Catálogo de publicaciones - libros

We propose to segment Multiple Sclerosis (MS) lesions overtime in multidimensional Magnetic Resonance (MR) sequences. We use a robust algorithm that allows the segmentation of the abnormalities using the whole time series simultaneously and we propose an original rejection scheme for outliers. We validate our method using the BrainWeb simulator. To conclude, promising preliminary results on longitudinal multi-sequences of clinical data are shown.

In the last five years, Deep Brain Stimulation (DBS) has become the most popular and effective surgical technique for the treatment of Parkinson’s disease (PD). The Subthalamic Nucleus (STN) is the usual target involved when applying DBS. Unfortunately, the STN is in general not visible in common medical imaging modalities. Therefore, atlas-based segmentation is commonly considered to locate it in the images. In this paper, we propose a scheme that allows both, to perform a comparison between different registration algorithms and to evaluate their ability to locate the STN automatically. Using this scheme we can evaluate the expert variability against the error of the algorithms and we demonstrate that automatic STN location is possible and as accurate as the methods currently used.

Tachyarrhythmias are pathological fast heart rhythms often caused by abnormally conducting myocardial areas (foci). Treatment by radio-frequency (RF) ablation uses electrode-catheters to monitor and destroy foci.The procedure is normally guided with x-rays (2D), and thus prone to errors in location and excessive radiation exposure. Our main goal is to provide pre- and intra-operative 3D MR guidance in XMR systems by locating the abnormal conduction pathways. We address the inverse electro-mechanical relation by using motion in order to infer electrical propagation. For this purpose we define a probabilistic measure of the onset of regional myocardial activation, derived from 3D motion fields obtained by tracking tagged MR sequences with non-rigid registration. Activation isochrones are then derived to determine activation onset.

We also compare regional motion between two different image acquisitions, thus assisting in diagnosing arrhythmia, in follow up of treatment, and in determining whether the ablation was successful.Difference maps of isochrones and other motion descriptors are computed to determine abnormal patterns. Validation was carried out using an electro-mechanical model of the heart, synthetic data, a cardiac MRI atlas of motion and geometry, MRI data from 6 healthy volunteers (one of them subjected to stress), and an MRI study on a patient with tachyarrhythmia, before and after RF ablation. A pre-operative MRI study on a second patient with tachyarrhythmia was used to test the methodology in a clinical scenario, predicting the abnormally conducting region.

This paper presents a method for the initial detection of ductal structures within 3D ultrasound images using second-order shape measurements. The desire to detect ducts is motivated in a number of way, principally as step in the detection and assessment of ductal carcinoma in-situ. Detection is performed by measuring the variation of the local second-order shape from a prototype shape corresponding to a perfect tube. We believe this work is the first demonstration of the ability to detect sections of duct automatically in ultrasound images. The detection is performed with a view to employing vessel tracking method to delineate the full ductal structure.

We have been developing an approach for automatically quantifying organ motion for adaptive radiation therapy of the prostate. Our approach is based on deformable image registration, which makes it possible to establish a correspondence between points in images taken on different days. This correspondence can be used to study organ motion and to accumulate inter-fraction dose. In prostate images, however, the presence of bowel gas can cause significant correspondence errors. To account for this problem, we have developed a novel method that combines large deformation image registration with a bowel gas segmentation and deflation algorithm. In this paper, we describe our approach and present a study of its accuracy for adaptive radiation therapy of the prostate. All experiments are carried out on 3-dimensional CT images.

A method is presented for the calculation of perfusion parameters in dynamic contrast enhanced MRI. This method requires identification of enhancement curves for both tumor tissue and plasma. Inter-operator variability in the derived rate constant between plasma and extra-cellular extra-vascular space is assessed in both canine and human subjects using semi-automated tumor margin identification with both manual and automated arterial input function (AIF) identification. Experimental results show a median coefficient of variability (CV) for parameter measurement with manual AIF identification of 21.5% in canines and 11% in humans, with a median CV for parameter measurement with automated AIF identification of 6.7% in canines and 6% in humans.

The Taylor spatial frame is a fixation device used to implement the Ilizarov method of bone deformity correction to gradually distract an osteotomized bone at regular intervals, according to a prescribed schedule. We improve the accuracy of Ilizarov’s method of osteogenesis by preoperatively planning the correction, intraoperatively measuring the location of the frame relative to the patient, and computing the final shape of the frame. In four of five tibial phantom experiments, we were able to achieve correction errors of less than 2 degrees of total rotation. We also demonstrate how registration uncertainty can be propagated through the planned transformation to visualize the range of possible correction outcomes. Our method is an improvement over an existing computer–assisted technique (Iyun  [3]) in that the surgeon has the same flexibility as in the conventional technique when fixating the frame to the patient.

Assessment of soft tissue in normal and abnormal joint motion today gets feasible by acquiring time series of 3D MRI images. However, slice-by-slice viewing of such 4D kinematic images is cumbersome, and does not allow appreciating the movement in a convenient way. Simply presenting slice data in a cine-loop will be compromised by through-plane displacements of anatomy and “jerks” between frames, both of which hamper visual analysis of the movement. To overcome these limitations, we have implemented a demonstrator for viewing 4D kinematic MRI datasets. It allows to view any user defined anatomical structure from any viewing perspective in real-time. Smoothly displaying the movement in a cine-loop is realized by image post processing, fixing any user defined anatomical structure after image acquisition.

Bio-engineered cartilage has made substantial progress over the last years. Preciously few cases, however, are known where patients were actually able to benefit from these developments. In orthopaedic surgery, there are two major obstacles between in-vitro cartilage engineering and its clinical application: successful integration of an autologuous graft into a joint and the high cost of individually manufactured implants. Computer Assisted Surgery techniques can potentially address both issues at once by simplifying the therapy, allowing pre-fabrication of bone grafts according to a shape model, individual operation planning based on CT images and providing optimal accuracy during the intervention. A pilot study was conducted for the ankle joint, comprising a simplified rotational symmetric bone surface model, a dedicated planning software and a complete cycle of treatment on one cadaveric human foot. The outcome was analysed using CT and MRI images; the post-operative CT was further segmented and registered with the implant shape to prove the feasibility of computer assisted arthroplasty using bio-engineered autografts.

Subject motion appears to be a limiting factor in numerous magnetic resonance imaging (MRI) applications. For head imaging the subject’s ability to maintain the same head position for a considerable period of time places restrictions on the total acquisition time. For healthy individuals this time typically does not exceed 10 minutes and may be considerably reduced in case of pathology. In particular, head tremor, which often accompanies stroke, may render certain high-resolution 2D and 3D techniques inapplicable. Several navigator techniques have been proposed to circumvent the subject motion problem. The most suitable for head imaging appears to be the orbital or spherical navigator methods. Navigators, however, not only lengthen the measurement because of the time required for acquisition of the position information, but also require additional excitation radio frequency (RF) pulses to be incorporated into the sequence timing, which disturbs the steady state. Here we demonstrate the possibility of interfacing the MR scanner with an external optical motion tracking system, capable of determining the object’s position with sub-millimeter accuracy and an update rate of 60Hz. The movement information on the object position (head) is used to compensate the motion in real time. This is done by updating the field of view (FOV) by recalculating the gradients and the RF-parameter of the MRI tomograph during the acquisition of k-space data based on the tracking data. Results of rotation phantom, in vivo experiments and the implementation in two different MRI sequences are presented.