Signal to noise ratio (SNR) and data quality
Coils
Head coil
Surface coil
•homogenous signal
•moderate SNR
•highest signal at hotspot
•high SNR at hotspot
Source: Joe Gati
Calculating Signal:Noise Ratio
Pick a region of interest (ROI) outside the brain free from artifacts (no ghosts,
susceptibility artifacts). Find mean () and standard deviation (SD).
Pick an ROI inside the brain in the area you care about. Find  and SD.
SNR = brain/ outside = 200/4 = 50
e.g., =4, SD=2.1
Alternatively SNR = brain/ SDoutside = 200/2.1 = 95
(should be 1/1.91 of above because /SD ~ 1.91)
When citing SNR, state which denominator you used.
e.g.,  = 200
Head coil should have SNR > 50:1
Surface coil should have SNR > 100:1
Source: Joe Gati, personal communication
What affects SNR?
Physical factors
PHYSICAL FACTORS
SOLUTION & TRADEOFF
Thermal Noise (body & system)
Inherent – can’t change
Magnet Strength
e.g. 1.5T  4T gives 2-4X increase in SNR
Use higher field magnet
Coil
e.g., head  surface coil gives ~2+X increase in
SNR
Use surface coil
Voxel size
e.g., doubling slice thickness increases SNR by root2
Use larger voxel size
Sampling time
Longer scan sessions
– additional cost and maintenance
– physiological noise may increase
– Lose other brain areas
– Lose homogeneity
– Lose resolution
– additional time, money and subject
discomfort
Source: Doug Noll’s online tutorial and Jody Culham’s web site
What affects SNR?
Physiological factors
PHYSIOLOGICAL FACTORS
SOLUTION & TRADEOFF
Cardiac and respiratory noise
Monitor and compensate
– very difficult to do
Head (and body) motion
Use experienced or well-trained subjects
– limited subject pool
Use head-restraint system
– subject discomfort
Post-processing correction
– often incompletely effective
Single trials to avoid body motion
Low frequency noise
Use smart design
Perform post-processing filtering
BOLD noise (neural and vascular fluctuations)
Use many trials to average out variability
Behavioral variations
Use well-controlled paradigm
Use many trials to average out variability
Source: Doug Noll’s online tutorial and Jody Culham’s web site
Physiological Noise
Respiration
• every 4-10 sec (0.3 Hz)
• moving chest distorts susceptibility
• deep breaths particularly problematic
(instruct subject well)
Cardiac Cycle
• every ~1 sec (0.9 Hz)
• pulsing motion, blood changes
Solutions
• gating
• avoiding paradigms at those frequencies
Low and High Frequency Noise
Head Motion: Main Artifacts
Head motion Problems
1) Rim artifacts
•
•
hard to tell activation from artifacts
artifacts can work against activation
time1
 time2
Playing a movie of
slices over time helps
you detect head motion
Looking at the negative
tail can help you
identify artifacts
2) Region of interest moves
•lose effects because you’re sampling outside
ROI
Head Restraint
Other:
• Thermoplastic Mask (used in
PET)
• Vacuum packs
• Tape across forehead
• Foam padding
Head Vise
(more comfortable than it
sounds!)
Bite Bar
(less comfortable
than head vice!)
Motion Correction Options
2D realignment
• fast
• 2 degrees of freedom (2 translations)
3D realignment
• slow
• more accurate
• 6 degrees of freedom (3 translations, 3 rotations)
• can lose parts of brain
Can realign within a run or within a session
Motion Correction Output
SPM output
raw data
gradual motions are usually well-corrected
linear trend removal
abrupt motions are more of a problem (esp
if related to paradigm
motion corrected in
SPM
Caveat: Motion correction can cause
artifacts where there were none
Caveat: Motion
correction in BV
doesn’t seem nearly as
good as SPM
Head Motion: Susceptibility Artifacts
Stationary Head
or Phantom
Bag of Saline on a Stick
• experimenter
moves saline left
and right every 20
sec without
touching subject or
phantom
Analyze data using saline motion as “paradigm”
Head Motion: Solution to Susceptibility
Solution:
• one trial every 10 or 20 sec
• fMRI signal is delayed ~5 sec
distinguish true activity from artifacts
Especially good for motor paradigms –
any artifact from the movement made
by the subject should be gone once the
critical data is collected!
action
artifact activity
fMRI
Signal
0
5
10
Time (Sec)
Effect of Filtering – spatial smoothing.
before
after
Source: Brain Voyager course slides
Trial-to-trial variability
Single trials
Average of all trials from 2 runs
Other Artifacts
Ghosts
Zebra Brains
Spikes
Metallic Objects (e.g., hair tie)
Other Artifacts
Poor shimming