Espectrometría de Fluorescencia 38 

Electric Impedance Tomography as a neuro-imaging method will be discussed separ- ately in 1.7. EIT can image both the neuronal activity directly as well as metabolic changes associated with neuronal activity, but does not provide a structural image of the brain.

Figure 1.4.: Saltatory conduction of the action potential on an axon showing the ’jumping’ depolarization along the axon (source: http://en.wikipedia.org).

1.4.1. Structural imaging methods

The currently most commonly used structural neuro-imaging methods are CT and MRI. CT uses x-rays and hence has the disadvantage that ionizing radiation is needed for a scan. It has the advantage of a very accurate imaging of the bony structures, blood and a reasonable contrast in the soft tissues especially when con- trast agents are used (Kretschmann and Weinrich [2004]). It is furthermore a very fast technique and it can usually be performed with monitoring equipment for the patient in place which makes an invaluable clinical tool for strokes or imaging in acute medical settings. MRI is an imaging method that makes use of the nuclear magnetic resonance of atoms and can provide good contrast between soft tissues. It is the best method to obtain a structural image of the brain and can be further enhanced with contrast agents (Kretschmann and Weinrich [2004]). It has the ad- vantage of not using any ionizing radiation, but is expensive and the scans take much longer than CT scans. Also, most standard monitoring equipment can not be brought into the MRI scanner which makes a MRI scan of intensive care patients difficult.

1.4.2. Imaging the metabolic changes associated with neuronal activity

Several imaging methods use the physiological increase in blood flow and metabolism caused by neuronal activity. Functional MRI, PET, functional ultrasound, and most optical imaging methods all make use of either changes in blood flow directly or their oxygen and glucose contents. The currently most commonly used functional imaging method is fMRI, in which the changes in blood supply of active areas are

used for mapping. In fMRI the different effect of oxygenated and deoxygenated haemoglobin in the blood on a magnetic field is used. If an area in the brain is activated its metabolic demands rise and more blood is directed to this area. The increase in blood flow leads to a higher oxygen content in active vs inactive areas of the brain, this effect is called the blood-oxygenation level dependent (BOLD) response (Logothetis [2008]). fMRI has a good spatial resolution of about 2mm (Hopfinger et al. [2000]), but it’s temporal resolution is intrinsically limited to the time it takes for the blood flow to respond to neural activity which makes it not useful to image the spread of neuronal activity. PET imaging of neuronal activity also makes use of the metabolic demands of activated neuronal tissue. To image neural activity a radionuclide, such as oxygen-15 or fluorodeoxyglucose, is injected and its emitted positrons are indirectly measured (Wieler et al. [1986]). Often a CT scan is performed on the same patient in the same machine to aid 3D imaging by co-registering the images. Ultrasound is another method that has recently been used for functional neuroimaging based on the changes in bloodflow for research purposes. To image with ultrasound (US) high frequency pulses are sent into the tissue and their echos are recorded. The pulses are scattered by the different tissues and their echoes can be used for image reconstruction. Moving blood can specifically imaged by the characteristic change it causes to the emitted pulses in the tissue by making use of the Doppler effect. New developments in ultrasound have allowed researchers to record changes in local blood flow in response to neuronal activity (Macé et al. [2011], van Raaij et al. [2011]). Some have enhanced this effect with an injectable contrast agent (van Raaij et al. [2012]). Further to the methods discussed, there are several optical imaging methods that have been applied to image neuronal function, mainly in research settings. Diffuse optical imaging uses the different spectra of oxyhaemoglobin and deoxyhaemoglobin while intrinsic optical imaging measures the absorption spectra of blood (Gibson et al. [2005], Franceschini et al. [2003]). Both of these methods have a good temporal resolutions but can only be used to image on directly underlying cortex, which limits its clinical usefulness.

1.4.3. Imaging the neuronal activity directly

EEG and MEG record a surface map of the electric and electromagnetic fields that neural activity produces, respectively. Both have a temporal resolution of milliseconds and both can be used for source modelling (Barkley and Baumgartner [2003], Jousmäki [2000]). However the inverse solution of EEG is not unique and no travelling activity can be imaged unless it is occurring directly on the cortical surface on which EEG electrodes are placed. Unfortunately, no 3D imaging of the travelling neural activity is possible with either EEG or MEG. Another way of imaging the neuronal activity directly are voltage sensitive dyes or calcium sensitive dyes, which can be used to image changes in neuronal cell wall due to activity or change in intracellular calcium concentration, respectively. These methods are used for research purposes, they provide excellent temporal and spatial resolution for surface recordings, however they require exposed brain and they are neurotoxic

which limits their clinical usefulness (Hillman [2007]).

1.5. Introduction to Electric Impedance Tomography