Start | basic neuroanat. | reading | zeki extracts | table of links | List of Refs | Journals

Functional Imagining of the Human Brain: Methods

[page 8 of handout]
1. PET Scans (Positron Emission Tomography)
Measures metabolism of oxygen or sugar. Takes half a minute or more. Only accurate to a centimetre or two. Requires ingestion of radioactive substances by subjects (e.g. carbon dioxide, water). There is a need to limit annual exposure to radiation.

2. SPECT (Single Photon emission Computed Tomography)
Measures blood flow, takes seconds, involves ingestion of special chemicals as radioactive substances. Has similar disadvantages to PET but is cheaper. Clinical uses.

3. (MRI) Magnetic Resonance Imaging
Producing anatomical, rather than functional information, is spatially accurate to millimetres rather than centimetres (the methods above: PET and SPECT). Requires exposure to high radio frequencies and to very strong magnetic fields. (About the strength of magnets used to pick up cars in junk yards.)

4. fMRI (Functional Magnetic Resonance Imaging)
Measures blood flow. Can be used without ingestion of any chemicals, or with ingestion of contrast agents that are not radioactive. Needs strong magnetic fields like MRI. (BOLD= “Blood oxygen level dependent”)

5. MEG/MSI (Magnetoencephalography/ Magnetic Source Imaging)
Measures the very small magnetic fields produced by the electrical activity of neurons. The spatial accuracy of the localization is about 2mm and it is claimed that the temporal resolution goes down to a single millisecond. Like the others, depends a lot on computational and statistical analyses of the raw data. Is non-invasive.


6. Direct (Electrical) Stimulation
of motor cortex of brain of animals, and direct recording from sensory cortex in response to peripheral stimulation (Grunbaum and Sherrington, 1908; Hubel and Weisel, 1977; Zeki, 1978).

7. Direct stimulation of the brain of patients
prepared for brain surgery(e.g. Penfield and Roberts, 1959: still used, sometimes to check on the imaging methods — see Lee et al., 2000; Kreiman, Koch & Fried, 2000; Blanke et al., 2002)

8. Transcranial magnetic stimulation (TMS)
of the cortex of human subjects. Large changes in magnetic fields are produced by currents passed through coils placed close to the scalp, which induces localized activites in neurons in the cortex or other structures. [There is a 1998 discussion of TMS in Brain online ]

9. Near Infra-Red Spectroscopy (NIRS)
(Gratton et al, 1995, Gratton et al, 1997). Like MEG and ERP is very non-invasive compared to the others mentioned.

10.Electroencephalography and Event-Related Potentionals
(EEG and ERP). These techniques have been used for many decades, and involve simply measuring brain activitiy via electrical potentials picked up by surface electrodes on the scalp. Originally did not include much information about the localization of brain activity, but greater computational sophistication means that claims can now be made for localized orgins of the electrical changes measured. (e.g. Somer et al, 1997, Duzel et al 1997).

Start | basic neuroanat. | reading | zeki extracts | table of links | List of Refs | Journals

last revised 10-2006