WO1993003449A1

WO1993003449A1 - Medical image archiving system and storage method

Info

Publication number: WO1993003449A1
Application number: PCT/US1992/006209
Authority: WO
Inventors: Marc Boroditsky; Kenneth Davenport; John Schumann
Original assignee: Novus Technologies, Inc.
Priority date: 1991-08-05
Filing date: 1992-07-31
Publication date: 1993-02-18

Abstract

An archiving system stores discrete medical radiological images on a record card (60) having an imaging surface thereon. The apparatus (20) for storing discrete medical images comprises means for receiving (28) a film writing device output signal from a medical imaging device (22), and means for capturing a single image frame from the output signal. The captured single image frame is digitized and associated with a selected location on a record card at which the digitized single frame is to be permanently stored by a processor/memory (32). A plurality of the digitized single frames and their respective permanent storage locations are temporarily stored in the processor/memory (32). The digitized frames are reversibly compressed, and permanently recorded (46) at the individually selected, discrete locations of the record card (60). The images may be later recalled from the record card (60), decompressed, and viewed on a monitor or written to film (36). These operations are all under the control of a graphical user controller (26).

Description

Medical Image Archiving System and Storage Method

Technical Field

This invention relates to the storage and management of information, and, more particularly, to the capture, transmittal, and archiving of medical images.

This application is a continuation-in-part of United States application Serial Number 07/741,256, filed August 5, 1991, for which priority is claimed.

Background Art

Medical imaging plays an important role in evaluating the health of patients and detecting and diagnosing any abnormalities. Medical imaging includes radiology and radiography, and includes, for example, X-ray imaging, sonography, nuclear medicine, computed axial tomography (CT) scans, magnetic resonance imaging (MRI) , and invasive and motion studies. In these procedures, images on film of the body are prepared and retained for study, to satisfy legal requirements, or as otherwise might be required.

Some medical procedures produce large amounts of film. As an example, the typical CT scanner performs 12-15 studies per day, generating 4 to 6 sheets of 14 inch by 17 inch film per study. Some radiology centers have several CT scanners and MRI machines, each producing on the order of 75 sheets of film per day. As more efficient machines are developed, even more films can be generated.

The storage of large amounts of medical film is costly and uses valuable space. It has been the experience of some radiology centers that the storage requirements for archiving (long term storage) of exposed medical films soon exceeds the storage capacity of the center. Off-site storage facilities must be used, and films must be retrieved from the archives when needed for viewing.

Moreover, current medical practices may require the patient to see a number of doctors. Each doctor requires the imaged information, as well as other aspects of ^'the patient^■s medical history. The patient may be required to carry ever-increasing volumes of films and other information to each doctor, or the material must be shipped from doctor to doctor. The misplacing, deterioration, or loss of medical films is an increasing problem, possibly leading to serious diagnostic and legal problems.

There have been several proposals to alleviate the shortage of storage space, and the transport and retrieval problems. Some have been based upon the approach of digitizing the images and storing them in a digital magnetic memory. This approach has the important drawback that in many states such magnetically stored images are not admissible in court proceedings because they can be readily altered. Thus, while magnetic storage of radiologic images has the potential to store large volumes of information, it may not be practical in medical applications due to the legal limitations.

Optical disks have been considered for storage of medical images. "Write once-read many" type optical disks meet legal requirements, because the digital information is permanently burned into the disk and cannot be altered. However, optical disk reading/writing equipment is so expensive that it may not be practical for each doctor's office so that the medical record can be supplemented as necessary.

There is a need for an improved approach to the archiving of medical images in a permanent form that is legally admissible in court proceedings. The improved approach must utilize space efficiently, and permit ready writing to and reading from the storage medium. The present invention fulfills this need, and further provides related advantages.

Disclosure of Invention The present invention provides an apparatus and method for the computerized capture of medical images from digital imaging equipment such as CT, MRI, and ultrasound, the output of these images into hard copy (film) , and the permanent storage of these images in digital form. The apparatus is compatible with a wide variety of medical image- producing devices, and is relatively inexpensive. The record card that archives the images is small in size, is readily transported, and can be used for both images and other aspects of the patient's medical history.

In accordance with the invention, an apparatus for storing discrete medical images, comprises means for receiving a film writing device output signal from a medical imaging device, and means for capturing a single image frame from the output signal. The apparatus further has means for digitizing the captured single image frame, means for selecting and associating with the digitized single image frame a location on a record card at which the digitized single frame is to be stored, means for storing a plurality of the digitized single frames and their respective permanent storage locations, and means for reversibly compressing the information of the image of the digitized single frame. There is also means for permanently recording the plurality of the compressed frames at the individually selected, discrete locations of the record card. More specifically, an archiving system stores discrete medical radiological images on a record card having an imaging surface thereon. The apparatus includes means for receiving a film writing device output signal from a medical imaging device. The output signal can be either analog (video) or digital. In the case of an analog output signal, a frame grabber selectively captures a single frame from the output signal. In the case of a digital output signal, no frame grabber is required, and the signal is captured directly. A digitizer receives the single frame from the frame grabber and provides a digitized frame output. (The frame grabber and digitizer are available as a single unit commercially.) A controller permits an operator to identify a location on the record card at which the digitized single frame is to be stored as a discrete record. Before the frame is recorded on the record card, it may be stored in a memory having a capacity sufficiently large to store a plurality of single frames. As the frame is transferred to the record card, a data compressor reversibly compresses the digitized single frame using differential pulse code modulation. A card writer writes a digitized compressed frame onto the imaging surface of the record card responsive to the location selected using the controller. The record card permits a large number of compressed frames to be stored thereon.

Each digitized single compressed frame is also controllably exposed onto a sheet of X-ray film to generate hard copy of the image identical in reconstituted content to that recorded onto the record card. A controller permits an operator to identify a placement location for a particular image on a current sheet of film. CT, MRI, and ultrasound nuclear medicine images are commonly recorded in a 512 x 512 pixel matrix configuration. Each sheet of film may typically have between l and 20 images, placed in a column/row configuration. The operator incrementally selects the location of each discrete image on the sheets as they are generated by the imaging equipment. The apparatus' controller permits the operator to simultaneously identify the image location on the film as well as the record card. The controller is graphically represented on a computer monitor with software which allows the operator to select the number of images per sheet, expose desired images in a selected location on the sheet, save the images to the record card, and print the images to film.

The present invention is concerned primarily with the capture, storage, and management of radiologic information. "Radiologic" devices are those that capture an image directly in digital, or digital converted to analog, form, and include, for example, CT, MRI, ultrasound, and nuclear medicine methodologies. "Radiographic" devices, on the other hand, first record the image in an analog manner, such as on film. The film can later be scanned to produce a digital signal, but there may be degradation of image quality during the scanning operation. A conventional X-ray recorded on film is an example of a radiographic process♦

Because the apparatus is structured to receive and use the film writing device output signal from a medical imaging device, it can utilize feeds from a variety of different imaging devices, such as various brands of CT scanners, MRI units, ultrasound devices, etc. The apparatus has analog-to-digital conversion capability, so that images can be stored and recalled on demand for output to film recording devices. Thus, the apparatus also provides for output of the plurality of images to a laser film camera, other film recording device, or a multimodality unit (MMU) that receives inputs from several different input devices.

The invention further encompasses a method for storing discrete medical images generated by a medical radiologic device on a record card having an imaging surface thereon. The method comprises the steps of furnishing a film writing device output signal from a medical imaging device, capturing a single image frame from the video feed, and digitizing the captured single frame. Additional steps include selecting and associating with the digitized single frame a location on a record card at which the digitized single frame is to be stored, temporarily storing a plurality of the compressed single frames and their respective permanent storage locations, compressing the information of the image of the digitized single frame, and permanently recording a selected group of the stored frames at the individually selected, discrete locations of the record card.

The present invention provides an important advance in the art of medical image management, control, and archiving. The present approach permits a large number of images to be recorded on a laser record card the size of a standard credit card. The card stores the images in an easily searched format. The facility that produces the images can keep a copy of the record card as a permanent record, and a copy can be furnished to the patient on the conveniently carried card. Other true copies can be produced easily and inexpensively, as needed. The record card can be kept in the patient's file, and separate storage facilities are not required. Other features and advantages will be apparent from the following more detailed description of the preferred embodiment, taken in conjunction with the accompanying drawing, which illustrates, by way of example, the principles of the invention.

Brief Description of Drawings

Figure 1 is a block diagram of the apparatus of the present invention, which reflects the method of the invention.

Best Mode for Carrying Out the Invention Figure 1 illustrates a medical archiving apparatus 20, incorporating the features of the present invention. A medical radiologic image is received by the apparatus 20 from a radiologic source 22, which is typically a CT scanner, an MRI device, an ultrasound device, or a nuclear medicine device. As used herein, an "image" is a final graphic representation of clinical diagnostic information as read by a physician, in either digital or analog form. That the graphic information is in such a final, processed form is significant. There is no capability to alter the image so that its graphic representation may be changed. The image is not raw, unprocessed information, so that the final image could vary depending upon the selection of a different operator than the person operating the source 22. Thus, the apparatus 20 receives a film writing device output image signal or feed 24 from the medical imaging device 22. That is, the apparatus 20 operates with an image that is suitable for output directly to a film writing device such as a laser camera. The image feed 24 is levelled and adjusted for optimal contrast by circuitry in the source 22. Commercial CT scanners and MRI systems, and comparable systems, have this capability built in. The source 22 is not a part of the present invention, and it is important only that the apparatus 20 is adapted to receive such a film writing device output image signal 24.

The image carried on the feed 24 from the source 22 is viewed by a person operating the system, such as a radiologist or technician, on a monitor of an operator controller 26. Utilizing the operator controller 26, the operator commands an analog (video) or digital interface board 28 within the apparatus to capture the selected image frame 24. (The software for operating the controller 26 is attached as Exhibit A. ) In the case of a video interface, the analog signal is captured by an ISA AT compatible interface board capable of digitizing and displaying video from the source 22. Video input can be programmed to capture imagery at 256 gray shades with alphanumeric overlay at bandwidths greater than 250 MHz (megahertz) . Video timing on the digitization is fully programmable and provides a pixel clock frequency range of between 0.25 and 150 megasamples per second. Display resolution is variable up to a 2048 x 2048 pixel matrix on the host operator controller 26. In the case of a direct digital interface, the digital signal is also captured by an ISA AT compatible interface board capable of receiving digital data from the source 22 at a rate of 4 MB (megabytes) per second along a parallel RS 485 communication link. The video or digital interface board 28 provides a number of digital signal processing capabilities including image sub-sampling, replication, bi¬ linear interpolation, rearrangement, and histogramming.

After the selected frame is captured and digitized (if an analog feed signal) in some preselected format, such as 256 x 256 or 512 x 512 pixels, the single frame is moved off of the interface 28 along the host processor BUS and into random access memory (RAM) 32. Host transfer rates along the BUS are 3 MB per second. Enough RAM is provided to temporarily store a plurality of the image frames.

The operator designates, utilizing the operator controller 26, an eventual storage location of each selected frame on both a current sheet of film being generated, numeral 36 and a record card, numeral 60, and that location is stored with the respective digitized frame in the host processor/memory 32. The film 36 records the information in the conventional matrix row/column format. The record card records the information in a series of linear tracks on the card. In the preferred format, the image information is stored first in order on the tracks, and any database information is stored last in order on the tracks. As an example, if the record card is sized to hold 80 images at predefined discrete image locations along the tracks numbered L1-L80, the operator may designate that a particular captured frame is to be written into any particular location, such as location L6. That location is stored in the label of that particular captured frame, and it is later written into the designated location L6 of the record card. After each image location of the record card 60 is selected for later use, it is indicated to the operator as no longer available for subsequent selection.

Normally, a group of images large enough to fill a designated sheet of film is first identified and placed into host processor/memory

32 prior to placement on the record card or generation into film. Also placed into memory is the location identifier for each selected frame and any other identifying information required. This permits the operator to review the selected images on the controller 26 before the actual permanent writing operation. The operator may choose to delete earlier selected frames in favor of later selected frames. Prior to the selected group of images being committed to the designated sheet of film, the operator may change which images are selected for permanent recording, and the locations on the record card at which they are to be permanently recorded. However, the operator cannot change the content of the image. Alternatively, the operator may permanently write each image as it is selected to the record card 60.

The commitment of images to conventional film 36 takes place through a direct transfer of the digital image frame(s) to a laser camera 38 via a digital interface 40. The digital interface board 40 sends images at a 0.05 to 4 MB per second programmable burst rate, and supports all major manufacturers of laser cameras 38. Utilizing the camera interface board 42, the operator controls a variety of camera functions on a bi-directional RS 232 serial communication link with a data rate of 80 to 9600 Baud. Camera interface functions include: Setting of printer parameters, "new sheet", "expose image" (position) , "print film", query of printer status, "erase image", "open study", and "close study". (In Figure 1, the output to the laser camera 38 is shown as first passing from the apparatus 20 through an optional multimodality unit (MMU) 44.) A film 36 might be prepared selectively with one, several, or all of the images stored in the memory 32 for a particular study, but is usually prepared with all of the images. In a typical application, the film 36 version of the captured images may be prepared for immediate use by a physician, and record cards 60 would be prepared for the patient and for the file. The operator controller 26 includes a graphically represented "keypad" 27 which allows the operator to capture images from the source 22 and place them into a desired location on a sheet of film, and indicate which images are to be saved to the optical card. The controller presents the operator with a column/row matrix format which emulates the current sheet of film, and which can be sized in varying formats (e.g., 1, 2, 4 , 9, 12, 15, or 20 images per sheet) at the operator's discretion. When the operator captures a selected image from the source 22, a sub-sampled "thumbnail" representation of that image appears in the chosen column/row location, enabling the operator to preview the sheet of film prior to its printing. The operator also has the ability to call any one of the thumbnail images, from current or past sheets, onto the screen in full resolution for on-screen viewing. The operator controller 26 also has a complete database on both a sheet-by- sheet and image-by-image basis of the images committed to the card and film during the study. In this way, specific images or sheets in the study can be called into the film matrix by the operator on demand. At the time that the operator has composed the group of images that are to be committed to the designated sheet of film they are moved along the host processor BUS to the digital interface 40 for output to film. At the same time, a copy of these image frames remains in the host processor/memory 32 where they are compressed and written to the record card 60. The writing to the record card 60 is accomplished via the writer/reader unit 46. The record card 60 is placed into the writer/reader unit 46, and a permanent record of the image frames is formed on the surface of the record card 60. In the preferred approach, the record card 60 is an optical data card termed a LaserCard, which is manufactured, by Drexler Technology of Mountain View, CA. The preferred optical data card 60, as well as the optical card writer/reader unit 46, are available commercially, and are described in US patents 4,542,288; 4,284,716; and 4,544,835, whose disclosures are incorporated by reference.

Briefly, part of the surface of the optical card 60 is covered with a medium that is sensitive to laser light of a particular frequency. A laser in the writer/reader unit 46 writes a pattern of digital pulses into the medium. When the laser beam impinges on the optical recording layer, it creates holes in the layer called "pits". When the laser beam is used for reading the card, the ratio of reflected light (reflectivity) varies, depending upon whether the beam impinges on holes (written data reflectivity) or an unwritten area of the recording layer (background reflectivity) . The differences in these reflectivities are picked up by a sensor and converted into signals. The pattern formed in the medium of the optical card 60 is permanent. It can be written only once, but can be read an indefinite number of times. The life of the optical card is expected to be at least 30-100 years, so that images can be stored for extended periods on the medium. The optical card 60 also has space thereon for recording comments on the images in the image locations, or other aspects of the patient's medical history.

Currently available optical cards 60 have a permanent storage capacity of 2.86 megabytes of information (with error correction) . The smaller the storage requirements of each image, the more images that may be recorded onto each optical card. In the discussion earlier, the number of image locations was indicated as 80. However, in a particular situation it may be desired to write a larger or smaller number of images onto the medium surface of the optical card 60. To write many images onto the optical card 60, it is desirable to compress the information in the images to reduce redundancy, thereby reducing the storage space required for each image and permitting more images to be stored. If the images are digitized in a 512 x 512 format using the apparatus 20 of the present invention, about 80 images may be archived on each record card 60. To this end, as each image is committed under user control from the host processor/memory 32 to the record card writer/reader unit 46 to be permanently written to the record card 60, it is processed by an image compressor (which also may be operated in the reverse direction to be an image decompressor, when images are to be read from the record card 60) . The image compressor resides in the host processor 32, and encodes each discrete image frame to reduce pixel redundancy. The user can select on the controller 26 a desired level of compression. The procedure analyzes an inputted information set, and produces an output that contains the same, or nearly the same, amount of information but with reduced storage requirements. A number of image compression techniques are known in the art. Typically, the greater the degree of compression, the greater the loss of information or distortion of the image. A balance must be struck between the degree of compression and information loss and distortion. In the context of medical radiological information, the degree of loss of information and distortion of the image must be small, both for medical and legal reasons. Within this limitation, however, any of the known image compression techniques is acceptable for reducing the storage requirement for each image. The image compressor preferably utilizes a combination of run length coding and differential pulse code modulation (DPCM) , an effective approach for compressing medical images that typically include large dark areas at the perimeters of the image. The run length coding identifies those areas at the perimeter of the image which are essentially of uniform black appearance in a highly compressed form using only a few bytes of information. The perimeter of the actual image is indicated by the beginning and end location of each non-black portion of the image on each raster line scan of the image. As to the actual image within this defined perimeter, the differential pulse code modulation technique permits a high compression ratio without noticeable impact on the utility of the image after subsequent decompression for viewing. Differential pulse code modulation is known in the art and is described in the following patents: 4,758,883; 4,725,885; 4,663,660; and 4,369,463, whose disclosures are incorporated by reference. Differential pulse code modulation avoids the blocky effect that results from some digital compression techniques. Moreover, it is well suited for medical imaging, where much of the image is monotone black.

The apparatus 20 may be used to read a card 60 prepared in a previous study. The card 60 is retrieved from a filing system and placed into the write/read device 46. A "Read Card" function on the operator controller 26 is employed to bring both images and database data on the study into the controller 26. Once the reading is completed, the operator can call into the film matrix any single sheet or any desired combination of images from the prior study for thumbnail viewing, selected full resolution viewing of single images, and printing to film. In addition, the operator controller utilizes a "critical image" subset function to identify those images deemed of greatest diagnostic use. Images can be tagged as critical either at the time they are captured, or later when the radiologist has rendered their findings on the study. "Critical images" may be brought into the film matrix immediately upon completion of the card reading process, if desired.

The apparatus of the present invention may be constructed in part by assembling commercially available hardware elements. The video or digital interface 28 is implemented as two distinct boards available as the DI Receive and VI from adaptive

Video, Inc. The digital output and camera interface are represented by the DI transmit and CI boards, also from Adaptive Video, Inc. The memory 32 is preferably single inline memory modules. The operator commands are preferably generated by a personal computer programmed for

"mouse" controlled operation with images and a location map of available space on the record card 60 being displayed on the monitor of the controller 26. The preferred optical card writer/reader 46 is a Nippon Conlux Model LC304. The preferred record card 60 is a LaserCard, manufactured by Drexler Technology, Mountain View, CA. The preferred laser camera 38 is a 3M Lasercam or a GE Laser camera Model 952.

The present invention provides an important advance in the art of medical image archiving technology. Images can be permanently stored and readily retrieved, from a record card that is inexpensive and achieves a reduction in volume of about 100 times in physical storage space requirements. Although particular embodiments of the invention have been described in detail for purposes of illustration, various modifications may be made without departing from the spirit and scope of the invention. Accordingly, the invention is not to be limited except as by the appended claims.

Claims

1. Apparatus for storing discrete medical images, comprising: means for receiving a film writing device output signal from a medical imaging device; means for capturing a single image frame from the output signal; means for digitizing the captured single image frame; means for selecting and associating with the digitized single image frame a location on a record card at which the digitized single frame is to be stored; means for temporarily storing a plurality of the digitized single frames and their respective permanent storage locations; means for reversibly compressing the information of the image of the digitized single frame; and means for permanently recording the plurality of the compressed frames at the individually selected, discrete locations of the record card.

2. The apparatus of claim 1, wherein the means for capturing is a frame grabber.

3. The apparatus of claim 1, wherein the means for compressing includes means for performing run length coding of an image.

4. The apparatus of claim 1, wherein the means for compressing including means for performing differential pulse code modulation of an image.

5. The apparatus of claim 1, wherein the means for temporarily storing is a random access memory.

6. The apparatus of claim 1, wherein the means for permanently recording includes a laser writer that records a digital image at the selected location of the record card.

7. The apparatus of claim 1, wherein the means for reversibly compressing further includes means for decompressing a previously compressed image.

8. The apparatus of claim 1, wherein the apparatus further includes a monitor upon which an image may be viewed.

9. The apparatus of claim 1, wherein the means for reversibly compressing is user controlled, with controllable compression levels.

10. Apparatus for storing discrete medical images on a record card having an imaging surface thereon, comprising: means for receiving a film writing device output signal from a medical imaging device; a frame grabber that selectively captures a single frame from the output signal; a digitizer that receives the single frame from the frame grabber and provides a digitized frame output; a controller including means for permitting an operator to identify a location on the record card at which the digitized single frame is to be stored as a discrete record; a memory having a capacity sufficiently large to temporarily store a plurality of single frames; a data compressor/decompressor that selectively compresses the digitized single frame using run length coding and differential pulse code modulation and selectively decompresses a previously compressed image; and a card writer that permanently writes a digitized compressed frame onto the record card responsive to the location selected using the controller.

11. A method for storing discrete medical images generated by a medical radiologic device on a record card having an imaging surface thereon comprises the steps of: furnishing a film writing device output signal from"a medical imaging device; capturing a single image frame from the output signal; digitizing the captured single frame; selecting and associating with the digitized single frame a location on a record card at which the digitized single frame is to be stored; storing a plurality of the digitized single frames and their respective associated permanent storage locations; compressing the information of the image of the digitized single frame; and permanently recording the plurality of digitized single frames at the individually selected, discrete locations of the record card.

12. The method of claim 11, including the additional steps, following the step of permanently recording, of: reading an image stored on the record card; and decompressing the image.

13. The method of claim 12, including the additional step, after the step of decompressing, of recording the decompressed image on a film.

14. The method of claim 11, wherein the step of compressing includes the substep of: performing run length coding of the image.

15. The method of claim 11, wherein the step of compressing includes the substep of: compressing the image using differential pulse code modulation.