Image analysis – Applications – Biomedical applications
Reexamination Certificate
1998-12-21
2003-08-19
Patel, Jayanti K. (Department: 2721)
Image analysis
Applications
Biomedical applications
C378S098700
Reexamination Certificate
active
06608915
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an image processing method and apparatus having a dynamic-range compression processing function for an image such as an X-ray chest image, using anatomical information.
2. Related Background Art
For example, an X-ray chest image is composed of an image of a lung field that X rays are easy to penetrate and an image of a mediastinum portion that X rays are very difficult to penetrate, so the range is very large in which pixel values are present.
Thus, it has been difficult to obtain an X-ray chest image that enables both the lung field and the mediastinum portion to be observed.
Consequently, for a practitioner to diagnose the chest, individual X-ray images (films) must be photographed and prepared for the diagnosis of the lung field and mediastinum respectively.
A method for avoiding this problem is a “self-compensation digital filter” (developed by Dr. Anan of National Cancer Center).
The self-compensation digital filter is expressed by the following expressions:
S
D
=S
org
+f
(
S
US
) (1)
S
US
=&Sgr;S
org
/M
2
(2)
where S
D
is a pixel value obtained after compensation (processing), S
org
is an original pixel value (an input pixel value), S
US
is an average pixel value obtained when taking the movement average of an original image (an input image) using a mask size of M×M pixels, and f(x) is a function having characteristics shown in
FIGS. 1A and 1B
.
The characteristics of the function f(x) are explained now. In the characteristic shown in
FIG. 1A
, f(x) is “0” when x>Th
a
where x is a signal value and where Th
a
is a threshold, and f(x) monotonously decreases with an intercept “Th
a
” and a slope “SLOPE
a
” when 0≦x≦Th
a
(the function f(x) having this characteristic is hereafter referred to as “f
a
(x)”).
Thus, when the equation (1) is executed using the original pixel value S
org
as a density equivalent, the density of the image is increased where its average is low.
In the characteristic shown in
FIG. 1B
, f(x) is “0” when 0<x≦BASE
b
where x is a signal value and where Th
b
is a threshold, and f(x) monotonously decreases into the negative area with an intercept “Th
b
” and a slope “SLOPE
b
” when x≧Th
b
(the function f(x) having this characteristic is hereafter referred to as “f
b
(x)”).
Thus, when the equation (1) is executed using the original pixel value S
org
as a density equivalent, the density of the image is reduced where its average is high.
When this “self-compensation digital filter” method is used for an image of the mediastinum that X rays are very difficult to penetrate, the density of the mediastinum portion of the X-ray chest image is increased due to the characteristic shown in
FIG. 1A
to provide an X-ray chest image that enables both the lung field and the mediastinum to be observed.
In addition to this self-compensation digital filter, a method is available that uses the results of anatomical segmentation to compress the dynamic range based on the difference in the characteristic amount of the anatomical site.
That is, as described in detail in “SPIE Medical Imaging 97 ‘Anatomic Region Based Dynamic Range Compression for Chest Radiographs Using Warping Transformation of Correlated Distribution’”, this method (hereafter referred to as the “dynamic-range-compression-based method”) subjects an X-ray chest image to predetermined image processing to define the mediastinum portion based on the results of the identification and extraction (hereafter referred to as “segmentation”) of the lung field portion, automatically determines an affine transformation function that transforms the pixel values of the lung field portion and/or mediastinum portion in order to analyze the distributions of the pixel values of the two image regions including the lung field portion and mediastinum portion and of the peripheral average value.
An X-ray chest image that enables both the lung field and mediastinum portion to be observed can also be obtained using this dynamic-range-compression-based method for an image of the mediastinum portion that X rays are very difficult to penetrate because it increases the density of this portion of the X-ray chest portion.
The conventional “self-compensation digital filter”, however, has no logical algorithm for automatically determining parameters such as slopes SLOPE
a
and SLOPE
b
. Thus, the desired results of a target image can be obtained depending on the settings of the parameters, but these results are not always obtained stably.
In addition, the conventional dynamic-range-compression-based method is for automating the compression of the lung field portion, and no method has been specified for compressing the mediastinum portion. In addition, the control of the amount of compression during dynamic-range compression is not specified. Thus, this method cannot control the contrast, thereby failing to provide quality X-ray images stably.
SUMMARY OF THE INVENTION
The present invention is provided to eliminate the above disadvantages, and its object is to provide an image processing method that can provide high-grade images stably.
Another object of the present invention is to provide an image processing apparatus, an image acquisition apparatus, and an image processing system that can provide high-grade images stably.
To achieve these objects, according to one aspect of the present invention, there is provided an image processing method including a step of dynamic-range-compression processing of an image, wherein the dynamic-range compression processing step comprises a sorting step of sorting the image into a first and a second sites, an extraction step of extracting the characteristic amounts of at least two pixels contained in at least either the first or second site, a statistical-amount acquisition step of determining the statistical amount of the characteristic amounts for at least either the first or second site, and a transformation step of transforming pixel values based on the statistical amount.
According to another aspect of the present invention, there is provided an image processing method including a step of dynamic-range-compression processing of an image, wherein the dynamic-range compression processing step comprises a sorting step of sorting the image into first and second sites, a boundary acquisition step of determining the boundary site between the first and second sites, an extraction step of extracting the characteristic amounts of at least two pixels contained in at least one of the first and second sites and the boundary site, a statistical-amount acquisition step of determining the statistical amount of characteristic amounts for at least one of the first and second sites and the boundary site, and a transformation step of transforming pixel values based on the statistical amount.
Furthermore, according to a preferred embodiment, the image is an X-ray image, and the sorting step includes a step of sorting as the first site a site in the X-ray image that X-rays are easy to penetrate while sorting as the second site a site in the X-ray image that X-rays are difficult to penetrate.
Further, according to another aspect of the present invention, there is provided an image processing apparatus for dynamic-range-compression processing of an image, comprising sorting means for sorting the image into first and second sites, extraction means for extracting the characteristic amounts of at least two pixels contained in at least either the first or second site, statistical-amount acquisition means for determining the statistical amount of the characteristic amounts for at least either the first or second site, and transformation means for transforming pixel values based on the statistical amount.
The other objects and features of this invention will be clear from the following detailed description of the embodiments referencing the drawings.
REFERENCES:
patent: 5228443 (1993-07-01), Tatar
patent: 5368033 (1994-11-01), Moshfeghi
patent: 5471987 (1995-1
Canon Kabushiki Kaisha
Choobin Barry
Patel Jayanti K.
LandOfFree
Image processing method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3127347