of acute interstitial Nd:YAG laser irradiation was investigated in the normal pig brain. In six pigs, a craniotomy was performed. In the frontal or temporal lobe, a thermal laser lesion was made using a 600-micron-diameter optical fiber at powers of 1 W, 2 W, and 4 W with exposure times of 5 min and 10 min. Ten to thirty minutes after laser irradiation, the pigs were sacrificed. Ultrasound imaging was performed before, during, and after laser irradiation. During laser irradiation, a clear hyperechogenic area was observed around the fiber tip. The onset of the changes and the extent of the lesion were dependent on the power and exposure time. Histologic examination showed thermal lesions consisting of coagulation necrosis and edema. The size of the lesions correlated well with size on ultrasound imaging. The maximal lesion dimension was 12 mm in diameter (4 W for 5 min). In conclusion, within the limitations of this experimental setup, it is feasible to visualize interstitial laser-induced lesions in the brain by ultrasonography. This method is safe and simple and may be helpful in future applications of interstitial thermotherapy in brain tissue.
Brain tumors are histologically heterogeneous. A technique for three-dimensional fusing of computed tomography (CT) or magnetic resonance images (MRI) with positron emission tomography (PET) images is described. This allows the anatomic detail provided by CT or MRI scans to be combined with the information about metabolic activity provided by PET scans. The fused images allowed selection of the most metabolically active portions of tumors. Fusion of CT and MRI images with PET scans has allowed first-pass diagnostic yield by providing the surgeon with a map of anatomical as well as functional (metabolic) detail. We describe a technique to allow routine fusion of MRI, CT, and PET information to help guide the neurosurgeon.
