Method for enhanced visualization of medical images
View Patent ↗A method to process medical images for a desired global and local contrast enhancement. A digital image is accessed and segmented, and a first histogram of the digital image is generated. An overall contrast enhanced histogram is generated using the first histogram. A local contrast enhanced histogram is generated using the first histogram. A combined histogram is generated using the overall contrast enhanced histogram and the local contrast enhanced histogram. A contrast-enhanced digital image is generated using the combined histogram.
1. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
segmenting the digital mammographic image to identify a region of interest based on gray level of dense breast tissue;
generating a first histogram of the segmented digital mammographic image;
generating a global contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram for the region of interest by applying spatial correlation histogram equalization to the first histogram using spatial relationships of pixels in the region of interest, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram using the global contrast enhanced histogram and the detail contrast enhanced histogram; and
using the combined histogram to generate a contrast-enhanced digital mammographic image.
2. The method of claim 1 , wherein the combined histogram is determined by a weighted combination of the global contrast enhanced histogram and the detail contrast enhanced histogram.
3. The method of claim 1 , wherein the combined histogram is generated by selecting a gray level value on the first histogram as a threshold.
4. The method of claim 1 , wherein the combined histogram is determined by an equally weighted combination of the global contrast enhanced histogram and the detail contrast enhanced histogram.
5. The method of claim 1 , wherein the combined histogram is determined by an unequally weighted combination of the global contrast enhanced histogram and the detail contrast enhanced histogram.
6. The method of claim 1 , wherein the combined histogram is determined by a combination of weighting factors and gray level spaces.
7. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, wherein the step of applying spatial correlation histogram equalization comprises:
(a) calculating a gray-level histogram of the digital mammographic image to identify dense breast regions; and
(b) calculating a spatial correlation histogram of the digital mammographic image, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram using the overall contrast enhanced histogram and the detail contrast enhanced histogram; and
using the combined histogram to generate a contrast-enhanced digital mammographic image.
8. The method of claim 7 , wherein the combined histogram is determined by a weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
9. The method of claim 7 , wherein the combined histogram is generated by selecting a gray level value on the first histogram as a threshold.
10. The method of claim 7 , wherein the combined histogram is determined by an equally weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
11. The method of claim 7 , wherein the combined histogram is determined by an unequally weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
12. The method of claim 7 , wherein the combined histogram is determined by a combination of weighting factors and gray level spaces.
13. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, wherein the applying step comprises:
(a) selecting a gray level on the first histogram as a threshold level; and
(b) dividing the gray level into a plurality of parts, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram by combining the overall contrast enhanced histogram and the detail contrast enhanced histogram;
calculating a cumulative distribution function and a yield mapping function; and
mapping the combined histogram to a contrast-enhanced digital mammographic image using the cumulative distribution function.
14. The method of claim 13 , wherein the combined histogram is determined by a weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
15. The method of claim 13 , wherein the combined histogram is determined by an equally weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
16. The method of claim 13 , wherein the combined histogram is determined by an unequally weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram.
17. The method of claim 13 , wherein the combined histogram is determined by a combination of weighting factors and gray level spaces.
18. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram, the step of generating a detail contrast enhanced image comprising steps of:
(a) calculating a gray-level histogram p(k) of the digital mammographic image using equation (1)
p
(
k
)
=
m
k
m
k
=
0
,
1
,
2
,
…
,
n
-
1
(
1
)
wherein m is the total number of pixels in the digital mammographic image, m k is the number of pixels of gray level k , and n is the total number of gray levels; and
(b) calculating a spatial correlation histogram p (k) of the digital mammographic image using equation (2)
p
_
(
k
)
=
∑
b
=
k
-
α
k
+
α
p
(
b
|
k
)
∑
a
=
0
n
-
1
∑
b
=
a
-
α
a
+
α
p
(
b
|
a
)
(
2
)
wherein the conditional probability p(b |a) is defined using equation (3)
p
(
b
|
a
)
=
∑
u
=
0
N
-
1
∑
v
=
0
M
-
1
[
∑
w
=
u
-
β
u
+
β
∑
z
=
v
-
β
v
+
β
H
(
a
,
q
(
u
,
v
)
)
·
H
(
b
,
q
(
w
,
z
)
)
]
/
σ
(
β
)
∑
u
=
0
N
-
1
∑
v
=
0
M
-
1
H
(
a
,
q
(
u
,
v
)
)
where
(
,
q
(
x
,
y
)
)
=
{
0
g
≠
q
(
x
,
y
)
=
(
x
,
y
)
(
3
)
and q(x, y) is the gray level at pixel (x, y), σ(β) =(2 β+1)×(2β+1); N and M are the width and height of a mammogram respectively; and the element q(x, y) denotes the probability of gray level being b for pixels in a neighborhood centered by a pixel with gray level being a ;
generating a combined histogram using the overall contrast enhanced histogram and the detail contrast enhanced histogram; and
using the combined histogram to generate a contrast-enhanced digital mammographic image.
19. The method of claim 18 , wherein the step of generating a combined histogram comprises steps of:
selecting a gray level value on the first histogram as a threshold k′;
dividing the gray level space into two parts:
K={k|p ( k *)< k′,p ( k*+ 1)≧ k′,k =0,1,2, . . . k*}and
K ={ 0,1, . . . , n− 1 }−K
calculating the combined histogram by combining the overall contrast enhanced histogram and the detail contrast enhanced histogram using equation (4)
p
(
k
)
new
=
{
p
(
k
)
,
k
∈
K
p
_
(
k
)
∑
i
∈
K
_
p
_
(
i
)
·
(
1
-
∑
i
∈
K
p
(
i
)
)
,
k
∈
K
_
(
4
)
calculating a cumulative distribution function F(k) new and a yield mapping function; and
mapping the combined histogram to a contrast-enhanced image based on the combined histogram using equation (5)
k′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
.
(
5
)
20. The method of claim 18 , wherein the step of generating a combined histogram comprises steps of:
producing a new histogram using equation (6)
p ( k ) new =γ p ( k ) +(1−γ) p ( k )* (0 <γ<1). (6);
calculating a cumulative distribution function F(k) new , and a yield mapping function; and
mapping the new histogram to a contrast-enhanced image based on the combined histogram using equation (7)
k′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
.
(
7
)
21. The method of claim 18 , wherein the step of generating a combined histogram comprises steps of
selecting a gray level value on the first histogram as a threshold k′;
dividing the gray level space into two parts:
K={k|p ( k *)< k′, p ( k*+ 1)≧ k′, k =0,1,2, . . . k*}and
K ={0,1, . . . , n − 1}− K;
producing a new histogram using equation (8)
{
p
(
k
)
new
=
γ
p
(
k
)
+
(
1
-
γ
)
p
_
(
k
)
*
(
0
<
γ
<
1
)
.
k
∈
K
p
(
k
)
new
=
γ
′
p
(
k
)
+
(
1
-
γ
′
)
p
_
(
k
)
*
(
0
<
γ
′
<
1
)
k
∈
K
_
;
(
8
)
calculating a cumulative distribution function F(k) new , and a yield mapping function; and
mapping the new histogram to a contrast-enhanced image based on the combined histogram using equation (7)
k′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
.
(
7
)
22. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram using the overall contrast enhanced histogram and the detail contrast enhanced histogram, the step of generating a combined histogram comprising steps of
(a) selecting a gray level value on the first histogram as a threshold k′;
(b) dividing the gray level space into two parts:
K={k|p ( k *)< k′, p ( k*+ 1)≧ k′, k =0,1,2, . . . k*}and
K ={0,1, . . . , n− 1}− K;
(c) calculating the combined histogram by combining the overall contrast enhanced histogram and the detail contrast enhanced histogram using equation (4)
p
(
k
)
new
=
{
p
(
k
)
,
k
∈
K
p
_
(
k
)
∑
i
∈
K
_
p
_
(
i
)
·
(
1
-
∑
i
∈
K
p
(
i
)
)
,
k
∈
K
_
(
4
)
(d) calculating a cumulative distribution function F(k) new and a yield mapping function; and
(e) mapping the combined histogram to a contrast-enhanced image based on the combined histogram using equation (5)
k
′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
;
(
5
)
and
using the combined histogram to generate a contrast-enhanced digital mammographic image.
23. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of:
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram using an unequally weighted combination of the overall contrast enhanced histogram and the detail contrast enhanced histogram, the generating a combined histogram comprising steps of:
(a) producing a new histogram using equation (6)
p ( k ) new , =γp ( k )+(1−γ) p ( k )* (0 <γ<1). (6);
(b) calculating a cumulative distribution function F(k) new and a yield mapping function; and
(c) mapping the new histogram to a contrast-enhanced image based on the combined histogram using equation (7)
k
′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
;
(
7
)
and
using the combined histogram to generate a contrast-enhanced digital mammographic image.
24. A method for enhancing the image contrast of a digital mammographic image, comprising using a computer to perform steps of
accessing the digital mammographic image;
generating a first histogram of the digital mammographic image;
generating an overall contrast enhanced histogram by applying histogram equalization to the first histogram;
generating a detail contrast enhanced histogram by applying spatial correlation histogram equalization to the first histogram, thereby preserving gray level detail at a tail end of the detail contrast enhanced histogram;
generating a combined histogram using the overall contrast enhanced histogram and the detail contrast enhanced histogram, the generating a combined histogram using a combination of weighting factors and gray scale level spaces and comprising steps of:
(a) selecting a gray level value on the first histogram as a threshold k′;
(b) dividing the gray level space into two parts:
K={k|p ( k *)< k′, p ( k*+ 1)≧ k′, k =0,1,2, . . . k*}and
K ={0,1, . . . , n− 1}− K;
(c) producing a new histogram using equation (8)
{
p
(
k
)
new
=
γ
p
(
k
)
+
(
1
-
γ
)
p
_
(
k
)
*
(
0
<
γ
<
1
)
.
k
∈
K
p
(
k
)
new
=
γ
′
p
(
k
)
+
(
1
-
γ
′
)
p
_
(
k
)
*
(
0
<
γ
′
<
1
)
k
∈
K
_
;
(
8
)
(d) calculating a cumulative distribution function F(k) new , and a yield mapping function; and
(e) mapping the new histogram to a contrast-enhanced image based on the combined histogram using equation (7)
k
′
=
n
·
F
(
k
)
new
=
n
·
∑
i
=
0
k
p
(
i
)
new
;
(
7
)
and
using the combined histogram to generate a contrast-enhanced digital mammographic image.