Ultrasound signal correlation phase estimation of lung motion
A method of estimating lung motion (e.g., local lung motion and local lung ventilation) includes collecting a time series of ultrasound images of a lung surface, the time series including a plurality of frames, identifying a lung surface in one of the plurality of frames, and subsetting each of the plurality of frames into at least one sub-image. The method further includes applying a high pass temporal filter and a spatial filter. The method still further includes calculating inter-frame motion data from complex data sets associated with one or more pairs of temporally successive frames. In a further method, lung surface longitudinal strain is also calculated.
1 . A method of estimating lung motion, the method comprising:
collecting a time series of B-mode ultrasound images of a lung surface, the time series including a plurality of frames;
identifying a lung surface in one of the plurality of frames;
subsetting, by a processor, each of the plurality of frames into at least one sub-image with the lung surface arranged along the x-axis of the at least one sub-image;
applying a spatial filter along an x-axis of each sub-image to transform each pixel of each sub-image into ultrasound B-mode data, the ultrasound B-mode data including a real component and an imaginary component, the ultrasound B-mode data compiled in a complex data set, each complex data set associated with one of the plurality of frames, and
calculating inter-frame motion data of a window location from the complex data sets associated respectively with a first frame of the plurality of frames and a second frame of the plurality of frames that is temporally successive to the first frame of the plurality of frames,
wherein the ultrasound B-mode data is pulse-range data represented by s(x)=A(x−U)e −i2πf(x−U) , where x is a digital spatial coordinate of a pixel, U is a position of a signal, A(x−U) is a signal amplitude profile, and f is a spatial center frequency of a complex signal, and
wherein a phase of s(x), the pulse-range data, is proportional to a difference of x, the digital spatial coordinate of the pixel, and U, the position of the signal.
2 . The method of claim 1 , wherein identifying a lung surface in one of the plurality of frames comprises drawing a line between a right edge and a left edge of the lung surface in one of the plurality of frames.
3 . The method of claim 2 , wherein the line is drawn by an operator using software on a computer.
4 . The method of claim 2 , wherein subsetting each of the plurality of frames into at least one sub-image is based on a location of the line.
5 . The method of claim 4 , wherein subsetting each of the plurality of frames of the time series includes,
for the one of the plurality of frames on which the line is drawn, identifying a sub-image having a fixed number of pixels above the line and below the line, and
for remaining frames of the plurality of frames, identifying a sub-image for each of the remaining frames having a same set of pixels as the sub-image of the one of the plurality of frames on which the line is drawn.
6 . The method of claim 1 , wherein identifying a lung surface in one of the plurality of frames and subsetting each of the plurality of frames into at least one sub-image is automated by use of image segmentation algorithms.
7 . The method of claim 1 , wherein the spatial filter is a synthetic phase filter or a gradient filter.
8 . The method of claim 1 , further comprising applying a spatial filter along a y-axis of each sub-image.
9 . The method of claim 1 , wherein calculating inter-frame motion data comprises extracting a first signal from a correlation window of the first frame and a second signal from the correlation window of the second frame, the correlation window being a same location in the first frame and the second frame, where the first signal is represented by s 1 (x)=A 1 (x−U 1 )e −i2πf(x−U 1 ) and the second signal is represented by s 2 (x)=A 2 (x−U 2 )e −i2πf(x−U 2 ) .
10 . The method of claim 9 , wherein calculating inter-frame motion data further comprises calculating a sum of a complex conjugate multiple of the first signal and the second signal, which is a zero lag covariance of s 1 and s 2 and is represented by
σ
(
0
)
=
∑
x
s
2
(
x
)
*
s
1
*
(
x
)
=
∑
x
A
2
(
x
-
U
2
)
e
-
i
2
π
f
(
x
-
U
2
)
A
1
(
x
-
U
1
)
e
i
2
π
f
(
x
-
U
1
)
11 . The method of claim 10 , wherein calculating inter-frame motion data further comprises calculating a phase angle of the zero lag covariance of s 1 and s 2 , which is a phase angle of the sum of the complex conjugate multiple of the first signal and the second signal, represented by
angle
(
σ
(
0
)
)
=
angle
(
s
2
(
x
)
s
1
*
(
x
)
)
=
2
π
f
(
U
2
-
U
1
)
.
12 . The method of claim 10 , wherein calculating inter-frame motion data further comprises calculating a lag 1 covariance, which is the complex conjuate multiple with s 1 shifted by one pixel, represented by
σ
(
1
)
=
∑
x
s
2
(
x
-
1
)
s
1
*
(
x
)
=
∑
x
A
2
(
x
-
1
-
U
2
)
e
-
i
2
π
f
(
x
-
U
2
-
1
)
A
1
(
x
-
U
1
)
e
i
2
π
f
(
x
-
U
1
)
.
13 . The method of claim 12 , wherein calculating inter- frame motion data further comprises calculating a phase angle of a complex conjugate multiple of the zero lag covariance and the lag 1 covariance as represented by angle(σ(1)σ(0))=2ρf.
14 . The method of claim 13 , wherein calculating inter-frame motion data further comprises calculating a displacement represented by ΔU 1,2 =2πf(U 2 -U 1 )/2πf.
15 . The method of claim 14 ,
further comprising identifying pairs of the plurality of frames, each frame having a temporally successive frame and identified as forming a pair that includes the frame and the temporally successive frame, and calculating inter-frame motion for all pairs of the plurality of frames from the complex data sets associated with the frame and the temporally successive frame of each pair, and
wherein the displacement is calculated for each window location of each sub-image for all pairs of the plurality of frames.
16 . The method of claim 1 , further comprising identifying pairs of the plurality of frames, each frame having a temporally successive frame and identified as forming a pair that includes the frame and the temporally successive frame, and calculating inter-frame motion for all pairs of the plurality of frames from the complex data sets associated with the frame and the temporally successive frame of each pair.
17 . The method of claim 1 , wherein a segmentation algorithm is applied to identify pixels from the lung surface, and wherein lung surface motion is calculated from displacement of pixels identified as the lung surface.
18 . The method of claim 1 , further comprising applying a high pass temporal filter, a band pass filter, or an all pass filter to each sub-image.