StudentShare
Contact Us
Sign In / Sign Up for FREE
Search
Go to advanced search...
Free

Functional Magnetic Resonance Imaging - Article Example

Summary
This article 'Functional Magnetic Resonance Imaging' explains the basic scientific technology about Functional Magnetic Resonance Imaging (fMRI), how it has enabled scientists and doctors to determine brain activity, and how fMRI graphical data is read and interpreted…
Download full paper File format: .doc, available for editing
GRAB THE BEST PAPER92.6% of users find it useful

Extract of sample "Functional Magnetic Resonance Imaging"

Mind Reading” and fMRI Data Analysis My My E-mail This article explains the basic scientific technology about Functional Magnetic Resonance Imaging (fMRI), how it has enabled scientists and doctors to determine brain activity, and how fMRI graphical data is read and interpreted. Using sample fMRI illustrations and substantial materials, the author summarises the procedures and methods for image acquisition, image processing, and data analyses using sample fMRI outputs using a General Electric MRI scanner. The key objective of fMRI data analysis is to minimise false positives and establish a high statistical probability of correlation and causation amongst human body operations, images of brain haemodynamics, and the respective brain operations. KEYWORDS: fMRI, BOLD signal, image acquisition, image processing, data analysis, statistical analysis. Introduction Magnetic resonance imaging or MRI is a technology that when applied to medicine allows medical doctors to capture an accurate image of the human body. MRI is founded on the scientific principles underlying nuclear magnetic resonance, which has the ability to capture energy released when the magnetic field of an atomic nucleus changes its alignment in the presence of an external stimulus such as a magnetic field or a radio wave signal. [1] The ability of physicists to capture an image of the internal structure of atomic molecules was eventually applied to medical diagnosis, with very useful and accurate results. 1. Medical Applications of fMRI MRI was at first useful in capturing images of the human body to determine the presence of tumours or blocked arteries. MRI was likewise very helpful for doctors in planning surgeries more accurately. Amongst medical fields that benefited greatly from MRI was neuroscience, the study of the human brain, because MRI offered a safe, non-invasive, and accurate procedure to study one of the most delicate organs of the human body. More recently, scientists discovered that MRI detected changes in blood flow inside the brain. This phenomenon is called haemodynamics, i.e. the flow of blood or haemoglobin, leading to what is now called functional MRI or fMRI. [2] Whilst conventional MRI focused mainly on traditional anatomical imaging, fMRI has the ability to map human brain function by identifying the specific parts of the brain where extra oxygen is being used, resulting in the expansion of blood vessels and changes in blood chemistry. This is called the haemodynamic response, evidence that a part of the brain is working and controlling other parts of the body. When a human being does something, the brain metabolism affected by this action goes up and the result is a change in the MRI signal of that part of the brain. Thus, scientists can find out which specific portion of the brain controls each human action. fMRI lets doctors and neuroscientists study mental operations by observing brain structure and find out which brain structures are involved in specific human functions such as movement, thinking, and so on by comparing the dynamics of the flow of blood in different brain regions. By visualising changes in the blood flow and chemical composition of different regions in the brain, or how fluids change their flow over a certain period of time, fMRI offers doctors a clear and accurate map of the functioning and physiology of the brain. [3] This can also be applied to the study of the other human organs. Applied to neuroscience, fMRI helps doctors “read the mind” by studying how neural networks develop and allow humans to speak, see, hear, move and perform all the other complex operations of the human body. In the future, fMRI can more accurately help doctors learn how humans think and make decisions, aside from allowing them to understand and learn more about how brain abnormalities develop, how stroke victims recover their brain and motor functions, and how severe nervous disease disorders such as schizophrenia, Alzheimer’s or Parkinson’s develop and affect people. [4] Such research and new knowledge would be helpful not only in discovering lasting cures but, perhaps more important to assure an over-all improvement in public health, in prevention in the first place. This has far-reaching implications not only in health care costs and government and insurance policies, but also in determining more effective strategies for medical intervention. 2. fMRI Data and Analysis Advances in computer software that allow doctors to capture MRI signals have helped realise advances in the usefulness of fMRI in the field of neuroscience. With the use of the proper imaging sequence, doctors are able to study and observe the functions of the various regions in the cortex: visual, motor, and Broca’s area of speech and language-related activities. [5] There are several advantages derived from fMRI in producing brain activity images related to a specific task or sensory process. For one, the signal does not use radioactive isotopes and is therefore safer for the patient and the doctors. Besides, fMRI is faster. The procedure takes only a short time, e.g., 1.5 to 2 minutes each depending on the scanning paradigm. Likewise, fMRI produces high in-pane resolution of the functional image: a 1.5 by 1.5 mm resolution is standard but more enhanced images could be captured by more advanced machines, making fMRI optimally suitable for analysis and making more accurate plans for neurosurgical or treatment interventions. [6] 3. Data Acquisition and Processing The technology of fMRI uses variations in MRI signals from functional brain activity. The most widely used method depends on Blood Oxygenation Level Dependent or BOLD signal changes due to blood flow and metabolic characteristics of neuronal responses. [7] In the brain, specific functions are localised at various sites, allowing fMRI to identify and map functional specialisation at high spatial resolutions. Neuroscience has allowed the matching of specific areas in the brain with their corresponding mental behaviour through BOLD-fMRI methods. Whilst imaging methods and procedures depend on the machine used and can be different from one another, this paper looks at the method applied using a 1.5T General Electric Magnetic Resonance Imaging System. The author is indebted for this section of the paper on the Columbia University MRI website that contains a detailed description of its fMRI system. [8] Columbia University’s General Electric system has an echo planar option that results in rapid acquisition of images at slice thicknesses set at 3-5 mm. Simultaneous images can be captured on as many as 16 contiguous slices oriented along any suitable plane. Getting 21 slices is possible but this would take longer. Each imaging series requires approximately 30 complete head volume acquisitions. [8] Processing the image requires a stand-alone facility separate from the image acquisition or scanner system. This facility is where the computations required are made to reconstruct the large numbers of images and to provide the statistical analyses that allows neuroscientists to determine which anatomical regions are active during specific tasks. [8] How is a patient scanned using fMRI whilst performing a simple task, such as tapping a finger or shaking a foot? As in a conventional scan, the patient is positioned in the scanner and plane lines are set based on conventional scanning methods. In a typical fMRI series, 30 images are acquired in a 90 sec run where the initial and last 10 images are baseline conditions and the middle 10 images at 30 secs are acquired during a task. At Columbia University, in the case of a typical task designed to identify eloquent brain tissue involved in hand and finger movement, the patient taps fingers and thumb during the activity epoch. [8] The start and end of this activity period is cued by a visual or auditory signal and occurs at images 10 and 20 respectively. Language, sensory, visual, auditory, and other targeted functions are imaged in a similar manner. [8] 4. Data Analysis The aim of data analysis is to determine from the raw imaging data the brain regions where BOLD signal changes occur when stimulation takes place. The volume element in an fMRI scan is a voxel or volumetric pixel. A voxel represents a value on a regular grid in three-dimensional space. The voxel is the 3D equivalent of a pixel that, in turn, is a 2D image data in a bitmap. Like pixels in a bitmap, voxels do not typically have their coordinates explicitly encoded along with their values. Instead, the position of a voxel is inferred based on its position relative to other voxels or its position in the data structure that makes up a single volumetric image. [9] Thus, voxels are able to accurately represent regularly sampled spaces that are non-homogeneous. Whilst there are many techniques to detect activation or to determine the level of confidence a scientist can place in the results, i.e., what is the probability that a purely random response could be falsely labelled as activation. [9] This is how statistical analysis of fMRI data can help eliminate such false positives. There is a vast amount of data from a single fMRI procedure. According to neuroscientists, there are over 50,000 voxels in the brain, and each voxel can be activated or not activated by the procedure, leading to the generation of false positives that can change the interpretation of the outcome. The purpose of data analysis is to reduce the number of false positives and generate reliable findings. [9] There are three steps in fMRI data analysis. The first step is pre-processing of the data to improve the detection of activation events. These include capture of the images, correcting for the movement of the subject whilst undergoing the experiment, and correcting the data to improve the ratio of the image signal to noise. The second step is the statistical analysis to detect the voxels that show a stimulus response. The third step is displaying the activation images and quoting the probability values that determine statistical confidence to be established. A common statistical process used in fMRI data analysis is Statistical Parametric Mapping or SPM. SPM is a software package that constructs and assesses spatially extended statistical processes to test hypotheses about functional imaging data. [10] Aside from fMRI data, SPM software can also analyse data from other scanning techniques such as PET, SPECT, EEG and MEG. SPM is based on the modelling of voxels. First, images are realigned, spatially normalised into a standard space, and smoothed. Then, the parametric statistical models are assumed at each voxel, using the General Linear Model or GLM to describe the data in terms of experimental and confounding effects, and residual variability. Lastly, for fMRI the GLM is used in combination with a temporal convolution model. Classical statistical inference is used to test hypotheses expressed in terms of GLM parameters. This uses an image whose voxel values are statistics, a Statistic Image, or a Statistical Parametric Map. fMRI data analyses use powerful mathematical software like Matlab to work with applications such as SPM. 5. Sample Images and Interpretations Figure 1 from the Columbia University Website shows a sensory task involving tactile stimulation or touching of the left hand. The image is a 5 mm axial slice with voxels indicated that show significant signal changes with left hand tactile stimulation. [8] Figure 1: Image of tactile stimulation [8]. The image of the brain region shows which part of the brain – in this case, the right hemisphere – is actively stimulated by the experiment. The coloured portions allow neuroscientists to identify that this specific part of the brain is activated by left hand stimulation. The white graph on the right show the activity of the voxels through the three stages from baseline (no stimulation) to stimulation (left hand movement) and recovery (after the tactile experiment). The graph shows that the voxels corresponding to the coloured portions behave in a way that is statistically significant, i.e., that its behaviour is not due to chance stimulation. [11] Figure 2, also from the Columbia University website, shows the successive stages of analyses for left hand finger-thumb tapping. [8] Figure 2: Multi-stage fMRI Analysis [8]. Figure 2 shows that, as the scientists conducting the experiment discovered, the parts of the brain where significant activity is registered, as shown by the fMRI images on the left portion of the illustration, coincided with what they expected. The scientists can also conclude that such a coincidence is beyond the realm of chance and is therefore statistically significant. Mapping the human brain this way with the use of fMRI technology, neuroscientists have gained new knowledge not only in how the brain operates and which specific parts of the brain can be linked with each human action, but perhaps more importantly, in finding ways to address illnesses related to brain function and help patients in the search for a cure. [12] 6. Conclusion Although fMRI has its flaws, it seems to be the best state-of-the-art technology available for studies of the human brain. Amongst its flaws are issues related to correlation and causation, the complexity and non-localised nature of brain processes, and the need to use statistical methods carefully to avoid false positives. [12] The ability of neuroscientists to capture an image of the brain with the use of fMRI technology has given them the power to isolate simultaneous and coordinated brain events. By allowing the study of the various brain functions to be done with high precision, fMRI data analysis has made it easier to “read the mind” and to expand and explore the boundaries of neuroscience. In the end, the scientific objectivity of what the doctors “read” depends greatly on the methods they use to gather and analyse the available data. Reference List [1] Belliveau, J., Kennedy, D., McKinstry, R., Buchbinder, B., Weisskoff, R., Cohen, M., Vevea, J., Brady, T. & Rosen, B. 1991, “Functional mapping of the human visual cortex by magnetic resonance imaging”, Science, vol. 254, 17 June, pp. 716-719. [2] Ogawa, S., Tank, D. W., Menon, R., Ellermann, J. M., Kim, S. G., Merkle, H. & Ugurbil, K. 1992. “Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging”, Proc Natl Acad Sci USA, vol. 89, pp. 5951-5955. [3] Blamire, A. M., Ogawa, S., Ugurbil, K., Rothman, D., McCarthy, G., Ellermann, J. M., Hyder, F., Rattner, Z. & Shulman, R. G. 1992. “Dynamic mapping of the human visual cortex by high-speed magnetic resonance imaging”, Proc Natl Acad Sci USA, vol. 89, pp. 11069-11073. [4] Schneider, W., Noll, D. C. & Cohen, J. D. 1993, “Functional topographic mapping of the cortical ribbon in human vision with conventional MRI scanners”, Nature, vol. 365, 9 September, pp. 150-153. [5] Kim, S. G., Ashe, J., Georgopoulos, A. P., Merkle, H., Ellermann, J. M., Menon, R. S., Ogawa, S. & Ugurbil, K. 1993, “Functional imaging of human motor cortex at high magnetic field”, J Neurophysiology, vol. 69, no. 1, pp. 297-302. [6] Hinke, R. M., Hu, X., Stillman, A. E., Kim, S. G., Merkle, H., Salmi, R. & Ugurbil, K. (1993). “Functional magnetic resonance imaging of Brocas area during internal speech”, NeuroReport, vol. 4, pp. 675-678. [7] Atlas, S.W., Howard II, R.S., Maldijian, J., Alsop, D., Detre, J.A., Listerud, J., DEsposito, M., Judy, K.D., Zager, E. & Stecker, M. 1996, “Functional magnetic resonance imaging of regional brain activity in patients with intracerebral gliomas: Findings and implications for clinical management”, Neurosurgery, vol. 38, no. 2, pp. 329-338. [8] The future role of functional MRI in medical applications, [Online], About Functional MRI: Columbia University Program for Imaging and Cognitive Sciences (PICS). Available from: [November 22, 2009]. [9] Clare, S. 1997, Functional magnetic resonance imaging: Methods and applications, {Online}, Unpublished PhD thesis. Available from: [November 22, 2009]. [10] Lazard, N. 2008, The Statistical Analysis of Functional MRI Data, Springer, New York. [11] Hirsch, J., De la Paz, R., Relkin, N., Victor, J., Li, T., Rubin, N. & Shapley, R. 1995, “A study of how illusory contours activate specific regions in human visual cortex: evidence from functional magnetic resonance imaging”, Proc Natl Acad Sci USA, vol. 92, pp. 6469-6473. [12] Logothetis, N. K. 2008, “What we can do and what we cannot do with fMRI”, Nature, vol. 453, no. 7197, pp. 869-878.. Read More

Applied to neuroscience, fMRI helps doctors “read the mind” by studying how neural networks develop and allow humans to speak, see, hear, move and perform all the other complex operations of the human body. In the future, fMRI can more accurately help doctors learn how humans think and make decisions, aside from allowing them to understand and learn more about how brain abnormalities develop, how stroke victims recover their brain and motor functions, and how severe nervous disease disorders such as schizophrenia, Alzheimer’s or Parkinson’s develop and affect people.

[4] Such research and new knowledge would be helpful not only in discovering lasting cures but, perhaps more important to assure an over-all improvement in public health, in prevention in the first place. This has far-reaching implications not only in health care costs and government and insurance policies, but also in determining more effective strategies for medical intervention. 2. fMRI Data and Analysis Advances in computer software that allow doctors to capture MRI signals have helped realise advances in the usefulness of fMRI in the field of neuroscience.

With the use of the proper imaging sequence, doctors are able to study and observe the functions of the various regions in the cortex: visual, motor, and Broca’s area of speech and language-related activities. [5] There are several advantages derived from fMRI in producing brain activity images related to a specific task or sensory process. For one, the signal does not use radioactive isotopes and is therefore safer for the patient and the doctors. Besides, fMRI is faster. The procedure takes only a short time, e.g., 1.

5 to 2 minutes each depending on the scanning paradigm. Likewise, fMRI produces high in-pane resolution of the functional image: a 1.5 by 1.5 mm resolution is standard but more enhanced images could be captured by more advanced machines, making fMRI optimally suitable for analysis and making more accurate plans for neurosurgical or treatment interventions. [6] 3. Data Acquisition and Processing The technology of fMRI uses variations in MRI signals from functional brain activity. The most widely used method depends on Blood Oxygenation Level Dependent or BOLD signal changes due to blood flow and metabolic characteristics of neuronal responses.

[7] In the brain, specific functions are localised at various sites, allowing fMRI to identify and map functional specialisation at high spatial resolutions. Neuroscience has allowed the matching of specific areas in the brain with their corresponding mental behaviour through BOLD-fMRI methods. Whilst imaging methods and procedures depend on the machine used and can be different from one another, this paper looks at the method applied using a 1.5T General Electric Magnetic Resonance Imaging System.

The author is indebted for this section of the paper on the Columbia University MRI website that contains a detailed description of its fMRI system. [8] Columbia University’s General Electric system has an echo planar option that results in rapid acquisition of images at slice thicknesses set at 3-5 mm. Simultaneous images can be captured on as many as 16 contiguous slices oriented along any suitable plane. Getting 21 slices is possible but this would take longer. Each imaging series requires approximately 30 complete head volume acquisitions.

[8] Processing the image requires a stand-alone facility separate from the image acquisition or scanner system. This facility is where the computations required are made to reconstruct the large numbers of images and to provide the statistical analyses that allows neuroscientists to determine which anatomical regions are active during specific tasks. [8] How is a patient scanned using fMRI whilst performing a simple task, such as tapping a finger or shaking a foot? As in a conventional scan, the patient is positioned in the scanner and plane lines are set based on conventional scanning methods.

Read More
sponsored ads
We use cookies to create the best experience for you. Keep on browsing if you are OK with that, or find out how to manage cookies.
Contact Us