Functional magnetic resonance imaging. Nikolay Avdievich - about new MRI devices and their capabilities Functional magnetic resonance imaging of the brain

TECHNOLOGY

E.I. Kremneva, R.N. Konovalov, M.V. Krotenkova

Scientific Center for Neurology of the Russian Academy of Medical Sciences (Moscow)

Since the 90s. XX century, functional magnetic resonance imaging (fMRI) is one of the leading methods for mapping the functional areas of the brain due to its non-invasiveness, lack of radiation exposure and relatively widespread use. The essence of this technique is to measure hemodynamic changes in response to neuronal activity (BOLD effect). For the success of an fMRI experiment, it is necessary: ​​the availability of appropriate technical support (high-field MRI scanner, special equipment for performing tasks), development of an optimal study design, post-processing of the data obtained. Currently, the technique is used not only for scientific purposes, but also in practical medicine. However, you should always remember some limitations and contraindications, especially when performing fMRI in patients with various pathologies. To properly plan a study and interpret its results, it is necessary to involve various specialists: neuroradiologists, biophysicists, neurologists, psychologists, since fMRI is a multidisciplinary technique.

Keywords: fMRI, BOLD contrast, study design, post-processing

For many centuries, scientists and doctors have been interested in how the human brain functions. With the development of scientific and technological progress, it became possible to lift the veil of this mystery. And the invention and introduction into clinical practice of such a non-invasive method as magnetic resonance imaging (MRI) has become especially valuable. MRI is a relatively young method: the first commercial 1.5 T tomograph began operating only in 1982. However, by 1990, continuous technical improvement of the method made it possible to use it not only to study the structural features of the brain, but also to study its functioning. This article will focus on a technique that allows mapping of various functional areas of the brain - functional magnetic resonance imaging (fMRI).

Basic principles of the fMRI technique_

fMRI is an MRI technique that measures the hemodynamic response (change in blood flow) associated with neuronal activity. It is based on two basic concepts: neurovascular interaction and BOLD contrast.

fMRI does not allow us to see the electrical activity of neurons directly, but does so indirectly, through local changes in blood flow. This is possible due to the phenomenon of neurovascular interaction - a regional change in blood flow in response to the activation of nearby neurons. This effect is achieved through a complex sequence of interconnected reactions occurring in neurons, the surrounding glia (astrocytes) and the endothelium of the vascular wall, since with increased activity, neurons need more oxygen and nutrients brought by the bloodstream. The fMRI technique allows one to directly assess changes in hemodynamics.

This became possible in 1990, when Seiji Ogawa and his colleagues from Bell Laboratories (USA) proposed using BOLD contrast to study brain physiology using MRI. Their discovery marked the beginning of the era

modern functional neuroimaging and formed the basis of most fMRI studies. BOLD contrast (literally - blood-oxygenation-level dependent, depending on the level of blood oxygenation) is the difference in the MR signal in images using gradient sequences depending on the percentage of deoxyhemoglobin. Deoxyhemoglobin has different magnetic properties from surrounding tissues, which during scanning leads to local disturbance of the magnetic field and a decrease in the signal in the gradient echo sequence. When blood flow increases in response to the activation of neurons, deoxyhemoglobin is washed out of the tissues, and it is replaced by oxygenated blood, which has magnetic properties similar to the surrounding tissues. Then the field disturbance decreases and the signal is not suppressed - and we see its local amplification (Fig. 1A).

Thus, summing up all of the above, the general fMRI scheme can be represented as follows: activation of neurons in response to the action of a stimulus and an increase in their metabolic needs leads to a local increase in blood flow, recorded during fMRI in the form of a BOLD signal - the product of neuronal activity and hemodynamic response ( Fig. 1B).

rice. 1: A - schematic illustration of the VOS contrast in the Oda\ga experiment with changes in the percentage of oxygen in the blood of rats; when inhaling ordinary air (21% oxygen), areas of decreased signal are determined in the cortex (in the upper part of the figure), corresponding to vessels with a high content of deoxyhemoglobin; when pure oxygen is inhaled, a homogeneous MR signal from the cerebral cortex is noted (in the lower part of the figure); B - general scheme for generating a WOS signal

Experiment planning

To conduct an fMRI study, you must have a high-field MR tomograph (magnetic field induction value - 1.5 T and above), various equipment for carrying out tasks during scanning (headphones, video glasses, a projector, various remote controls and joysticks for feedback from subjects, etc. .). An important factor is the willingness of the subject to cooperate.

Schematically, the scanning process itself (using the example of visual stimulation) looks like this (Fig. 2): the subject is in the tomograph; through a special system of mirrors fixed above his head, he has access to images displayed through a video projector on the screen. For feedback (if this is implied in the task), the patient presses a button on the remote control. The supply of incentives and monitoring of task completion is carried out using the console in the control room.

The tasks that the subject performs can be different: visual, cognitive, motor, speech, etc., depending on the goals set. There are two main types of presentation of stimuli in a task: in the form of blocks - block design, and in the form of individual isolated stimuli - discrete design (Fig. 3). A combination of both of these options is also possible - a mixed design.

The most widespread, especially for motor tasks, is the block design, when identical stimuli are collected in blocks alternating with each other. An example is the task of squeezing a rubber ball (each squeeze is a separate stimulus) for a certain period of time (on average 20-30 s), alternating with periods of rest of similar duration. This design has the greatest statistical power because it sums the individual BOLD signals. However, it is, as a rule, predictable for patients and does not allow assessing the response to a single stimulus, and therefore is not suitable for some tasks, in particular cognitive ones.

rice. 2: Scheme of the fMRI experiment (based on materials from the resource http://psychology.uwo.ca/fmri4newbies, with modifications)

Block

Discrete (event-related)

A 11 i A D1 iil iitU I I,

rice. 3: Main types of fMRI study designs

Functional magnetic resonance imaging

For this purpose, there is a discrete design, when stimuli are presented in a chaotic order at different intervals of time. For example, a subject with arachnophobia is shown neutral images (flowers, buildings, etc.), among which images of a spider appear from time to time, which makes it possible to assess brain activation in response to unpleasant stimuli. With a block design, this would be difficult: firstly, the subject knows when the block will appear and is already preparing for it in advance, and secondly, if the same stimulus is presented for a long time, the reaction to it is dulled. It is discrete design that can be used in fMRI as a lie detector or in marketing research, when volunteers are shown different versions of a product (its packaging, shape, color) and their unconscious reactions are observed.

So, we chose a task design and carried out scanning. What do we get as a result? First, there is a 4D series of functional data in a gradient echo sequence, which represents multiple repeated scans of the entire brain volume during the task. And secondly, a high-resolution 3D volume of anatomical data: for example, 1 x 1 x 1 mm (Fig. 4). The latter is necessary for accurate mapping of activation zones, since functional data have low spatial resolution.

Data post-processing_

Changes in the MR signal in areas of brain activation under various conditions are only 3-5%, they are elusive to the human eye. Therefore, the obtained functional data are then subjected to statistical analysis: a curve of the dependence of the intensity of the MR signal on time is plotted for each voxel of the image under different conditions - experimental (stimulus delivery) and control. As a result, we obtain a statistical activation map combined with anatomical data.

But before directly carrying out such an analysis, it is necessary to prepare the “raw” data obtained at the end of the scan and reduce the variability of the results that is not related to the experimental task. The preparation algorithm is a multi-step process, and it is very important for understanding possible failures and errors when interpreting the results obtained. Currently, there are various programs -

Ш -.V w<# %>

40 4"r h® Ф W

rice. 4: Series of functional (A) and anatomical (B) data obtained at the end of the scan

New software for preliminary processing of obtained data, produced by both manufacturers of MRI scanners and independent fMRI research laboratories. But, despite the differences in the methods used, their names and data presentation, all stages of preparation come down to a few basic steps.

1. Correction of the subject’s head movement. When performing tasks, this is inevitable, despite the use of various devices for fixing the head (masks, clamps on the head coil, etc.). Even minimal movement can result in large artificial changes in MR signal intensity between successive data volumes, especially if the head movement is associated with the performance of an experimental task. In this case, it is difficult to distinguish “true” BOLD activation from “artificial” activation - arising as a result of the movement of the subject (Fig. 5).

It is generally accepted that the optimal head displacement is no more than 1 mm. In this case, the displacement perpendicular to the scanning plane (the “head - feet” direction) is significantly worse for correct statistical processing of results than the displacement in the scanning plane. At this stage, a rigid-body transformation algorithm is used - a spatial transformation in which only the position and orientation of the object changes, and its dimensions or shape are constant. In practice, the processing looks like this: a referent (usually the first) functional volume of images is selected, and all subsequent functional volumes are mathematically aligned with it, similar to how we align paper sheets in a stack.

2. Core registration of functional and anatomical data.

Differences in the position of the subject's head are minimized. Computer processing and comparison of high-resolution anatomical data and very low-resolution functional data are also carried out to enable subsequent localization of activation zones.

rice. 5: Example of patient's head displacement during scanning while performing a motor paradigm. In the upper part of the figure is a graph of the movement of the subject’s head in three mutually perpendicular planes: the middle curve reflects the patient’s displacement along the z-axis (the “head-toe” direction), and it clearly deviates at the beginning of the movement and at its end. In the lower part are statistical maps of activation of the same subject without motion correction. Typical motion artifacts are identified in the form of half rings along the edge of the brain matter

In addition, differences associated with different scanning modes are minimized (usually for functional data this is the “gradient echo” mode, for anatomical data - T1). Thus, the gradient echo mode can give some stretching of the image along one of the axes compared to high-resolution structural images.

3. Spatial normalization. It is known that the shape and size of the human brain varies significantly. To compare data obtained from different patients, as well as to process the entire group as a whole, mathematical algorithms are used: the so-called affine transformation. In this case, the images of individual regions of the brain are transformed - stretching, compression, stretching, etc. - followed by reduction of structural data to a unified spatial coordinate system.

Currently, the two most common spatial coordinate systems in fMRI are the Thaleras system and the Montreal Neurological Institute system. The first was developed by French neurosurgeon Jean Talairach in 1988 based on post-mortem measurements of the brain of a 60-year-old French woman. Then the coordinates of all anatomical regions of the brain were given relative to the reference line connecting the anterior and posterior commissures. Any brain can be placed in this stereotaxic space, and areas of interest can be described using a three-dimensional coordinate system (x, y, z). The disadvantage of such a system is that it contains data from only one brain. Therefore, the more popular system is developed at the Montreal Neurological Institute (MNI) based on the total calculation of T1 image data from 152 Canadians.

Although in both systems the counting is carried out from the line connecting the anterior and posterior commissures, the coordinates of these systems are not identical, especially as they approach the convexital surfaces of the brain. This must be kept in mind when comparing the results obtained with data from the work of other researchers.

It should be noted that this stage of processing is not used for preoperative mapping of functional activation zones in neurosurgery, since the purpose of fMRI in such a situation is to accurately assess the location of these zones in a particular patient.

4. Smoothing. Spatial normalization is never accurate, so homologous regions, and therefore their activation zones, are not 100% consistent. To achieve spatial overlap of similar activation zones in a group of subjects, improve the signal-to-noise ratio and thus increase the reliability of the data, a Gaussian smoothing function is used. The essence of this stage of processing is to “blur” the activation zones of each subject, as a result of which the areas of their overlap increase during group analysis. Disadvantage: spatial resolution is lost.

Now, finally, we can move directly to statistical analysis, as a result of which we obtain data on the zones of activation in the form of color maps superimposed on the anatomical data. The same data can

Functional magnetic resonance imaging

Statistics: p-va/ues adjusted for search volume

set-level non-lsotroplc adjusted cluster-level voxel-level

R "- - - ---- mm mm mm

^ conected "E ^ uncorrected PFWE-con ^ FDR-con T (U ^ unconected

0.000 80 0.000 0.000 0.000 6.26 6.04 0.000 -27 -24 60

0.000 0.000 6.00 5.81 0.000 -33 -18 69

0.002 46 0.001 0.009 0.000 5.20 5.07 0.000 27 -57 -21

0.123 0.004 4.54 4.45 0.000 18 -51 -18

0.278 6 0.179 0.076 0.003 4.67 4.58 0.000 51 21 -21

0.331 5 0.221 0.081 0.003 4.65 4.56 0.000 -66 -24 27

0.163 9 0.098 0.099 0.003 4.60 4.51 0.000 -48 -75 -27

0.050 17 0.029 0.160 0.005 4.46 4.38 0.000 -21 33 27

0.135 10 0.080 0.223 0.006 4.36 4.28 0.000 3 -75 -33

0.668 1 0.608 0.781 0.024 3.83 3.77 0.000 6 -60 -9

rice. 6: Example of presentation of statistical post-processing results. On the left - activation zones when performing a motor paradigm (raising - lowering the right index finger), combined with volumetric reconstruction of the brain. On the right - statistical data for each activation zone

be presented in digital format indicating the statistical significance of the activation zone, its volume and coordinates in stereotaxic space (Fig. 6).

Applications of fMRI_

In what cases is fMRI performed? First, for purely scientific purposes: this is a study of the functioning of the normal brain and its functional asymmetry. This technique has revived the interest of researchers in mapping brain functions: without resorting to invasive interventions, you can see which areas of the brain are responsible for a particular process. Perhaps the greatest advances have been made in our understanding of higher cognitive processes, including attention, memory, and executive functions. Such studies have made it possible to use fMRI for practical purposes far from medicine and neuroscience (as a lie detector, in marketing research, etc.).

In addition, fMRI is being actively used in practical medicine. Currently, this technique is widely used in clinical practice for preoperative mapping of basic functions (motor, speech) before neurosurgical interventions for space-occupying brain lesions or intractable epilepsy. In the USA, there is even an official document - a practical guide compiled by the American College of Radiology and the American Society of Neuroradiology, which describes the entire procedure in detail.

Researchers are also trying to introduce fMRI into routine clinical practice for a variety of neurological and psychiatric diseases. The main goal of numerous works in this area is to assess changes in the functioning of the brain in response to damage to one or another of its areas - loss and (or) switching of zones, their displacement, etc., as well as dynamic observation of the restructuring of activation zones in response to drug therapy therapy and (or) rehabilitation measures.

Ultimately, fMRI studies conducted on patients of various categories can help determine the prognostic value of various options for functional restructuring of the cortex to restore impaired functions and develop optimal treatment algorithms.

Possible failures of the study_

When planning fMRI, you should always keep in mind various contraindications, limitations and possible

sources of error in the interpretation of data obtained for both healthy volunteers and patients.

These include:

Any factors affecting neurovascular interaction and hemodynamics and, as a result, BOLD contrast; therefore, one should always take into account possible changes in cerebral blood flow, for example, due to occlusions or severe stenoses of the main arteries of the head and neck, or taking vasoactive drugs; There are also known facts of a decrease or even inversion of the BOLD response in some patients with malignant gliomas due to impaired autoregulation;

The patient has contraindications common to any MRI examination (pacemakers, claustrophobia, etc.);

Metal structures in the area of ​​the facial (cerebral) parts of the skull (non-removable dentures, clips, plates, etc.), producing pronounced artifacts in the “gradient echo” mode;

Lack (difficulty) of cooperation on the part of the subject during the task, associated both with his cognitive status and with decreased vision, hearing, etc., as well as with a lack of motivation and proper attention to completing the task;

Pronounced movement of the subject while performing tasks;

Incorrectly planned study design (choice of control task, duration of blocks or the entire study, etc.);

Careful development of tasks, which is especially important for clinical fMRI, as well as when studying a group of people or the same subject over time to be able to compare the resulting activation zones; tasks must be reproducible, that is, the same throughout the entire period of the study and can be completed by all subjects; One possible solution for patients who cannot independently perform movement-related tasks is the use of passive paradigms using various devices to move the limbs;

Incorrect selection of scanning parameters (echo time - TE, repetition time - TR);

Incorrect data post-processing parameters at various stages;

Incorrect interpretation of the obtained statistical data, incorrect mapping of activation zones.

Conclusion

Despite the above limitations, fMRI is an important and versatile modern neuroimaging technique that combines the advantages of high spatial resolution and non-invasiveness with the absence of the need for intravenous contrast.

amplification and exposure to radiation. However, this technique is very complex, and to successfully complete the tasks assigned to a researcher working with fMRI, a multidisciplinary approach is required - involving in the research not only neuroradiologists, but also biophysicists, neurophysiologists, psychologists, speech therapists, clinical doctors, and mathematicians. Only in this case is it possible to use the full potential of fMRI and obtain truly unique results.

Bibliography

1. Ashburner J., Friston K. Multimodal image coregistration and partitioning - a unified framework. Neurolmage 1997; 6 (3): 209-217.

2. Brian N. Pasley, Ralph D. Freeman. Neurovascular coupling. Scholarpedia 2008; 3(3):5340.

3. Chen C.M., HouB.L., Holodny A.I. Effect of age and tumor grade on BOLD functional MR imaging in preoperative assessment of patients with glioma. Radiology 2008; 3: 971-978.

4. Filippi M. fMRI techniques and protocols. Humana press 2009: 25.

5. Friston K.J., Williams S., HowardR. et al. Movement-related effects in fMRI time-series. Magn. Reson. Med. 1996; 35: 346-355.

6. Glover, G.H., Lai S. Self-navigated spiral fMRI: Interleaved versus single-shot. Magn. Reson. Med. 1998; 39: 361-368.

7. Haller S, Bartsch A.J. Pitfalls in fMRI. Eur. Radiol. 2009; 19: 2689-2706.

8. Hsu Y.Y., Chang C.N., Jung S.M. et al. Blood oxygenation level-dependent MRI of cerebral gliomas during breath holding. J. Magn. Reson Imaging 2004; 2: 160-167.

9. Huettel S.A., Song A.W., McCarthy G. Functional magnetic resonance imaging. Sinauer Associates, Inc. 2004: 295-317.

10. Ogawa S., Lee T.M. Magnetic resonance imaging of blood vessels at high fields: In vivo and in vitro measurements and image simulation. Magn. Reson. Med. 1990; 16 (1): 9-18.

Magnetic resonance imaging is indispensable in the diagnosis of many diseases and allows for detailed visualization of internal organs and systems.

The MRI department of the NACFF clinic in Moscow is equipped with a high-field tomograph Siemens MAGNETOM Aera with an open tunnel design. The power of the tomograph is 1.5 Tesla. The equipment allows examination of people weighing up to 200 kg, the width of the apparatus tunnel (aperture) is 70 cm. In our clinic you can do an MRI of the spine, joints, internal organs, including with the introduction of a contrast agent, as well as undergo magnetic resonance imaging of the head brain The cost of diagnostics is affordable, while the value of the results obtained is incredibly high. In total, more than 35 types of magnetic resonance examinations are performed.

After the MRI diagnosis, the doctor conducts a conversation with the patient and issues a disk with the recording. The conclusion is transmitted via email.

Preparation

Most magnetic resonance examinations do not require special preparation. However, for example, for MRI of the abdominal cavity and pelvic organs, it is recommended to refrain from eating and drinking 5 hours before the examination.

Before visiting the magnetic resonance imaging center (on the day of the examination), you must wear comfortable clothing without any metal elements.

Contraindications

Contraindications to magnetic resonance imaging are due to the fact that during the study a powerful magnetic field is generated that can affect electronics and metals. Based on this, an absolute contraindication to MRI is the presence of:

• pacemaker;
• neurostimulator;
• electronic middle ear implant;
• metal clips on vessels;
• insulin pumps

Installed pacemaker, neurostimulator, electronic middle ear implant, metal clips on blood vessels, insulin pumps.

Restrictions on carrying out

If you have large metal structures installed (for example, a joint endoprosthesis), you will need a document about the possibility and safety of performing MRI. This may be a certificate for the implant (usually issued after the operation) or a certificate from the surgeon who performed the intervention. Most of these structures are made of medical grade titanium, which does not interfere with the procedure. But, in any case, before the examination, tell the doctor at the radiology department about the presence of foreign objects in the body - crowns in the oral cavity, piercings, and even tattoos (metal-containing paints could be used in the latter).

The price of magnetic resonance imaging depends on the part of the body being examined and the need for additional procedures (for example, contrast injection). So an MRI of the brain will cost more than a tomography of one hand. Sign up for the study by phone in Moscow: +7 495 266-85-01 or leave a request on the website.

Gives the researcher a lot of information about the anatomical structure of an organ, tissue or other object that comes into view. However, in order to develop a holistic picture of the processes occurring, there is not enough data on functional activity. And for this purpose there is BOLD-functional magnetic resonance imaging (BOLD - blood oxygenation level dependent contrast, or contrast depending on the degree of oxygen saturation of the blood).

BOLD fMRI is one of the most applicable and widely known methods for measuring brain activity. Activation results in increased local blood flow with changes in the relative concentration of oxygenated (oxygen-enriched) and deoxygenated (oxygen-poor) hemoglobin in the local circulation.

Fig.1.Scheme reactions brain blood flow V answer on excitation neurons.

Deoxygenated blood is paramagnetic (a substance that can be magnetized) and will cause MRI signal levels to drop. If there is more oxygenated blood in the brain area, the level of the MRI signal increases. Thus, oxygen in the blood acts as an endogenous contrast agent.

Fig.2.Volume brain blood supply (A) And BOLD-answer fMRI (b) at activation primary motor barkperson. Signal passes V 4 stages. 1 stage – due to activation neurons rises consumptionoxygen, increases quantity deoxygenated blood, BOLD— signal A little decreases (on graphicsNot shown, decrease minor). Vessels are expanding, due to what some decreasesblood supply cerebral fabrics. Stage 2 – long-term increase signal. Potential actions neuronsends, But flow oxygenated blood increases inertially, Maybe due to impactbiochemical markers hypoxia. Stage 3 – long-term decline signal due to normalizationblood supply. 4 stage – post-stimulus recession – called slow restoration originalblood supply

To activate the work of neurons in certain areas of the cortex, there are special activating tasks. Task design usually comes in two types: “block” and “event-related”. Each type assumes the presence of two alternating phases - an active state and a resting state. In clinical fMRI, tasks of the “block” type are more often used. When performing such exercises, the subject alternates so-called ON- (active state) and OFF- (resting state) periods of equal or unequal duration. For example, when identifying the area of ​​the cortex responsible for hand movements, tasks consist of alternating finger movements and periods of inactivity, lasting on average about 20 seconds. The steps are repeated several times to increase the accuracy of the fMRI result. In the case of an event-related task, the subject performs one short action (for example, swallowing or clenching a fist), followed by a period of rest, while actions, in contrast to the block design, alternate unevenly and inconsistently.

In practice, BOLD fMRI is used in preoperative planning of resection (removal) of tumors, diagnosis of vascular malformations, and during operations for severe forms of epilepsy and other brain lesions. During brain surgery, it is important to remove the lesion as accurately as possible, while at the same time avoiding unnecessary damage to adjacent functionally important areas of the brain.

Fig.3.

A – three-dimensional MRI— image head brain. Arrow indicated location motor bark Vprecentral gyrus.

b – map fMRI—activity brain V precentral gyrus at movement hand.

The method is very effective in studying degenerative diseases, such as Alzheimer's and Parkinson's diseases, especially in the early stages. It does not involve the use of ionizing radiation or radiopaque agents, and it is non-invasive. Therefore, it can be considered quite safe for patients who require long-term and regular fMRI examinations. fMRI can be used to study the mechanisms of formation of epileptic seizures and allows one to avoid removal of the functional cortex in patients with intractable frontal lobe epilepsy. Monitoring brain recovery after strokes, studying the effects of medications or other therapies, monitoring and monitoring the treatment of psychiatric diseases - this is not a complete list of possible applications of fMRI. In addition, there is also resting fMRI, in which complex data processing allows us to see the brain networks functioning at rest.

Sources:

1. How well do we understand the neural origins of the fMRI BOLD signal? Owen J. Arthur, Simon Boniface. TRENDS in Neurosciences Vol.25 No.1 January 2002
2. The physics of functional magnetic resonance imaging (fMRI) R. B. Buxton. Rep. Prog. Phys. 76 (2013)
3. Application of functional magnetic resonance imaging in the clinic. Scientific review. Belyaev A., Peck Kung K., Brennan N., Kholodny A. Russian electronic journal of radiology. Volume 4 No. 1 2014
4. Brain, cognition, mind: An introduction to cognitive neuroscience. Part 2 . B. Baars, N. Gage. M.: Binom. 2014 pp. 353-360.

Text: Daria Prokudina

Changes in blood flow activity are recorded by functional magnetic resonance imaging (fMRI). The method is used to determine the localization of arteries, to assess the microcirculation of the centers of vision, speech, movement, and the cortex of some other functional centers. A feature of mapping is that the patient is asked to perform certain tasks that increase the activity of the desired brain center (read, write, talk, move legs).

At the final stage, the software generates an image by summing conventional layer-by-layer tomograms and images of the brain with functional load. The complex of information is displayed by a three-dimensional model. Spatial modeling allows specialists to study the object in detail.

Together with MRI spectroscopy, the study reveals all the metabolic features of pathological formations.

Principles of functional MRI of the brain

Magnetic resonance imaging is based on recording the altered radio frequency of hydrogen atoms in liquid media after exposure to a strong magnetic field. Classic scanning shows soft tissue components. To improve the visibility of blood vessels, intravenous contrast with the paramagnetic gadolinium is performed.

Functional MRI records the activity of individual areas of the cerebral cortex by taking into account the magnetic effect of hemoglobin. After releasing oxygen molecules to tissues, the substance becomes paramagnetic, the radio frequency of which is picked up by the device’s sensors. The more intense the blood supply to the brain parenchyma, the better the signal.

Tissue magnetization is further enhanced by glucose oxidation. The substance is necessary to ensure the processes of tissue respiration of neurons. Changes in magnetic induction are recorded by the device’s sensors and processed by a software application. High-field devices create high-quality resolution. The tomogram shows a detailed image of parts with a diameter of up to 0.5 mm in diameter.

Functional MRI studies record signals not only from the basal ganglia, cingulate cortex, and thalamus, but also from malignant tumors. Neoplasms have their own vascular network, through which glucose and hemoglobin enter the formation. Signal tracking allows you to study the contours, diameter, and depth of tumor penetration into the white or gray matter.

Functional diagnostics of MRI of the brain requires the qualifications of a radiology doctor. Different zones of the cortex are characterized by different microcirculation. Saturation with hemoglobin and glucose affects the signal quality. The structure of the oxygen molecule and the presence of alternative substitute atoms should be taken into account.

A strong magnetic field increases the half-life of oxygen. The effect works when the device power is more than 1.5 Tesla. Weaker installations cannot fail to study the functional activity of the brain.

It is better to determine the metabolic intensity of the blood supply to the tumor using high-field equipment with a power of 3 Tesla. High resolution will allow you to register a small lesion.

The effectiveness of the signal is scientifically called the “hemodynamic response.” The term is used to describe the speed of neural processes with an interval of 1-2 seconds. The blood supply to tissues is not always sufficient for functional studies. The quality of the result is improved by additional administration of glucose. After stimulation, peak saturation occurs after 5 seconds, when scanning is carried out.

Technical features of a functional MRI study of the brain

Functional MRI diagnostics are based on an increase in neuronal activity after stimulation of brain activity by a person performing a specific task. An external stimulus causes stimulation of sensory or motor activity of a specific center.

To track the area, a gradient echo mode is enabled based on a pulsed echo-planar sequence.

Analysis of the active zone signal on MRI is done quickly. Registration of one tomogram is performed at an interval of 100 ms. Diagnostics are performed after stimulation and during the rest period. The software uses tomograms to calculate foci of neuronal activity, overlaying areas of amplified signal on a three-dimensional model of the brain at rest.

For treating physicians, this type of MRI provides information about pathophysiological processes that cannot be tracked by other diagnostic methods. The study of cognitive functions is necessary for neuropsychologists to differentiate mental and psychological diseases. The study helps to verify epileptic foci.

The final mapping map does not only show areas of increased functional stimulation. The images visualize zones of sensorimotor and auditory speech activity around the pathological focus.

Mapping the location of brain canals is called tractography. The functional significance of the location of the optic pyramidal tract before planning surgical intervention allows neurosurgeons to correctly plan the location of incisions.

What does fMRI show?

High-field MRI with functional tests is prescribed according to indications when it is necessary to study the pathophysiological basis of the functioning of the motor, sensory, visual, and auditory areas of the cerebral cortex. Neuropsychologists use research in patients with disorders of speech, attention, memory, and cognitive functions.

Using fMRI, a number of diseases are detected at the initial stage - Alzheimer's, Parkinson's, demyelination in multiple sclerosis.

Functional diagnostics in different medical centers is performed using different installations. The diagnostician knows what an MRI of the brain shows. A consultation with a specialist is required before the examination.

High quality results are achieved by scanning with a strong magnetic field. Before choosing a medical center, we recommend that you find out the type of device installed. The qualifications of a specialist who must have knowledge about the functional, structural components of the brain are important.

The future of functional MRI diagnostics in medicine

Functional studies have recently been introduced into practical medicine. The capabilities of the method are not used enough.

Scientists are developing techniques for visualizing dreams and reading minds using functional MRI. It is proposed to use tomography to develop a method of communication with paralyzed people.

• Neuronal excitability;
• Mental activity;
• Degree of saturation of the cerebral cortex with oxygen and glucose;
• The amount of deoxylated hemoglobin in the capillaries;
• Areas of blood flow expansion;
• Level of oxyhemoglobin in blood vessels.

Advantages of the study:

1. High-quality temporary picture;
2. Spatial resolution higher than 3 mm;
3. Possibility of studying the brain before and after stimulation;
4. Harmlessness (when compared with PET);
5. Lack of invasiveness.

The widespread use of functional MRI of the brain is limited by the high cost of equipment, each single examination, the impossibility of directly measuring neuronal activity, and cannot be done on patients with metallic inclusions in the body (vascular clips, ear implants).

Registration of the functional metabolism of the cerebral cortex has great diagnostic value, but is not an accurate indicator for the dynamic assessment of changes in the brain during treatment, after surgery.

Magnetic resonance imaging (MRI) is a method of obtaining tomographic medical images for non-invasive examination of internal organs and tissues, based on the phenomenon of nuclear magnetic resonance (NMR). The technology appeared several decades ago, and today you can undergo an examination using such a device in many modern clinics. However, scientists continue to work to improve the accuracy of the technology and develop new, more efficient systems. , senior researcher at the Max Planck Institute in Tübingen (Germany), is one of the leading specialists who develops new sensors for experimental ultra-high-field MRI. The day before, he conducted a special course in the master’s program “ RF Systems and Devices» ITMO University, and in an interview with ITMO.NEWS he spoke about his work and how new research in the field of MRI will help make the diagnosis of diseases more effective.

For the last few years you have been working in the High-field Magnetic Resonance Department of the Max Planck Institute. Please tell us what your current research is focused on?

I am developing new radio frequency (RF) sensors for MRI. What MRI is is probably now known to most people, since over the past 40 years, since this technology was developed, it has managed to come to a huge number of clinics and become an indispensable diagnostic tool. But even today people are working to improve this technology by developing new MRI systems.

An MRI is primarily a huge cylindrical magnet into which a patient or volunteer is placed to produce a three-dimensional image. But before this image is created, a huge amount of research needs to be done. It is led by engineers, physicists, doctors and other specialists. I am one of the links in this chain and am engaged in research at the intersection of physics and engineering. More specifically, we are developing sensors for ultra-high-field experimental MRI, which is used at the stage of excitation, reception and processing of the signal obtained as a result of the physical effect of NMR.

One of the main directions is the development of new experimental ultra-high-field MRI systems, that is, using a higher constant magnetic field, which allows improving image resolution or reducing scanning time, which is very important for many clinical studies and diagnostics.

Conventional clinical tomographs use constant fields up to 3 T, but experimental tomographs with magnetic fields of 7 T and higher are now appearing. It is customary to call tomographs with a magnetic field of 7 T and higher ultra-high-field. There are already about a hundred tomographs with a field of 7 T in the world, but developments are underway to further increase the magnetic field. For example, at the Max Planck Institute in Tübingen we have a 9.4 T MRI machine.

But even with the transition from 7 to 9.4 T, many technical problems arise that require serious scientific and technical developments, including the calculation and design of sensors for a new generation of MRI.

What are these difficulties?

An increase in the constant magnetic field results in a corresponding increase in the frequency of the RF sensors. For example, clinical 3 T tomographs use sensors with a resonant frequency of about 120 MHz, while a 7 T tomograph requires sensors with a frequency of 300 MHz. This primarily leads to a shortening of the wavelength of the RF field in human tissue. If the frequency of 120 MHz corresponds to approximately a wavelength of 35-40 centimeters, then at a frequency of 300 MHz it decreases to a value of about 15 cm, which is much smaller than the size of the human body.

As a result of this effect, the sensitivity of RF sensors can be greatly distorted when studying large objects (longer than the wavelength). This leads to difficulties in interpreting images and diagnosing clinical diseases and pathologies. In a field of 9.4 T, which corresponds to a sensor frequency of 400 MHz, all these problems become even more critical.

That is, such pictures become virtually unreadable?

I wouldn't say that. More precisely, in some cases this makes them difficult to interpret. However, there are groups developing techniques for obtaining MR images of the entire human body. However, the tasks of our group are primarily focused on brain research.

What exactly are the opportunities for medicine that research in the field of ultra-high-field MRI opens up?

As you know, during MRI a person must lie still: if you start to move during measurements, the picture will turn out distorted. At the same time, some MRI techniques can take up to an hour, and it is clear that it is difficult not to move during this entire time. The increased sensitivity of ultra-high-field tomographs makes it possible to obtain images not only with higher resolution, but also much faster. This is primarily important when studying children and elderly patients.

It is also impossible not to mention the possibilities for magnetic resonance spectroscopy ( MRS, a method that allows you to determine biochemical changes in tissues in various diseases based on the concentration of certain metabolites - editor's note ).

In MRI, the main signal source is the hydrogen atoms of water molecules. But besides this, there are other hydrogen atoms found in other molecules that are important for the functioning of the human body. Examples include various metabolites, neurotransmitters, etc. Measuring the spatial distribution of these substances using MRS can provide useful information for studying pathologies associated with metabolic disorders in the human body. Often the sensitivity of clinical tomographs is insufficient to study them due to their low concentration and, as a consequence, lower signal.

In addition to this, it is possible to observe the NMR signal not only from hydrogen atoms, but also from other magnetic atoms, which are also very important for diagnosing diseases and medical research. However, firstly, their NMR signal is much weaker due to the lower gyromagnetic ratio and, secondly, their natural content in the human body is much less than hydrogen atoms. The increased sensitivity of ultra-high-field MRI is extremely important for MRS.

Another important area of ​​MRI techniques for which increased sensitivity is critical is functional MRI, an important technique for cognitive studies of the human brain.

So far, the vast majority of clinics in the world do not have high-field tomographs. What are the prospects that 7 T and later 9 T tomographs will be able to be used in routine diagnostics?

In order for a tomograph to come to the clinic, it must be certified, checked for safety conditions, and appropriate documentation must be drawn up. This is a rather complicated and lengthy procedure. So far there is only one company in the world that has begun to certify not only the sensors we make, but also the device itself. This is Siemens.

There are 7 T tomographs, but there are not many of them, and they cannot yet be called completely clinical. What I called is a preclinical option, but this device has already been certified, that is, it can potentially be used in clinics.

Predicting when 9.4 T tomographs will appear in clinics is even more difficult. The main problem here is the possible local heating of tissue by the RF field of the sensor due to the strong reduction in wavelength. One of the important areas of engineering research on ultra-high-field MRI is detailed numerical modeling of this effect to ensure patient safety. Despite the fact that such studies are carried out within scientific institutions, the transition to clinical practice requires additional research.

How is cooperation currently developing between the Max Planck Institute and ITMO University? What joint results have you already achieved?

The work is progressing very well. Now he is working with us, a graduate student from ITMO University. We recently published an article in a leading journal on technical developments in MRI. In this work, we experimentally validated previous theoretical studies to improve the sensitivity of ultra-high-field RF sensors by using modified and optimized dipole antennas. The result of this work, in my opinion, turned out to be very promising.

Now we are also working on several more articles that are devoted to the use of similar methods, but for other tasks. And recently Georgy received a grant to travel to Germany. Next month he comes to us for six months, and we will continue to work together on further development of sensors for MRI.

This week you conducted a special course in the master's program “Radio Frequency Systems and Devices”. What are the main topics you covered?

The course covers various technical aspects of developing MRI sensors. There are many intricacies that need to be known in this area, so I have presented a number of basic techniques that are used to design and manufacture these sensors. In addition, I presented a lecture on my latest developments. In total, the course includes eight lectures of two academic hours, which are designed for four days. There is also a demonstration at the end to help explain these techniques more clearly.

Master's students are currently in the process of choosing their future direction, so I think this course will give them additional information to evaluate their prospects.

And if we talk in general about education in the field of MRI technologies, what knowledge and skills, in your opinion, are primarily required from such specialists today?

Despite the fact that our field has now become very popular and promising for use in clinical diagnostics, there are currently no engineering courses that would train highly specialized specialists involved in the manufacture of MRI coils. A gap has formed. And I think that together we can just fill it.

Elena Menshikova

Editorial office of the news portal