Sei sulla pagina 1di 5

INTRODUCTION

Diagnostic imaging is an invaluable tool in medicine today. Magnetic resonance imaging (MRI), computed tomography (CT), digital mammography, and other imaging modalities provide an effective means for noninvasively mapping the anatomy of a subject. These technologies have greatly increased knowledge of normal and diseased anatomy for medical research and are a critical component in diagnosis and treatment planning. A medical image processing problem may (but doesnt always) contain the following steps:

Image formation, which is performed by the image acquisition device. Image enhancement has the main goal of making objects visible, by eliminating noise, improving edges and contrast, equalizing histograms. Image segmentation refers to finding and distinguishing objects in the image, separating objects from background or separating relevant objects from irrelevant ones. Segmentation can also refer to separating a region of interest (ROI) from the other parts of the image. For example: brain and non-brain in an MRI image.

Sometimes in medical imaging, during an investigation, not only one 2-D picture is taken, but several parallel ones or different views of the same volume. In such cases, the 2-D segmentation is followed by the reconstruction of the 3-D surface of objects.

Sometimes the segmentation is performed in 3-D. In this case there is one more operation that is performed simultaneously, called image registration, which has the main goal to define the actual placement of the 2-D slices or different views.

Image interpretation is absolutely a problem specific question. In a brain MR slice it is likely to look for white and gray matter, while in an X-ray lung scan we are probably interested in the contour of the lung. On the other hand, different gray levels of an MR image are closely related to the tissue type that is being scanned, while in ultrasound imaging the intensity values of individual pixels have hardly any meaning.

This list may be completed with a further step called visualization, which is not anymore image processing: it is an issue of computer graphics and machine vision. Computerized medical image processing has the main goal to perform steps 2-6 of the list presented above as automatically as possible. Magnetic resonance imaging (MRI) is one of the ways to look inside the human body, that is, to take pictures of its cross sections without cutting (invasive technique). Magnetic resonance imaging (MRI) has become an essential part of the diagnostic radiology now days. MRI, like ultrasound and x-ray computed tomography (CT), is used to generate two dimensional images of sections of the body. It can be extended for 3D visualization of anatomical and functional views of living system. MRI, since its invention in the early 70s, has evolved into one of the most powerful non-invasive techniques for producing anatomical images from living systems. There are many techniques used to obtain MR images, but all involve excitation by radio frequency (RF), magnetic field of the nuclear spins in a sample of interest. The precession of these spins in a carefully controlled magnetic field causes the sample to echo back with a composite electromagnetic signal whose analysis can reveal the types and positions of the nuclei within the sample. The high contrast sensitivity to soft tissues differences and the inherent safety to the patient resulting from the use of non-ionizing radiation have been key reasons, why MRI has replaced many CT and projection radiography methods. Magnetic resonance imaging (MRI) uses two natural forces a magnetic field and radio frequency pulses- to enable noninvasive , high resolution , three dimensional images of the bodys internal structure. The local magnetic resonance signal which has a information of the body tissue, is produced as a result of the interaction of the particular pulse sequence parameter with local tissue properties. Magnetic resonance imaging is also applied to study the behavior of cells and how they react with disease and treatment. Subatomic particles (electron, proton and neutron) can be imagined as spinning on their axes. MRI system has been active area for researchers. Due to fast computing technology available, the performance of MRI system has been improving, especially, in time reduction , improvements in resolution, removal of artifacts and flexibility in operation . The new tools are being used in various steps of MRI etc Among all other computed tomography (CT) methods, MRI has several advantages:

1. Neither the constant magnetic field, nor the radio frequency modulated one, which is present in the MRI device, represent a danger to humans. Other CT methods, for example the positron emission tomography, use X-rays to create the image. 2. The image contrast of a CT scan only depends on the electron density of the tissues considered. The MRI signal is determined by the proton density of the tissue, T1 and T2 relaxation times, the type of sequence used, and the selected acquisition parameters. These parameters give the opportunity to enhance the image contrast between two tissues by cleverly choosing the type of sequence and acquisition parameters, and thus optimize the differentiation between tissue structures. 3. Some MRI devices give us the opportunity to produce multi-spectral images, that is, two or more images of the same cross section with different parameter settings.

In order to prevent possible misinterpretations, most MR images are acquired such a way that the tissue contrast of various images is determined mainly on a single tissue parameter. In this context, T1, T2, and PD-weighted images are produced.

Digital signal processing (DSP) describes the science that tries to analyze, generate and manipulate measured real world signals with the help of a digital computer. These signals can be anything that is a collection of numbers, or measurements and the most commonly used signals include images, audio (such as digitally recorded speech and music) and medical and seismic data. In most digital signal processing applications, the frequency content of the signal is very important. The Fourier transform (FT) is probably the most popular transform used to obtain the frequency spectrum of a signal. Noise removal or de-noising is an important task in image processing. Image enhancement is a collection of techniques that improve the quality of the given image that is making certain features of the image easier to see or reducing the noise. In general, the results of the noise removal have a strong influence on the quality of the image processing techniques. Noise generated by electronic components in instrumentation is a common type of random signal that is present in much biomedical data even though contemporary electronic design minimizes this noise. Often those components of a signal which are not understood are classified as noise. The ultimate basis for deciding what constitutes noise should be derived from considerations about the experimental or clinical measurements and the
3

source of a signal. Ideally when a priori knowledge for judging whether certain components of a signal represent the desired measurement or not is known then the signal processing method is chosen to enhance the desired signal and reduce undesired signal components. In some cases this information may not be known and it may be necessary to examine the results of the signal processing steps to assess whether the output signal exhibits some apparent separation into desired and noise components. The field of imaging provides many examples of both biomedical images and biomedical image processing. Magnetic resonance imaging (MRI) is excellent for showing abnormalities of the brain such as: stroke, haemorrhage, tumour, multiple sclerosis or lesions. In the MRI basic signals are currents induced in a coil caused by the movement of molecular dipoles as the molecules resume a condition of random orientation after having been aligned by the imposed magnetic field. Signal processing is required to detect and decode them, which is done in terms of the spatial locations of the dipoles (which is related to the type of tissue in which they are located). Much of the associated signal processing is based on Fourier transform. Since MRI utilizes two-dimensional Fourier transforms the basic concepts are the same.

Fig 1.1. Sequence of MR image slices

Fig 1.2. 3-D model visualization

Image processing can be defined as the manipulation of an image for the purpose of either extracting information from the image or producing an alternative representation of the
4

image. There are numerous specific motivations for image processing but many fall into the following categories: (i) to remove unwanted signal components that are corrupting the image and (ii) to extract information by rendering it in a more obvious or more useful form. Segmentation is a process of partitioning the image into some non-intersecting regions such that each region is homogeneous and the union of no two adjacent regions is homogeneous. In medical imagery, segmentation can be defined as the process of partitioning and labelling an individual point of a data set by tissue type, based on properties of the observed intensities as well as anatomical knowledge about normal subjects. The trained eye of a medical expert is able to segment regions of interest not only by detecting edges, but by prior knowledge of anatomy (shape) and by exploiting the coherency of the tissue or the organ under study as well

Potrebbero piacerti anche