1. BOLD fMRI BIAC Graduate fMRI Course October 1, 2003

2. Why do we need to know physics/physiology of fMRI? To understand the implications of our results –Interpreting activation extent, timing, etc. –Determining the strength of our conclusions –Exploring new and unexpected findings To understand limitations of our method –Choosing appropriate experimental design –Combining information across techniques to overcome limitations To take advantage of new developments –Evaluating others’ approaches to problems –Employing new pulse sequences or protocols

3. Developments allowing functional MRI Echoplanar imaging methods –Proposed by Mansfield in 1977 Ready availability of high-field scanners –Technological developments –Clinical applicability  insurance reimbursement  clinical prevalence Discovery of BOLD contrast mechanism

4. Contrast Agents Defined: Substances that alter magnetic susceptibility of tissue or blood, leading to changes in MR signal –Affects local magnetic homogeneity: decrease in T2* Two types –Exogenous: Externally applied, non-biological compounds (e.g., Gd-DTPA) –Endogenous: Internally generated biological compound (e.g., dHb)

5. External Contrast Agents Most common are Gadolinium-based compounds introduced into bloodstream –Very large magnetic moments Create field gradients within/around vessels –What type of contrast would this generate? Large signal changes: 30-50% –Delay in activation change until agent bolus passes through MR imaging volume –Width of activation change depends on delivery of bolus and vascular filtering –Degree of signal change depends on total blood volume of area

6. Belliveau et al., 1990 CBV Maps (+24%) Slice Location NMR intensity change (CBV)

7. Common Contrast Agents Compound Longitudinal Relaxivity Transverse Relaxivity Magnetic Susceptibility GdCl2111 MnCl20.963.830.51 GdDTPA0.520.51 DyDTPA0.030.041.78 GDTPA albumin1.6-- Iron oxide particle (3nm)0.410.6340.7 Iron oxide particle 253nm)4.415.5148

8. Hemodynamic Measures of Blood Flow

9. Does blood flow serve function? “These facts seem to us to indicate the existence of an automatic mechanism by which the blood supply of any part of the cerebral tissue is varied in accordance with the activity of the chemical changes which underlie the functional action of that part. Bearing in mind that strong evidence exists of localisation of function in the brain, we are of opinion that an automatic mechanism, of the kind just referred to, is well fitted to provide for a local variation of the blood supply in accordance with local variations of the functional activity.” - Roy and Sherrington (1890)

10. Apparatus of Angelo Mosso, as described by James (1890)

11. PET Imaging

12. PET Studies of Brain Function Injection of radioactive tracer –Measure presence of tracer during performance of task –Quantization of activity during single conditions Advantages –Conceptually simple –Allows functional measurement Disadvantages –Invasive contrast mechanism –Limited repeatability –Very low temporal resolution Image from M. Raichle

13. History of BOLD fMRI

15. Blood Deoxygenation affects T2 Recovery Increasing Blood Oxygenation Decreasing Relaxation Time T2 T1 Thulborn et al., 1982

16. Ogawa et al., 1990a Subjects: 1) Mice and Rats, 2) Test tubes Equipment: High-field MR (7+ T) Results 1: –Contrast on gradient-echo images influenced by proportion of oxygen in breathing gas –Increasing oxygen content  reduced contrast –No vascular contrast seen on spin-echo images Results 2: –Oxygenated blood in tube leads to little signal change, either on spin- or gradient-echo images –Deoxygenated blood leads to large susceptibility effects on gradient-echo images

17. Ogawa et al., 1990b 100% O 2 90% O 2, 10% CO 2 Breathing a mix including CO2 results in increased blood flow, in turn increasing blood oxygenation. There is no increased metabolic load (no task). Therefore, BOLD contrast is reduced. Under anesthesia, rats breathing pure oxygen have some BOLD contrast (black lines).

18. Gradient Echo Ogawa 1990b Oxyhemoglobin Spin Echo Deoxyhemoglobin

19. Kwong et al., 1992  VISUAL   MOTOR 

20. Ogawa et al., 1992 High-field (4T) in humans Patterned visual stimulation at 10 Hz Gradient-echo (GRE) pulse sequence used –Surface coil recorded Significant image intensity changes in visual cortex Image signal intensity changed with TE change –What form of contrast?

21. Blamire et al., 1992 This was the first event-related fMRI study. It used both blocks and pulses of visual stimulation. Hemodynamic response to long stimulus durations. Hemodynamic response to short stimulus durations. Gray Matter White matter Outside Head

22. BOLD Endogenous Contrast Blood Oxyenation Level Dependent Contrast –Deoxyhemoglobin is paramagnetic, oxyhemoglobin is less so. –Magnetic susceptibility of blood increases linearly with increasing oxygenation Oxygen is extracted during passage through capillary bed –Arteries are fully oxygenated –Venous (and capillary) blood has increased proportion of deoxyhemoglobin –Difference between oxy and deoxy states is greater for veins  BOLD sensitive to venous changes

23. from Mosley & Glover, 1995

24. T2 Decay MR Signal T1 Recovery MR Signal 50 ms 1 s T2* and T2 Contrast

25. Discarded and Deleted Acquisitions (DISDAQs) Images (TRs) Raw MR Signal (arbitrary units)

26. Relation of BOLD Activity to Neuronal Activity

27. 1. Assumptions fMRI response varies with pooled neuronal activity in a brain region –Behavior/cognitive ability determined by pooled activity Alternatively, if single neurons governed behavior, fMRI activation may be epiphenomenal

28. BOLD response reflects pooled local field potential activity (Logothetis et al, 2001)

29. Calcarine Sulci Fusiform Gyri fMRI Hemodynamic Response 1500ms 500ms 100ms

30. * Calcarine Fusiform 500ms 100ms 1500ms

31. 2. Measuring Deoxyhemoglobin fMRI measurements are of amount of deoxyhemoglobin per voxel We assume that relative oxygenation changes with amount of deoxygenated hemoglobin

32. 3. Different changes in CBF & CMRO 2 Cerebral Blood Flow (CBF) and Cerebral Metabolic Rate of Oxygen (CMRO 2 ) are coupled under baseline conditions –PET measures CBF well, CMRO 2 poorly –fMRI measures CMRO 2 well, CBF poorly CBF about.5 ml/g/min under baseline conditions –Increases to max of about.7-.8 ml/g/min under activation conditions CMRO 2 only increases slightly with activation –Note: A large CBF change may be needed to support a small change in CMRO 2

34. The Hemodynamic Response

35. Impulse-Response Systems Impulse: single event that evokes changes in a system –Assumed to be of infinitely short duration Response: Resulting change in system = Impulses Convolution Response Output

36. Basic Form of Hemodynamic Response Baseline Rise Peak Undershoot Sustained Response Initial Dip

37. Baseline Period Why include a baseline period in epoch? –Corrects for scanner drift across time

38. Initial Dip (Hypo-oxic Phase) Transient increase in oxygen consumption, before change in blood flow –Menon et al., 1995; Hu, et al., 1997 Shown by optical imaging studies –Malonek & Grinvald, 1996 Smaller amplitude than main BOLD signal –10% of peak amplitude (e.g., 0.1% signal change) Potentially more spatially specific –Oxygen utilization may be more closely associated with neuronal activity than perfusion response

39. Early Evidence for the Initial Dip C AB Menon et al, 1995

40. Rise (Hyperoxic Phase) Results from vasodilation of arterioles, resulting in a large increase in cerebral blood flow Inflection point can be used to index onset of processing

41. Peak – Overshoot Over-compensatory response –More pronounced in BOLD signal measures than flow measures Overshoot found in blocked designs with extended intervals –Signal saturates after ~10s of stimulation

42. Sustained Response Blocked design analyses rest upon presence of sustained response –Comparison of sustained activity vs. baseline –Statistically simple, powerful Problems –Difficulty in identifying magnitude of activation –Little ability to describe form of hemodynamic response –May require detrending of raw time course

43. Undershoot Cerebral blood flow more locked to stimuli than cerebral blood volume –Increased blood volume with baseline flow leads to decrease in MR signal More frequently observed for longer- duration stimuli (>10s) –Short duration stimuli may not evidence –May remain for 10s of seconds

44. Issues in HDR Analysis Delay in the HDR –Hemodynamic activity lags neuronal activity Amplitude of the HDR Variability in the HDR HDR as a relative measure

45. The Hemodynamic Response Lags Neural Activity Experimental Design Convolving HDR Time-shifted Epochs Introduction of Gaps

46. Percent Signal Change Peak / mean(baseline) Often used as a basic measure of “amount of processing” Amplitude variable across subjects, age groups, etc. 500 505 200 205 1%

47. Amplitude of the HDR Peak signal change dependent on: –Brain region –Task parameters –Voxel size –Field Strength Kwong et al, 1992

48. Variability in the Hemodynamic Response Across Subjects Across Sessions in a Single Subject Across Brain Regions Across Stimuli

49. Relative vs. Absolute Measures fMRI provides relative change over time –Signal measured in “arbitrary MR units” –Percent signal change over baseline PET provides absolute signal –Measures biological quantity in real units CBF: cerebral blood flow CMRGlc: Cerebral Metabolic Rate of Glucose CMRO 2 : Cerebral Metabolic Rate of Oxygen CBV: Cerebral Blood Volume

50. Detection vs. Estimation Detection power –The ability of a design to determine whether or not an area is active Power: ability to detect an effect that is there (1-ß) Alpha: odds that a detected effect is due to chance Estimation efficiency –The ability of a design to characterize the form of the BOLD changes –With a priori knowledge of HDR timing, shape, etc., one can describe temporal properties of brain regions We can think of these challenges as spatial and temporal, respectively.

51. Analyzing data using BIAC tools

52. Basic Steps of Analysis Reconstruction of K-Space Data Pre-processing Steps –Slice Timing Correction –Motion Correction –Coregistration, Normalization, Smoothing Epoch averaging / Regression –Statistical comparison of data with convolved hemodynamic response