IIIT Hyderabad A METHOD FOR MOTION DETECTION AND CATEGORIZATION IN PERFUSION WEIGHTED MRI Rohit Gautam, Jayanthi Sivaswamy CVIT, IIIT Hyderabad, Hyderabad, India Ravi Varma KIMS Hospital, Hyderabad, India
Feb 05, 2016
IIIT
Hyd
erab
ad
A METHOD FOR MOTION DETECTION AND CATEGORIZATION IN PERFUSION
WEIGHTED MRI
Rohit Gautam, Jayanthi SivaswamyCVIT, IIIT Hyderabad, Hyderabad, India
Ravi VarmaKIMS Hospital, Hyderabad, India
IIIT
Hyd
erab
ad
Introduction
• MRI for brain - Application of nuclear magnetic resonance (NMR) to create images of human brain.
From MRI to Perfusion MRI
• Many neurological disorders can be detected using abnormal blood flow.
• Perfusion MRI utilizes this blood flow information in disease diagnosis.
MRI
PerfusionMRI
IIIT
Hyd
erab
ad
What is Perfusion MRI ?
• Perfusion is the delivery of oxygen and nutrients to the cells via capillaries.
• A bolus injected into patient’s blood is tracked over time.
• It provides information regarding rate of blood flow, which helps to determine the affected regions in brain on the onset of disorder.
• Acquired data is a 3D time-series.
1 Nnwin nwout
Time-points
Before Bolus wash-in
After Bolus wash-out
Bolus in transit
Volume
IIIT
Hyd
erab
ad
Problem
Patient motion during MRI scan (> 2 minutes) misaligns (corrupts) the acquired data.
Aim - Motion detection and categorization of volumes corrupted due to patient motion.
Difficulties•Simultaneous local (non-uniform variation in image contrast due to bolus) and global (motion) changes.•Current scenario: Motion correction is a time-limiting
step in PWI analysis [ Straka et al. JMRI 07].
IIIT
Hyd
erab
ad
N
1
Motion
Motion
Motion
Beforebolus
wash-in
After bolus
wash-out
Bolus intransit
No variation in intensity
Non-uniform Variation in intensity
No variation in intensity
IIIT
Hyd
erab
ad
Why motion correction ?
TTP: Time to PeakCBV: Cerebral Blood Volume
Perfusion parameters obtained from motion corrupted data vary with degree of motion.
Error in CBV estimation Error in TTP estimation
IIIT
Hyd
erab
ad
Proposed Divide and Conquer Strategy
IIIT
Hyd
erab
ad
Division of Time-Series
• The signal intensity in perfusion MRI varies proportionally with bolus concentration.
• A gamma-variate function (GVF) can model the change in concentration of bolus with time.
• We fit GVF on the mean-intensity perfusion curve µa(n) to estimate GVF-fit mean intensity curve µg(n).
• Using µg(n), we divide the time-series into 3 sets: Set1: pre-wash-in, Set2: transit and Set3: post-washout setsWash-in
Time pointWash-outTime point
IIIT
Hyd
erab
ad
Divide and Conquer Strategy
IIIT
Hyd
erab
ad
Intensity Correction
• Bolus is present only in Set2. Hence, intensity correction is required for bolus affected regions in these volumes before motion detection.
• A volume (F) is segmented into normal (Fnormal) and bolus affected (Fbolus) regions using clustering technique.
• F is then intensity corrected :
where, µg(n) is the GVF-fit-mean intensity curve.
IIIT
Hyd
erab
ad
Intensity CorrectionSlice 1 Slice 2
Intensity CorrectedSlice 2
AbsoluteDifference
AbsoluteDifference
Ideally, these should be 0
Reduction in absolute intensity
difference
IntensityCorrection
Slice 1
IIIT
Hyd
erab
ad
Divide and Conquer Strategy
IIIT
Hyd
erab
ad
Motion Detection Workflow
IIIT
Hyd
erab
ad
Motion Detection
For each mean shifted blocks of fixed size around each pixel, the flow vector is given by:
where S is the cross power spectrum.
Extract Central Slices
Block wise Phase Correlation
Process is accelerated by down-sampling of central slices.
Block wise Phase Correlation
IIIT
Hyd
erab
ad
Motion Flow Maps
Slice 1 Slice 2
Un Vn
Bolus present andNo motion
Slice 1 Slice 2
Un Vn
Bolus absent andMinimal motion
Slice 1 Slice 2
Un Vn
Bolus absent andMild motion
IIIT
Hyd
erab
ad
Net Entropy Metric
• For a given time series {Fn; n=1…N} , the net entropy (Hn) of flow fields (Un;Vn)is given by:
• The net entropy is 0 for no motion and increases with degree of motion.
IIIT
Hyd
erab
ad
Zero net entropy even in the
presence of bolus.
Net Entropy Profile
1 5 8
33 40Wash-in
Time-pointWash-in
Time-point
IIIT
Hyd
erab
ad
A non-zero net entropy even in the absence of
motion
Does Intensity Correction help ?
IIIT
Hyd
erab
ad
Slice Resolution
Block Size Mean time per slice pair (sec)
Total time(sec)
128x128 32x32 0.00 + 3.48 = 3.48 132.21
128x128 16x16 0.00 + 3.99 = 3.99 151.69
128x128 8x8 0.00 + 4.34 = 4.34 164.84
64x64 16x16 0.01 + 0.77 = 0.78 29.71
64x64 8x8 0.01 + 0.97 = 0.98 37.38
32x32 8x8 0.01 + 0.19 = 0.20 7.68
Time Analysis of motion detection
Large reduction in computation time
IIIT
Hyd
erab
ad
Divide and Conquer Strategy
IIIT
Hyd
erab
ad
Motion Categorization
• Peak entropy of flow fields is used to quantify the degree of motion.
• The peak entropy Hpeak of the flow fields is found by:
where, Hn is the net entropy.
IIIT
Hyd
erab
ad
Such a small motion cannot be
detected.
Peak entropy can distinguish
between different motion categories.
Entropy values for different motion categories for image size – 32x32 and block size 8x8
IIIT
Hyd
erab
ad
Comparison of our approach
• We compare the efficiency of motion detection method by applying it prior to existing motion correction algorithms.
[1] Kosior et al., JMRI 2007.[2] Straka et al., JMRI 2010.[3] Tanner et al., MICCAI 2000.
IIIT
Hyd
erab
ad
Conclusions
• We proposed a motion detection method that is immune to intensity changes due to injected contrast agent.
• We achieved a large reduction in time (~37%) required for motion correction by rejecting the stationary volumes.
• The detection method can be made to be fast but the sensitivity to minimal motion maybe compromised.
IIIT
Hyd
erab
ad
Questions ?
IIIT
Hyd
erab
ad
Thank you