Image analysis – Histogram processing
Reexamination Certificate
1999-08-17
2001-04-10
Tran, Phuoc (Department: 2721)
Image analysis
Histogram processing
C382S172000
Reexamination Certificate
active
06215900
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to a method for displaying an image composed of a plurality of digital values, and more particularly to a method for mapping a plurality of digital values onto a gray scale transfer function of a display medium, to display a radiogram.
BACKGROUND OF THE INVENTION
There exists significant activity in the development of digital x-ray image data capture systems. In such systems direct conversion to an electrical signal of the incident radiation is obtained using a plurality of sensors in an array. The sensor output is almost invariably immediately converted to a digital signal and further processed and stored in a databank for use in the eventual display of the data as a radiogram. U.S. Pat. No. 5,313,066 issued to Lee et al. and U.S. Pat. No. 5,315,101 issued to Hughes et al describe typical such sensor arrays and their contents are incorporated herein by reference. Even though several different technologies are being utilized, the output data is quite similar.
A major advantage of digital data detection systems is the wide dynamic range of signal capture. Display media, such as radiographic film or Cathode Ray Tube (CRT) displays, on the other hand, have a substantially more limited dynamic range. A typical digital x-ray capture system can have a useful dynamic detection range of greater than a 1,000:1. However, the useful image data is generally limited to a dynamic range of less than 100:1. There is, therefore, need to determine and select the optimal limited range of useful data for diagnostic display, and then properly display such range on the available display medium.
This problem which reduces to a need for a method whereby the exposure sensor output is mapped onto the display device density transfer function has been addressed by the art in numerous ways. Typically the sensor output is digitized and a histogram of the frequency of occurrence of digital values representing detected exposure is constructed. Following construction of the histogram, cutoff points eliminating values under selected minimum occurrence for both ends of the scale are determined and the digital values in the remaining range are mapped onto the display transfer function. These steps are rather fundamental and intuitive. What is significant and the subject of continuing research is the specific selection process for the two cutoff points and the manner in which the remaining density values are actually mapped on the transfer function.
U.S. Pat. No. 5,164,993 issued Nov. 17, 1992 to Capozzi et al. together with U.S. Pat. Nos. 5,046,118 issued to Ajewole et al. and 4,868,651 issued to Chou et al. are believed to represent the current state of the art in explaining and solving the problems associated with such displays.
The currently available solutions do not adequately address the problem of adapting the digital value mapping to a particular type of both patient and examination. For example, a different portion of the data generated by the radiation sensors is of interest in displaying a radiogram depending on whether the radiogram is one of an extremity or a chest cavity, whether the patient is thin or obese, and what is the area desired to be displayed with maximum diagnostic efficiency. There is, therefore, still need for a system which addresses these problems and which with simple operator input automatically maps the available data in a way as to optimize the display medium density range for a particular set of examination type, patient characteristics and display medium capabilities.
It is an object of this invention to provide a method for automatically identifying the range of useful digital values to be used for diagnostic display, and to provide an appropriate gray scale transfer to optimize the diagnostic value of the final displayed image, either hard or soft copy.
SUMMARY OF THE INVENTION
In its broadest aspect, the present invention is a method for displaying on a display medium an image comprising a plurality of pixels having various optical densities stored in a data bank as a plurality of digital values corresponding to a detected exposure for each of said plurality of pixels, the method comprising the following steps:
I) In a programmed computer comprising a memory, storing a first look-up table (LUT) representing at least one gray scale transfer function for the display medium, a second LUT containing a plurality of experimentally determined constants, and a plurality of algorithms;
II) Constructing a histogram representing the frequency of occurrence of a range of digital values stored in said data bank.
III) Obtaining the integral of the histogram;
IV) Determining a low point on the histogram, DVlow, corresponding to a first predetermined percentage of the histogram integral and a point DVedge wherein the histogram has dropped to a predetermined percentage from a peak value;
V) selecting a type of radiographic examination from a list of radiographic examinations, retrieving from the second LUT preset constant values related to the selected radiographic examination type and using a preselected one or more of the plurality of stored algorithms related to the examination type with the retrieved constants to calculate:
a) a value DVmin,
b) a value DVmax
VI) Replacing all digital values in the data bank which are lower than DVmin with DVmin, and all digital values which are higher than DVmax, by DVmax to obtain a new range of data digital values.
VII) using the new range of data digital values to enter the first LUT and to obtain a mapped range of display data values and
VIII) displaying the image on the display medium using the display data values.
In a preferred embodiment the algorithms stored in the memory for calculating DVmin and DVmax are:
1. DVmax=DVmin+((SF/C)×Log (B)); and
DVmin=DVlow−SF×Log (A)
2. DVmax=DVmin+(SF×Log (B)−&Dgr;Range)/C and
DVmin=DVmin−SF×Log (A)
3. DVmax=DVmin+(1/L)×(DVedge−DVmin) and
DVmin=(DVlow−SF×Log (A))−((1/C)×(DVedge−DVmin))−DS
4. DVmax=DVmin+(1/L)×(DVpeak−DVmin)
DVmin=(DVlow−SF×Log (A))-DS
In an alternate embodiment, the process uses a plurality of gray scale transfer functions stored in the first LUT, each such gray scale transfer function being adapted so that DVedge produces a preselected output density on the display medium, such density being dependent on a specific desired visual appearance of the displayed data selected by the viewer. In essence, this multiplicity of gray scale transfer functions provides the operator with the ability to simulate in digital radiography the possibility of using different type photographic films and film screen combinations in traditional radiography, to optimize the visual appearance of the radiogram.
REFERENCES:
patent: 4868051 (1989-09-01), Chou et al.
patent: 5046118 (1991-09-01), Ajewola et al.
patent: 5164993 (1992-11-01), Capozzi et al.
patent: 5313063 (1994-05-01), Netzer
patent: 5315101 (1994-05-01), Hughes et al.
patent: 5633511 (1997-05-01), Lee et al.
patent: 5675624 (1997-10-01), Relihan et al.
Schwenker Emily J.
Schwenker Ronald P.
Williams Cornell L.
Direct Radiography Corp.
Ratner & Prestia
Schwenker Emily J.
Tran Phuoc
LandOfFree
Image data processing for digital x-ray detectors 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 data processing for digital x-ray detectors, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image data processing for digital x-ray detectors will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2481684