Document Actions
Segmentation, Classification and Quantification
Up one levelThe following CMIV projects conducts research related to Segmentation, Classification and Quantification.
3D and 4D Adaptive Filtering for Image Denoising and Patient Safety
The aim of this medical image science project is to increase patient safety in terms of improved image quality and reduced exposure to ionizing radiation in CT. The means to achieve these goals is to develop and evaluate an efficient adaptive filtering (denoising/image enhancement) method that fully explores true 3D and 4D image acquisition modes. Four-dimensional (4D) medical image data are captured as a time sequence of image volumes. During 4D image acquisition, a 3D image of the patient is recorded at regular time intervals. The resulting data will consequently have three spatial dimensions and one temporal dimension. Increasing the dimensionality of the data impose a major increase the computational demands. The initial linear filtering which is the cornerstone in all adaptive image enhancement algorithms increase exponentially with the dimensionality. On the other hand the potential gain in Signal to Noise Ratio (SNR) also increase exponentially with the dimensionality. This means that the same gain in noise reduction that can be attained by performing the adaptive filtering in 3D as opposed to 2D can be expected to occur once more by moving from 3D to 4D.
3D Modelling of the Mastoid Bone. Structural and developmental analysis of the human air cell system.
The mastoid bone contains a set of numerous small air cells, which are connected together and located in the backside of the middle ear underneath the cranium. Development of middle ear infections (otitis media = OM) together with sequelae concerning hearing lost are a common factor for a net absorption of gas from the mastoid bone that leads to a negative pressure in the middle ear (MEP). The origin of this “gas exchange” results in a difference in gas concentration between the air phase and the blood phase (via the mucosa). The structure of the air cell system in the mastoid bone with its numerous ramifications implies a relative high surface area compared to its volume which results in a high surface area to volume ratio, and is therefore considered to be very effective for gas exchange. The related clinical problems are important and lead to a main case for ear surgery treatment in order to improve hearing. <p> Similarly, the lungs have numerous ramifications (bronchi) and air cells (alveoli), which also lead a high surface area to volume ratio leading to an effective gas exchange, by inhalation of oxygen through the trachea. Other similarities can be found between the structure and the functionalities of the lungs and the mastoid: 1) histologi-cal structure of the flat epithelium with rich vascularization of the underlying network of capillaries, and 2) the central neural feedback regulation for the both the lungs and the mastoid bone is performed in the same cranial nerve (IX and X) and located in the same area of the brainstem. These similarities have led to the naming of the mastoid bone as a miniature lung. <p> In spite of the important role of the mastoid bone in the development of OM and sequelae, research concerning the structure and the functionality of the mastoid bone is rather poor because of its inaccessibility. Improved CT scanning methods generate more detailed pictures, which can be used as the basis for advanced image processing techniques with possible 3D reconstruction and derivation of quantitative data describing the inherent structure of study. Such investigations have been performed for the lungs and lead to a better analysis of the lung functionality. Due to the similarities of the mastoid bone structure and function with the ones from the lungs, the use of a similar method, new knowledge concerning the development of OM as well as corresponding sequelae could be derived in order to improve treatments about preserving hearing.
Absolute Quantification of Multinuclear Magnetic Resonance Spectroscopy
The major aim of this project is in collaboration with several clinics to expand the scope of medical magnetic resonance methods of water to a large number of metabolites and other functional tissue properties in order to significantly enhance the level of todays applications of clinical MR. The work covers developing novel acquisition technologies and hardware, as well as clinical applications of quantitative molecular spectroscopy and imaging. A major long-term aim is to shift MR-applications from a qualitative to a quantitative mode.
Adolescent Idiopathic Scoliosis - Deformity in Three Dimensions
Until now, scoliosis has been diagnosed with a standing X-ray to determine the size of the deformity in the frontal plane, the Cobb angle. New technology based on low-dose CT makes it possible to reconstruct the deformity in 3D. This provides better possibilities to assess the vertebral rotation. The rotation seems to be of greater importance than previously thought. The project consists of the following four studies: 1. Investigation of the correlation between the size of the deformity on pictures from standing X-ray and supine computer tomography. The deformity increases when put under pressure. 2. Development and evaluation of a new software/method to take measurements and do calculations in 3D-reconstructions of scoliosis based on low-dose CT. The software will be developed in collaboration with Sectra. 3. Investigation of the correlation between the rotation of the vertebrae and the size of the deformity. The correlation will be investigated both pre- and post-operatively. 4. Comparison of how well two different surgical approaches work to correct the deformity in scoliosis. The project has natural ties to CMIV. The low-dose CT examinations are performed at CMIV and require optimal use of the new CT equipment there and the new software will be developed and evaluated in collaboration with Sectra. It also seems very natural to work with CMIV because of their great knowledge in 3D-reconstructions from different radiological examinations. Moreover, collaboration with researchers in the technical field will also most likely be beneficial for the project.
An uncertainty-aware visualization pipeline for volume rendering
Within Direct Volume Rendering (DVR) there is one important aspect of knowledge and data representation that has been largely overlooked - Uncertainty. In all measurement and simulations there are inherent inaccuracies and throughout the visualization pipeline additional uncertainties in processing, rendering and interaction are introduced. If these uncertainties are not conveyed the result may be misinterpretations and false conclusions. In medical visualization the problem is particularly pertinent with the hazards of wrong diagnosis and mistreatment of medical conditions. The overall goal in this project is: To develop an uncertainty aware real-time Direct Volume Rendering pipeline based on domain knowledge.
Automated Generation of Patient Specific Models for Visual and Haptic Simulation of Hip Fracture Surgery
The goal of this project is an auto-generated patient specific model for haptic and visual simulation of hip fracture surgery. Osteoporotic fractures constitute a problem of increasing clinical importance. A problem with the cervical type of hip fracture is the great risk of complications. A patient-specific simulation model would enable the surgeon to perform simulated surgery on the patient. Instead of discussing alternative techniques using plain X-ray films, the surgeon would have the chance to test several operative approaches, resulting in a safer and more rapid real operation. In addition, these models would be useful in the training of surgeons and development of new techniques. The first step in the generation of the model is segmentation of the bone in a CT-volume. Then, the local bone density will be estimated from the CT-data. The resulting information will then be converted to fit models suitable for visual and haptic simulation.
Classification of multi-variate medical datasets using deformable models
Computer systems tailored for diagnostic tasks can help increase resource efficiency in the healthcare. With systems that help in classifying medical image data, the time spent on time-consuming manual tasks can be cut and diagnostic accuracy can be increased. The recent development of new imaging techniques and scanners, such as synthetic MR and dual-energy CT, has given rise to datasets that comprise of several measurements at a single spatio-temporal position. The added information presents new possibilities to extract relevant features from the data.
Clinical Application of Synthetic MRI on Patients with Malignant Gliomas
Patients with malignant gliomas are treated with surgery, chemo- and radiotherapy and then followed with MRI-examinations to detect early signs of tumour recurrence. Almost 25% of patients react to the treatment with oedema and contrast-enhancement in the affected area of the brain, and this image is difficult to distinguish from tumour recurrence. Since the image is non-specific, radiologists in these cases look to more quantitative methods such as MR Spectroscopy and PET-CT, but there are still cases where it is unclear if the patient should be treated with further surgery or just followed with further imaging. Synthetic MRI is a quantitative MR-method that enables quantitative measurement of the tissue. If it is possible to find tumour specific quantitative values, it might be possible to distinguish tumour from treatment effects and thereby improving the diagnostic arsenal in these difficult cases.
Effect of reperfusion on infarct size and cardiac function evaluated with MRI and echocardiography - MrSTEMI
Mechanical opening of the infarct related artery (primary PCI) in patients with ST-elevation myocardial infarction (STEMI) seems to produce better results than iv thrombolysis. Our hypothesis is that primary PCI saves myocardium, that the size of myocardial damage is best quantified with contrast-enhanced MR (CEMR), that salvaged myocardium translates into better cardiac function, and that the time to opening of the artery is directly related to the size of the infarct. We will attempt to study the effect of the delay between the start of symptoms and the opening of the infarct related artery. The infarct risk area will be estimated from echocardiography performed in the cath lab during initiation of treatmentand expressed in terms of wall motion score index, WMSI. The final damage will be assessed from a comprehensive MR study with late enhancement performed at 6 weeks post PCI.
Functional evaluation of scaphotrapeziotrapezoid fusion: A retrospective study.
The indication for surgery with STT - fusion includes Kienböck´s disease, STTarthrosis and radiocarpal instability combined with pain. Methods: 25 patients, 15 women and 10 men, underwent STT arthrodesis between January 2000 and June 2005. 15 patients suffered from arthrosis, 7 from Kienböck´s disease and 3 had pain combined with radiocarpal instability. Conventional X-ray was performed approxymately 10 weeks after surgery. A long term follow up is planned where the patients will be re-examined at a mean follow –up time of 35 months. Active range of motion (AROM) will be verified with a goniometer; grip strength measured with a JAMAR-Dynamometer II. The Key pressure grip measured with pinch gauge. Pain will be evaluated by a visual analogue scale (VAS) from zero to ten for stress and during resting conditions. We will use a Disabilities of the Arm, Shoulder and Hand questionnaire (DASH) to capture the patients´ upper-extremity function. Occupational conditions shall be evaluated. CT scan with 3D reformation will be used to evaluate degree of bone fusion, scafolunar angel and signs of arthrosis.
HEART4FLOW - Improved Diagnosis and Management of Heart Disease by 4D Blood Flow Assessment
The primary purpose of the cardiovascular system is to drive, control and maintain blood flow to all parts of the body. Despite the primacy of flow, cardiac diagnostics still rely almost exclusively on tools focused on morphological assessment. The objective of the HEART4FLOW project is to develop the next generation of methods for the non-invasive quantitative assessment of cardiac diseases and therapies by focusing on blood flow dynamics, with the goals of earlier and more accurate detection and improved management of cardiac diseases.
Histological and functional effects of aortic valve disease on left ventricular function prior to and after surgical intervention
This research project aims to apply, modify and validate cardiac magnetic resonance imaging (cMRI) as a diagnostic tool for identification of fibrotic changes in the heart muscle due to pressure and volume overload caused by aortic valve disease. Furthermore, we hypothesize that these tissue changes (the amount and the location of the fibrotic tissue) can be connected to the impairment of the left ventricular function (LVF) in the advanced natural history of aortic valve disease. By using cMRI we hope to gain further information by non-invasive means on whether this impairment is reversible following surgical therapy. In addition to histological and functional studies at rest we plan to survey the anaerobe (physical) capacity of the study persons by performing cardiopulmonary exercise testing pre- and postoperatively and study the relationship between physical performance capacity and left ventricular function (LVF).
Identification of cognitive processes with fMRI and auditory stimulation in hearing impaired with and without hearing aids
Within this project, we will investigate the neural correlates of cognitive processes during speech intelligibility in noise. The data will be analysed according to the working memory framework of Ease of Language Understanding (ELU) developed by Rönnberg and colleagues. This model states that the demands on cognitive (‘explicit’) processing increase when speech comprehension is impaired by background noise, hearing loss, or altered by the type of signal processing in the hearing instrument.
Imaging of Brown Adipose Tissue in Man
We will develop non-invasive methods for measurements of human Brown Adipose Tissue (hBAT) tissue mass and activity. Our hypothesis is that this can be achieved by means of magnetic resonance imaging (MRI) and dual energy computed tomography (DECT). Initial studies will be performed using rodents (mice and rats). An important next step will be to use human postmortem material, which will enable us to confirm the true identity of hBAT images by genetic and morphological analysis of biopsies. The validated methods will then be used for in vivo studies. We will use phase sensitive reconstruction of complex images acquired with the water and fat resonance in- and out- of-phase, so called Dixon imaging.
Implementation of Synthetic MRI into the clinical workflow.
Synthetic MRI is the approach of rapid quantification of MRI parameters and the subsequent synthesis of a whole range of contrast images based on the quantified data. This implies that a single scan is sufficient to generate any conventional T1- or T2 weighted image. It is even possible to visualize far stronger, non-physical contrast such as tissue specific imaging. Application of Synthetic MRI might save up to a third of the patient examination time and will make MRI more reliable and quantitative. The project aims at the clinical implementation of the approach into the PACS system. The technique of rapid quantification is more or less mature but the general use of Synthetic MRI in daily clinic needs to be introduced and validated. In addition to the investigation of the quality of the resulting images the specification of the time-saving aspect will be important. Cardiac Late Enhancement is implemented first and the validation is on-going. Synthetic Brain imaging is implemented at the moment. Future directions will concern the liver.
Managing large scale data in medical visualization
The data sets produced by state-of-the-art medical imaging modalities can be extremely large, and there is a clear trend that they are rapidly increasing even further. Such data sets cannot be handled by traditional viewing systems. This project aims to develop new visualization and data management techniques that enables efficient human anlysis of large medical data sets in the clinical environment.
Manifold-Valued Signal Proocessing
A manifold is a mathematical concept which generalizes surfaces to higher dimensions. Examples of 2-dimensional manifolds are for instance the surface of a sphere and the the surface of a torus, both being examples of non-linear manifolds. Locally however, manifolds are flat and equivalent to the an Euclidean space. Features found in signals can often be described using manifolds. This is often not stated explicitly, but instead various parameterizations of manifolds are used. In order to describe a quantity which can be seen as a point on a sphere, spherical coordinates are commonly used. This has some drawbacks however which we wish to avoid if possible. We see a need for manifolds in the field of medical image analysis. Medical doctors express a wish to objectively quantify various features in medical images, such as local texture, shape and orientation of organs. We know from previous research that manifolds can do the job, but we lack a generic framework for dealing with manifold-valued signals in signal processing. In fact, we believe that such a framework will be useful in other areas signal processing too. The goal of this project is to explore a specific flavour of signal processing and continue the development of methods 1) to learn or identify manifold-valued representations from examples and 2) apply signal processing on manifold-valued signals which is analogous to filtering and interpolation using convolution operators in classic signal processing.
Morphology Guided fMRI
As the potentials for treating neurological disorders have increased tremendously the last decades, there is also a growing need for safe, reliable and cost-effective diagnostic tools. fMRI is valuable both for an improved description of normal brain function and for assessment of patients with neurological disorders. The core theoretical idea in the project is that by including/developing tools for reconstruction of the brains cortical surface new and highly significant local spatial priors can be included in the fMRI data analysis and in this way significantly improve detection performance.
MOVIII Demonstrator Project: Bio-Feedback Using Real-Time fMRI
Despite the enormous complexity of the human mind, fMRI techniques are able to partially observe the state of a brain in action. In this paper we describe an experimental setup for real-time fMRI in a bio-feedback loop. One of the main challenges in the project is to reach a detection speed, accuracy and spatial resolution necessary to attain sufficient bandwidth of communication to close the bio-feedback loop. To this end we have banked on our previous work on real-time filtering for fMRI and system identification, which has been tailored for use in the experiment setup. In the experiments presented the system is trained to estimate where a person in the MRI scanner is looking from signals derived from the visual cortex only. We have been able to demonstrate that the user can induce an action and perform simple tasks with her mind sensed using real-time fMRI.The technique may have several clinical applications, for instance to allow paralyzed and "locked in" people to communicate with the outside world. In the meanwhile, the need for improved fMRI performance and brain state detection poses a challenge to the signal processing community. We also expect that the setup will serve as an invaluable tool for neuro science research in general.
MRI in the diagnosis of Neuroborreliosis
This project attempts to study whether MRI can contribute to the diagnosis of Borrelia infection iin the brain. A unique material of acute and chronic patients with proven or suspected neuroborreliosis will, in addition to clinical, biochemical and immunological methods for diagnosis, be examined with MRI of the brain. An extensive MRI protocol is used, including Gd-enhanced images and diffusion-weighted imaging, as well as methods for quantitative assessment of MR parameters and synthetic MRI developed at CMIV.
Dissertation by Daniel Forsberg
