 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
This Invention relates to medical imaging and identifying and archiving
medical image information relating to a subject. More particularly, the
invention relates to a method of automatically extracting information from
a text frame image and organizing text information and images in a
particular manner which facilitates ease of storage and retrieval.
2. Background Information and Description of the Prior Art
Medical imaging techniques and devices for obtaining high quality medical
images have greatly improved in recent years. With the improvement in
techniques and quality of images obtained, an increasing amount of imaging
is being performed which results in a greater amount of image information
to be organized and archived for later reference by radiologists, other
medical personnel or other interested parties.
Many medical imaging devices are currently being used in radiology.
Although these devices produce images in a variety of forms, the images
can all be converted to a form suitable for digital archival and retrieval
systems. Digital images can be produced directly by techniques including
computed tomography (CT), nuclear magnetic resonance imaging (MRI),
ultrasound, digital subtraction angiography and gamma ray imaging In
addition, traditional x-ray projection images continue to provide
information about many biological systems and these film images can be
digitized to a form compatible with digital image management systems. It
is envisioned that computer technology which is now being applied to
radiology will provide the ability to view, print, and store images
developed from any of the aforementioned devices. Additionally, physicians
will be interested in retrieving image data about a particular subject
generated using all modalities including CT, MRI and x-ray. The basic
concepts of various image modalities and image viewing of data generated
by such devices are described in U.S. Pat. No. 4,833,625.
Historically, images were stored on films which were filed in various
ordering systems. A large percentage of such films are lost, or the films
become physically separated from file jackets and it then becomes
impossible to identify the subject to whom the film relates. More
recently, digital image data is stored in computer archival systems. In
conventional MRI and CT systems, many images are generated, each having
its own identification number and header information. The header
information includes textual information about the patient and about the
study being conducted. Typically, groups of these images are arranged to
appear together on separate pages, with possibly twelve to twenty images
per page. A technician must perform the task of entering the
identification information and arranging the images into pages. Physicians
typically prefer to review images in a page-by-page format and on film
rather than viewing separate original images on screen. As noted
hereinbefore, historically the information was archived by saving the
films and deleting the original image data. More recently, digital
archival systems have been developed which store each of the small images
individually.
A medical image filing system was disclosed in U.S. Pat. No. 4,879,665.
This involves assigning information to image data and using the
information in order to retrieve image data which has been stored. Each
separate medical picture or image must have retrieving data associated
with it in order to retrieve the image This results in a large number of
entries in the directory and time is sacrificed in processing textual
information associated with every individual image which is archived.
An image data management system was disclosed in U.S. Pat. No. 4,653,112.
U.S. Pat. No. 4,653,112 employs a ring-type architecture data transmission
network having nodes comprising data input/output points Node interfaces
are provided which allow data to flow between an associated image
acquisition device or display console and a host computer. Although a
method of utilization of storage capacity for image data is provided, a
more efficient means of storing textual information is not disclosed. In
addition, U.S. Pat. No. 4,653,112 does not disclose a method of retrieving
images from storage based upon such textual information.
Other types of image data acquisition devices and archiving systems have
been disclosed. For example, U.S. Pat. No. 4,430,749 discloses a medical
imaging apparatus and method for furnishing difference images U.S. Pat.
No. 4,437,161 discloses a medical imaging apparatus containing a device
for detecting radiation emitted from an object under examination. Both of
the these patents relate to magnetic resonance imaging.
Imaging sofware is disclosed in U.S. Pat. No. 4,585,992. A
three-dimensional imaging display system was disclosed in U.S. Pat. No.
4,737,921. A method of dual image recording of medical data was disclosed
in U.S. Pat. No. 4,665,004 which involves recording a digital
representation of an image and saving the raw image data.
As can be seen from the above discussion, methods of image generating and
systems for archiving and storing medical images have been known. However,
there remains a need for efficient storage of images and textual
information in relation to those images. At present there are two classes
of digital image archive systems--vendor specific and general purpose. The
vendor specific devices are very complex and cannot be easily integrated
with equipment from other vendors The general purpose systems suffer from
an inability to communicate the textual information associated with an
image and hence require a technician to enter the relevant textual
information at the beginning of a study as well as prior to archiving
image data to be saved. This involves dual entry of the same information.
The additional step can lead to typographical and other errors which in
many systems cannot be corrected.
In addition, there remains a need for a digital interface communication
system which allows immediate retrieval of images from a variety of image
acquisition devices. The present trend is for each vendor to provide a
proprietary method for the management of image data produced by image
acquisition equipment supplied by that vendor. It can be extremely
difficult to arrange for systems from different vendors to interact with
one another in an efficient manner. Recently, standard protocols have been
proposed for image archival and retrieval systems, however, such protocols
are extremely expensive and complex and have not been universally adopted
by vendors. Although it is extremely desirable, it is not currently
practical for image acquisition equipment from multiple vendors to be
interconnected into a single image management system in a way suitable for
use in radiology departments.
Moreover, there remains a need for a more efficient method of organizing
directories which are then used for retrieving image information. At
present, the textual data archived by means of digitizing analog signals
must be entered a second time as to each set of image data to which it
relates. Further, textual information is stored in the computer archiving
system for each image generated. Many images are required for one study
and this means that there is a great amount of duplication of information
stored. As noted hereinbefore, conventional systems involve image-by-image
archival systems for which an entry is maintained in the data base for
each of the individual small images. The size of such a data base grows
very rapidly (about ten million records per year in some institutions) and
it becomes progressively more difficult for data to be retrieved.
SUMMARY OF THE INVENTION
These and other needs have been satisfied by the present invention which
provides a method of organizing and archiving data developed during an
imaging study of a subject. More specifically, the invention involves
generating a text frame image which is an image of identifying information
relating to a subject in the study to be conducted. This text frame image
is then "frame grabbed" or digitized to create digital text information.
This digital text image information is converted into character
information. The character information is searched and a header is created
in a predetermined format using the character information. In accordance
with one aspect of the invention, this character information is then
transferred back into image information and stored with a page of images
such that when the image page is retrieved, the first image block on the
page is a textual block which identifies the images. The identifying
information may include patient name, social security number, and the
nature of the study being conducted.
In accordance with another aspect of the invention, image data is stored in
a page-by-page format. A single page typically contains a group of twelve
to twenty images. Physicians generally prefer to review images in a
page-by-page format. Thus, storing the images in such a format avoids the
need to reconstruct pages which have been stored in an image-by-image
format.
In accordance with yet a further aspect of the invention, a second set of
identifying information is retrieved from the text file created upon
capture of the text image. This second set of information is translated
from character information into image information and is inserted at a
predetermined location on each page of image data. A universal location
for the textual information identifies each particular image and allows
such identification to appear in a permanent form when a film is printed
of the page. This reduces the chances of misidentification of the printed
film. In addition, the universal identification can provide a method for
searching for the image and far retrieving all image data on a particular
patient regardless of the image device which is used to create the images.
It is an object of the invention to provide a method of archiving image
data which reduces the required storage capacity necessary for storing
large amounts of textual information and which facilitates ease of
retrieval of the image data.
It is another object of the invention to provide a page-by-page image
storage system.
It is another object of the invention to provide a system for storing
textual information with image data in a manner which is transparent to
the technologist and which avoids the requirement of dual entry of
identifying information by a technologist.
It is yet a further object of the invention to provide universal textual
identification on each page of image data which is stored.
It is yet a further object of the invention to store images in the same
format as seen by radiologists or other medical personnel for the purpose
of diagnosis.
It is yet a further object of the invention to provide a system which is
easily adaptable to many types of image generating devices manufactured by
a variety of vendors.
These and other objects of the present invention will be more fully
understood from the following description of the invention with reference
to the illustrations appended hereto.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic block diagram of the image generating apparatus and
the workstation used when performing a study on a subject.
FIGS. 2, 3, and 4 are schematic diagrams of the page-by-page image storage
system of the present invention.
FIG. 5 is a schematic diagram of the method of constructing an image page
for the page-by-page image storage system of the present invention.
FIG. 6 is a flowchart of one aspect of the method of the present invention.
FIG. 7 is a flowchart of another aspect of the present invention including
the image storage and automatic header-extraction process of the present
invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
As used herein, "subject" includes members of the animal kingdom including
human beings or inanimate objects subjected to imaging techniques.
For convenience of disclosure, the present invention is described with
reference to the medical imaging field. However, it is equally applicable
for use in the field of industrial non-destructive testing.
A typical image generating system with associated peripheral devices is
shown in FIG. 1. A conventional image-generating device 3, such as a CT or
MRI, for example, emits the required excitation, radiation, ultrasound
waves or other method of generating an image towards a subject 5, as
schematically illustrated by arrow 7. The image generating apparatus 3
receives image data signals as schematically illustrated by arrow 9. Image
data is sent to image processor 15 which processes the data and sends a
video signal to be displayed on image display 17. A technician's
workstation which includes user interface 19 has suitable means for
allowing the technician to enter textual information identifying the
patient, the study to be conducted and the like. A print mechanism, such
as film printer 21, is also provided in order that images which are filmed
of subject 5 can be printed for later review by the radiologist or other
medical personnel. During the course of the study, a technician may
archive certain image frames by initiating the appropriate instructions
through user interface 19 to store the image information in storage
archive 23.
As mentioned hereinbefore, radiologists and other medical personnel
generally prefer to view images on film sheets such as those described
above. However, when the images are being acquired by the technician they
are conventionally stored on an image-by-image basis. When the technician
then proceeds to print a film of images, the image scheme of twelve,
sixteen or twenty images which appear on a page of film must be
reconstructed from the individual images.
In accordance with the present invention, it is preferred to store image
data on a page-by-page basis. Referring to FIGS. 2 through 4, page 40 is
shown in FIG. 2, page 42 is shown in FIG. 3, and page 44 is shown in FIG.
4. In accordance with this aspect of the present invention, the separate
images 50 though 72 (FIG. 2) are stored on a single page. For example,
each of the images 50 through 72 may consist of 512 lines with 512 pixels
per line, and the page 40 may consist of the 3 by 4 array of small images
shown in FIG. 2. As would be understood by those skilled in the art, the
page 40 is constructed in digital memory or on computer disk by repeatedly
concatenating lines from the small images, in the proper order, as shown
schematically in FIG. 5.
Referring to FIG. 5, box 75 represents schematicaly the memory of storage
archive 23 (FIG. 1). The numerals above the internal boxes represent the
relative pixel locations in the memory of storage archive 23. The first
line of image 50 is placed in memory first at, for example, pixel
locations 0 to 511, followed in order by line 1 of image 52 which is
stored at pixel locations 512 to 1023, and line 1 of image 54 which is
stored at pixel locations 1024 to 1535. Thereafter, preferably line 2 of
image 50 (pixel locations 1536 to 2047) is stored, and then line 2 of
image 52 (pixel locations 2048 to 2559) is stored. This process continues
as each line of images 50 through 54 is recorded. This is schematically
indicated with the three periods designated by reference character 77.
Then the process is continued with images 56, 58, and 60 with the first
line of image 56 following immediately after the final line of image 54,
and this point is designated by reference character 79 in FIG. 5 which
corresponds with pixel location 786432. The remaining images (62 through
72) are recorded similarly. One skilled in the art would readily
understand how to write appropriate software to perform these steps.
When a radiologist later decides to review the films, the images are
retrieved on a single page such as page 40 (FIG. 2) and can be immediately
printed in this format without the necessity of having the technician
reconstruct the page format.
In order to further increase the efficiency of storage and retrieval of the
page-by-page system, the present invention further involves an automatic
method of extracting textual identification information and storing this
information with the images to which it relates. As explained
hereinbefore, the conventional image-by-image archival system maintains an
entry in its data base for each of the individual small images. The size
of such a data base grows rapidly and it becomes progressively more
difficult for data to be retrieved. To avoid this problem, the present
invention includes storage of the textual identification information on
the first page of images in the page-by-page system. This reduces the size
of the data base by a factor of about twelve.
This also serves to reduce errors and to allow errors which are made to be
corrected. More particularly, in present practice, a major contributor to
errors in the retrieval of image information results from entry errors
occurring at the time when an examination is performed. When the
technologist begins a study, the technologist enters information such as
patient name, social security number, and other relevant information. This
same information must be entered yet again when each image is archived.
Errors often occur during this dual entry process. The probability that a
case, which is archived under the wrong header, is found upon later
request decreases significantly. In general, the entry errors cannot be
corrected even when recognized.
The present invention involves an automatic creation of a standard header
which is then embedded preferably in the first archived page such as pages
40 of FIG. 2. If desired, the header can be embedded in each page. The
header is verifiable and correctable. The method is also entirely
transparent to the technologist. More specifically, with reference to the
flowchart of FIG. 6, the study is started with an appropriate command as
indicated in block 80 of the flowchart. Generally, the technologist at the
beginning of each filming task enters text header information associated
with the examination. This feature is typically available on the scanners
currently manufactured. This step is shown in the flowchart of FIG. 6 at
step 82. As the textual information is entered, a text frame is created as
shown in block 84. This text frame is displayed as an image on the image
display 17 of the image generating apparatus 3 (FIG. 1).
At this point the text frame image is generated by a text frame analog
video signal. In accordance with the present invention, the text image is
then "frame-grabbed" or digitized as indicated in the block 88 of FIG. 6.
Preferably, this digitization is initiated automatically and transparently
when the technician requests the printing of the text image while it is
being displayed. This occurs without any need for the technologist to
perform any task or enter any instructions beyond current normal operating
procedure, and in fact, this will occur without the technologist even
being aware of it occurring. The digital image information representative
of the text image is then translated to character information by a
computer character recognition program using conventional techniques based
on convolution of the text image data with a digital description of the
character font. The character recognition software is guided by a template
which is customized for each particular image acquisition system to
include a general description of the format of the text display. This
program will then recognize the character information. Briefly, text is
recognized when pixel portions of a frame are examined for light and dark
intensity values. The program is designed to recognize patterns as certain
characters as would be understood by those skilled in the art. The program
as used in accordance with the present invention preferably will recognize
patterns corresponding to patient name, social security number,
physician's name and the study being conducted. Of course, any other
information desired could also be included.
As stated, the text image is translated into character information as
indicated in block 90. In other words, a text file is created. From this
text file, needed information is extracted and a first header of text
information is created as designed by block 92. This header will serve as
a directory address for the archival systems as well as a message embedded
in the first page of images of a study. For example, page 40 (FIG. 2) is
the first page of images for a particular study. The text header is
preferably stored as the first image block, 50. As shown in block 94 (FIG.
6) the textual information is stored with the images. It is then used both
as a directory address for retrieving the image information as well as a
message embedded in the page. If a study includes three pages of images,
it is preferred to include the text image only on the first page of the
image study. Then, when the images are printed on film the top left-hand
corner of the first page will contain all information related to the
patient and the study. It should be understood that this process is
entirely transparent to the technologist. Further, it is not vendor
specific, so it can be used with a variety of image-generating machines.
The frame-grabbed text image may be included without modification as the
text image on the image page, block 50 of page 40 for example, or
alternately, the character data extracted from the frame-grabbed image may
be translated back to image data in a standard format to create a new text
image block for the page. In order to view the images on image display 17,
the digital image information is converted back to an analog video
original as would be readily understood by one skilled in the art.
In order to facilitate convenient retrieval of image information whether it
be through CT, MRI, digital angiography, gamma ray or x-ray techniques, it
is preferred to have a universal format for indexing and identifying each
image page. Therefore, the present invention also provides a method for
identifying each image page. When the text frame is "frame-grabbed" in
addition to constructing the header to be inserted in the top left image
block in any page such as page 40 of FIG. 2, a second brief header of
image information will also be extracted. Preferably, this information
will be stored on all pages of images. For example, one examination may
involve three pages such as the pages 40, 42 and 44 as shown in FIGS. 2, 3
and 4, respectively. Each page 40, 42 and 44 has, in accordance with this
aspect of the invention, a second header embedded in lower portion 96 on
page 40 (FIG. 2), lower portion 97 on page 42 (FIG. 3) and lower portion
98 on page 44 (FIG. 4). This information is located at an identifiable
pixel location on the image page. The information which is embedded on
each page may be information such as date of examination, patient name and
social security number. This text information will appear on a print of
the film or on a display screen such as image display 17 (FIG. 1). This
avoids the common problem arising when films become separated from the
file jacket are then not identifiable. Furthermore, this information may
be used to retrieve all image data on a particular patient regardless of
the image device which is used to create the image.
Referring now to FIG. 7, this aspect of the method of the present invention
will be described in further detail. In accordance with the method of the
present invention, a signal will be generated to indicate the beginning of
a study, as shown in block 100 of FIG. 7. Textual information will have
been entered by the technician identifying the study. This textual
information will appear as a text frame on image display 17 (FIG. 1). This
text information is "frame-grabbed" or digitized to digital image
information as shown in block 104.
The present invention involves translating the image information to
character information, 106. A text file is thereby created as shown in
block 108. The text file may be searched and predetermined character
information is selected from the text file or the entire text file may be
used as the text header as depicted in block 110. This information is the
header which is then stored and embedded in the top left hand image block
of a first page of images for the study as shown in block 112. Images are
then obtained, 114. The images are arranged in pages in the manner
described hereinbefore as shown in block 116. The second header of
reference information is then retrieved from the character information in
the text file as shown in step 118. In step 120, this second header
information is inserted at a predetermined pixel location on each page.
The images are now stored in a page-by-page format. The first page of
images contains a full description of the patient and the study.
Additionally, all pages can be retrieved using the second header as a
directory address. Furthermore, the text headers also appear in human
readable form on the page itself. The second header appears at a
predetermined location on each page and it can also serve as a method of
identifying each separate page. This is helpful not only in identifying a
printed film, but additionally, this information could alternatively be
used to retrieve stored images.
It will be appreciated that the present invention provides efficient means
of archiving image information and a more efficient manner of retrieving
this information. The manner of archiving the images reduces the size of
the data base significantly. In addition, this system is transparent to
the technologist and thereby avoids dual entry and the greater risk of
typographical errors when dual entry is required. The text header
information can be corrected if necessary. The images are also archived in
the same format as preferred by radiologists. It should be appreciated
that the system is easily adaptable to essentially all available image
generating devices by all manufacturers.
It should also be understood that the invention has application outside the
medical field. For example, the invention is useful in connection with
imaging for purposes of non-destructive industrial testing.
Whereas particular embodiments of the present invention have been described
above for purposes of illustration, it will be appreciated by those
skilled in the art that numerous variations of the details can be made
without departing from the invention as described in the appended claims.
* * * * *
|
|
 |
|
 |
Description |
|
