Neuroimiging Flashcards
4) Discuss methodological limitations and barriers to
translational implementation
1) Describe the physiological basis of sMRI
There are different versions of sMRI but most of them use signals from water in the brain to extract information about neuroanatomy.
1) Water
- Water makes up two thirds of the human body and consists of 2 types of atoms: oxygen & hydrogen. Within an atom there are protons (+), electrons (-) and neutrons (neutrally charged).
- The MRI scanner contains a magnet that produces radiofrequency waves.
- The energy in these waves is absorbed by protons within hydrogen atoms in water, which begin to move (oscillate) between 2 points.
- After a short period protons return to their original state by remitting this energy in the same radiofrequency (a process called resonance)
2) Image
- The energy remitted by the protons is measured by the scanner in order to infer the distribution of water across the brain.
- As different brain tissues have different water concentrations, this information can be used to build a 3D image of the brain.
- In psychiatric and neurological disorders, the molecular environment of water is altered and this makes sMRI a useful clinical tool.
What is Voxel-based morphometry (VBM)?
It is a widely used analytical technique for the analysis of sMRI data.
It allows detection of significant differences in gray or white matter:
- between 2+ groups of interest (e.g. patient group vs control group)
- as a function of one or more variables of interest (e.g. years of disease, test-score, etc)
What are the steps in VBM?
- sMRI scans acquisition
- Pre-processing (normalization, segmentation, smoothing of the images)
- Statistical analysis of the data
Acquisition of the data, facts and issues
- It takes 8-12 minutes to acquire an image.
- Spatial resolution can be as small as 1.5mm.
- Within the same investigation images must be acquired using the same scanner and imaging parameters (sequence, number of slices, etc).
- Subjects must keep as still as possible to minimize movement artefacts in the images.
Pre-processing steps in more detail (3)
1) Normalization: The MRI images are “registered” or “mapped” to a standard template in a three-dimensional stereotactic space.
2) Segmentation: The normalized MRI images are then segmented into (i) gray matter; (ii) white matter (iii) cerebrospinal fluid.
3) Smoothing: The segmented gray and white matter images are “smoothed”, resulting in a loss of spatial resolution.
Applications of sMRI in FEP (4)
1) Excluding the presence of neoplasm or other neurological disorders that might be causing psychosis-like symptoms. (e.g. tumour)
- At present this is the most common clinical application of sMRI in early psychosis.
- However individuals presenting with psychosis-like symptoms due to other problems are fairly rare (about 2%).
2) Development of “diagnostic” biomarkers.
- UHR vs HC? FEP vs HC?
3) Development of “prognostic” biomarkers.
UHR-t vs UHR-nt? FEP with multiple relapses vs none?
4) Development of “predictive” biomarkers.
UHR/FEP who benefit from a specific treatment and those who do not?
Issues with using sMRI as biomarkers
- Brain has been proven to have plasticity qualities (depending on a job e.g. London cab driver, number or languages spoken and their proficiency, stressful life events)
= the interpretation of brain differences is more complicated because we cannot be sure which effect are we observing
What are physiological bases for fMRI and PET?
1) Chemistry
- The presentation of a stimulus increases neural activity in specific brain regions.
- This activity elicits (i) electrical signals from nerves to arterioles and (ii) the synthesis of Nitric Oxide.
- Both (i) electrical signals and (ii) Nitric Oxide ensure muscle relaxation around microvessels and mediate the increase in cerebral blood flow which occurs within few hundred milliseconds.
2) Oxygen
- Neuronal activity also elicits an increase in the metabolic rate of oxygen consumption.
- The increase in the metabolic rate of oxygen consumption is much smaller than the increase in local blood flow.
- Because of this imbalance, local deoxyhemoglobin concentration decreases during brain activation.
3) The increase in blood flow is the basis for measuring brain activation with PET.
The decrease in local deoxyhemoglobin concentration is the basis for measuring
brain activation with fMRI.->BOLD signal (blood-oxygen-level-dependent )
How do PET and fMRI compare? (7)
PET vs fMRI :
1. Measures Cerebral Blood Flow vs Measures Deoxyhemoglobin Concentration
2. Radiation (12 measurements) vs No Radiation (unlimited measurements)
3. Temporal resolution = 30 seconds vs Temporal resolution = 2-3 seconds
4. Spatial Resolution: 2-3 mm vs Spatial resolution = 2-3 mm
5. Trials can only be Blocked vs Trials can be Blocked or Randomized
6. Continuous sensitivity in space vs Non-Continuous sensitivity in space (less sensitive in some brain regions)
7. More directly related to neural activity vs Less directly related to neural activity
What is “functional localization” or “functional specialisation”?
- Most functional imaging studies are based on the principle that different regions of the brain serve different functions.
- Studies based on this principle compare responses across different conditions (e.g. reading > rest) to identify areas “specialised” for certain stimuli or tasks.
- There is a range of software (e.g. SPM, XBAM, AFNI, etc) that can be used to test for functional localization.
What is “functional localization” or “functional specialisation”?
- Most functional imaging studies are based on the principle that different regions of the brain serve different functions.
- Studies based on this principle compare responses across different conditions (e.g. reading > rest) to identify areas “specialised” for certain stimuli or tasks.
- There is a range of software (e.g. SPM, XBAM, AFNI, etc) that can be used to test for functional localization.
What is functional integration?
- Studies that are based on the principle of functional integration rely on the fact that different regions communicate with each other within the brain.
- For instance, the neural correlates of reading in the temporal and frontal cortex can be identified using written words because visual regions interact with reading-specific temporal and frontal regions.
- Thus, alternative approaches have been developed to examine the dynamic interaction between regions.
- These alternative approaches consider the brain as a network of regions rather than a modular system.
What is functional connectivity?
It refers to the temporal correlations between neuronal activity in different areas
Can be estimated simply by measuring the correlation in activity between 2 or more areas over time
What is effective connectivity?
It refer to the dynamic influence that one brain region exerts on another
Can be computed using more complex methods such as Structural Equation Modelling (SEM) and Dynamic Causal Modelling (DCM).
