JP2009039411A - Quality evaluation method of radiation image, quality evaluation apparatus and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve the efficiency of quality evaluation regardless of the difference of physical features of respective QC phantoms. <P>SOLUTION: In the quality evaluation method of radiation images, image data for which the QC phantom in which an evaluation member having a size or a shape corresponding to an image quality evaluation item is inherent is photographed are acquired, the image data are binarized to generate a first binary image, a continuous area having a prescribed signal value is extracted from the first binary image as a first binary area, the feature amount of the first binary area is calculated, and a predetermined evaluation operation is executed to the first binary area on the basis of the feature amount of the first binary area. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a radiation image quality evaluation method, a quality evaluation apparatus, and a program.

  In recent years, in the medical field, a radiation image reading system has been used to obtain an image for diagnosing a medical condition of a patient. In the radiographic image reading system, a part of the radiation energy transmitted through the subject by radiography is temporarily stored inside the stimulable phosphor sheet, and the stimulable phosphor sheet in which the radiation energy is accumulated is two-dimensionally with excitation light. Scanning is performed to generate stimulated emission, and this stimulated emission is photoelectrically read to obtain an image signal, which is A / D converted to obtain digital image data.

  If the image quality of the image obtained in such a radiation image reading system is poor, an appropriate diagnosis cannot be performed. Therefore, there is an inspection method for performing QA (Quality Assurance) or QC (Quality Control) of an image reading apparatus or a cassette by an operator or a service person.

  As this inspection method, a radiographic image of a QC phantom is obtained by photographing a QC phantom in which a member that shields or absorbs radiation is arranged on a radiation transmissive substrate, and this radiographic image is read as an image to be inspected. There is known an inspection that is read by an apparatus and that performs an operation (hereinafter referred to as a QC operation) for obtaining a quality evaluation result on image data of a read radiation image.

When performing QC calculation on image data obtained from a radiation image of a QC phantom, it is necessary to specify the positions of various members included in the QC phantom.
As a technique for specifying the positions of the various members, for example, a radiographic image (grayscale image) obtained by photographing a QC phantom is processed, an image component is determined, and an image component and a predetermined image component are determined. When they match, the position of the mark is specified, and the region of interest (ROI) is calculated. That is, the mark has a mark for specifying the position of each member in the QC phantom, and is compared with a predetermined template. Discloses a technique for determining the positions of various members (see Patent Document 1).

Also, an IC for storing a marker for detecting the position of each part, identification information for identifying the type of the QC phantom, and various information related to the inspection member (calibration value information, position information of each part, etc.) A radiation image reading system using a QC phantom with a memory is disclosed (see Patent Document 2).
JP 2001-54514 A JP 2004-223138 A

However, since it is necessary to hold the position information of various members and the reference position information in the QC phantom, the production cost of the QC phantom is increased. Furthermore, since it is necessary to switch identification information and templates for each QC phantom, the number of inspection steps for quality evaluation increases.
In addition, when the QC phantom is damaged, only the user who knows that the QC phantom is damaged can not verify the evaluation result considering that the user is damaged. In some cases, the evaluation result cannot be verified in consideration of the damaged portion, and therefore there is a problem that the erroneous evaluation result is verified.

  In view of the above problems, an object of the present invention is to improve the efficiency of quality evaluation regardless of the difference in physical characteristics between QC phantoms.

  According to the first aspect of the present invention, an image acquisition step of acquiring a radiographic image obtained by photographing a QC phantom in which a first evaluation member having a size or shape corresponding to an image quality evaluation item is included, and the image A first binarization step for binarizing the radiation image acquired by the acquisition step to generate a first binarized image, and a first binarization generated by the first binarization step A first extraction step of extracting a continuous region having a predetermined signal value from an image as a first binarization region, and a feature amount of the first binarization region extracted by the first extraction step A first feature value calculation step to be calculated, and an evaluation operation associated in advance based on the feature value of the first binarized area calculated by the first feature value calculation step is the first binary value. A first radiological image comprising: Evaluation method is provided.

  According to a second aspect of the present invention, in the radiological image quality evaluation method according to the first aspect, the comparison result by the specifying step corresponds to the feature amount of any of the first binarized regions. An abnormality information generation step of generating abnormality occurrence information when there is no predetermined feature amount of the first evaluation member.

  According to a third aspect of the present invention, in the radiological image quality evaluation method according to the first or second aspect, the feature amount is a size and a shape for each of the regions, and the shape is for each of the regions. Of circularity, perimeter, or aspect ratio.

  According to a fourth aspect of the present invention, in the radiological image quality evaluation method according to any one of the first to third aspects, the QC phantom includes a second evaluation member, and the identification A region extraction step of extracting an image of a region in a predetermined positional relationship from the radiation image acquired by the image acquisition step with reference to any region where the corresponding first evaluation member is specified by the step; A second binarization step for binarizing the image extracted by the region extraction step to generate a second binarized image; and a second binarization generated by the second binarization step A second extraction step of extracting a second binarized region from the image; a second feature amount calculating step of calculating a feature amount of the second binarized region extracted by the second extraction step; , The second feature amount calculated by the second feature amount calculation step The feature value of the binarized region and the predetermined feature value of the second evaluation member are compared with each other, and the second evaluation member corresponding to the second binarization region is compared according to the comparison result. A determination step for determining presence / absence, and a second calculation for executing an evaluation calculation corresponding to the second evaluation member for the second binarized region in which the second evaluation member corresponding to the determination step is specified. And a process.

  According to the invention described in claim 5, the image acquisition means for acquiring a radiographic image obtained by photographing a QC phantom in which a first evaluation member having a size or shape corresponding to an image quality evaluation item is included, and the image A first binarization unit that binarizes the radiation image acquired by the acquisition unit to generate a first binarized image; and a first binarization generated by the first binarization unit. First extraction means for extracting a continuous area having a predetermined signal value from the image as a first binarized area, and feature amounts of the first binarized area extracted by the first extraction means A first feature value calculating means for calculating, and an evaluation operation associated in advance based on the feature value of the first binarized area calculated by the first feature value calculating means is the first binary value. A quality evaluation apparatus comprising: first computing means for executing on a conversion area It is provided.

  According to a sixth aspect of the present invention, in the quality evaluation apparatus according to the fifth aspect, the comparison result by the specifying unit is determined in advance corresponding to the feature amount of any of the first binarized areas. In addition, when there is no feature amount of the first evaluation member, abnormality information generation means for generating abnormality occurrence information is provided.

  According to a seventh aspect of the present invention, in the quality evaluation apparatus according to the fifth or sixth aspect, the feature amount is a size and a shape for each region, and the shape is a circularity for each region. , Perimeter or aspect ratio.

  According to an eighth aspect of the present invention, in the quality evaluation apparatus according to any one of the fifth to seventh aspects, the QC phantom includes a second evaluation member, and is handled by the specifying unit. A region extracting unit that extracts an image of a region having a predetermined positional relationship from a radiation image acquired by the image acquiring unit with reference to any region in which the first evaluation member to be specified is specified, and the region extracting unit A second binarization unit that binarizes the image extracted by step (b) to generate a second binarized image, and a second binarized image generated by the second binarization unit. A second extraction means for extracting two binarized areas; a second feature quantity calculating means for calculating a feature quantity of the second binarized area extracted by the second extraction means; Of the second binarized area calculated by the feature amount calculating means of 2 The collected amount and the predetermined characteristic amount of the second evaluation member are compared with each other, and the presence or absence of the second evaluation member corresponding to the second binarized region is determined according to the comparison result. A determination unit; and a second calculation unit that executes an evaluation calculation corresponding to the second evaluation member on the second binarized region in which the second evaluation member corresponding to the determination unit is specified. Prepare.

  According to the invention described in claim 9, the image acquisition means for acquiring a radiographic image obtained by photographing the QC phantom in which the first evaluation member having a size or shape corresponding to the image quality evaluation item is included, A first binarization unit that binarizes the radiographic image acquired by the image acquisition unit to generate a first binarized image; a first binary generated by the first binarization unit; First extraction means for extracting a continuous area having a predetermined signal value from the binarized image as a first binarized area, and a feature quantity of the first binarized area extracted by the first extracting means. An evaluation operation associated in advance based on the first feature value calculation means to be calculated and the feature value of the first binarized area calculated by the first feature value calculation means is the first binary value. Function as the first calculation means to be executed on the conversion area Because of the program.

  According to the present invention, since the evaluation calculation is performed according to the feature amount of the image of the evaluation member, there is no need to hold the reference position information for identifying the position of the evaluation member in the QC phantom. Even if the position of the evaluation member in the QC phantom is changed, it is not necessary to set new reference position information again, and the evaluation can be performed, and there is a variation in the position of each evaluation member for each QC phantom. Even if it exists, since evaluation of each QC phantom can be performed with high accuracy, the efficiency of quality evaluation can be improved regardless of the difference in physical characteristics of each QC phantom.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
First, the configuration will be described.
FIG. 1 shows an overall configuration diagram of the radiation image reading system A.
As shown in FIG. 1, the radiographic image reading system A includes an imaging device 10, a reading device 20, a console 30, an imager 40, and a service center 50. The reading device 20, the console 30 and the imager 40 are connected to each other via a communication network N1 so that data can be transmitted and received. The console 30 and the service center 50 are connected to each other via a communication network N2 so as to be able to send and receive data. Yes. The number of each device is not particularly limited.

  The communication networks N1 and N2 can apply various line forms such as a LAN (Local Area Network), a WAN (Wide Area Network), and the Internet. Note that wireless communication or infrared communication may be used as long as it is permitted within a medical institution such as a hospital. However, since important patient information is included, it is preferable to encrypt information to be transmitted and received. In general, the DICOM (Digital Image and Communications in Medicine) standard is used as a hospital communication method. In communication between the devices on the communication networks N1 and N2, the DICOM MWM (Modality Worklist Management) and DICOM MPPS (Modality Performed Procedure Step) are used.

Hereinafter, each component apparatus of the radiation image reading system A will be described.
The imaging apparatus 10 includes a radiation irradiation switch (not shown), a radiation source 11 such as an X-ray source, and the like, and performs imaging by irradiating the subject M with radiation. When photographing is performed with the photographing apparatus 10, a cassette 12 containing a stimulable phosphor sheet is disposed in a direction opposite to the radiation source 11 of the subject M.

  The cassette 12 is a portable radiation conversion medium in which a radiation image conversion sheet including a stimulable phosphor sheet that accumulates radiation energy is accommodated, and the subject M has a radiation dose from the radiation source 11 of the imaging apparatus 10. The radiation dose according to the radiation transmittance distribution is accumulated in the stimulable phosphor layer of the provided stimulable phosphor sheet, and the radiation image information of the subject M is recorded in this stimulable phosphor layer. Further, a cassette ID (cassette identification information) for individually identifying the cassette 12 is affixed to the surface of the cassette 12.

  When the cassette 12 containing the photostimulable phosphor sheet is loaded, the reading device 20 reads a radiation image recorded on the loaded cassette 12, and controls the entire device according to a program and performs various processes. CPU (Central Processing Unit) to be executed, ROM (Read Only Memory) for storing various programs and data, RAM (Random Access Memory), etc., control unit, operation unit, housed in or incorporated in the loaded cassette 12 A reading unit that reads a radiation image of the subject M recorded on the photostimulable phosphor sheet to obtain image data, an erasing unit that erases an afterimage on the photostimulable phosphor after reading by the reading unit, and a loaded cassette 12 A barcode reader that reads barcodes affixed to the computer, a communication unit that controls data transmission / reception with each device on the communication network N1, and the like Configured. The photostimulable phosphor sheet is a recording medium for recording a radiation image.

  The reading unit two-dimensionally scans the stimulable phosphor sheet in which the radiation energy transmitted through the subject M is stored with the excitation light, and generates the stimulated light emission. An image signal is obtained by reading and A / D conversion is performed to obtain digital image data. For example, a laser beam is used as excitation light, the laser beam is deflected in the main scanning direction and incident on the photostimulable phosphor sheet, and the stimulable phosphor sheet is scanned and conveyed in the sub-scanning direction. By scanning the photostimulable phosphor sheet two-dimensionally, propagating the photostimulated luminescence light with a light propagator and detecting it with a photodetector such as a CCD (Charge Coupled Device) or a photomultiplier. The radiation image recorded on the photostimulable phosphor is read.

  The console 30 is a photographing target (selection of a photographing menu (there is a QC menu for photographing for quality evaluation processing and a general photographing menu for photographing a patient)), registration of a cassette 12 and the like used for photographing, and photographing for general photographing. Registration of imaging order information such as patient information and imaging region of a patient, association and storage of image data read by the reading device 20 and imaging order information, image processing for image data, transmission of image data to the imager 40, and Further, in the present embodiment, it is an apparatus for transmitting and receiving various reports generated when quality evaluation processing is performed to the service center 50.

  FIG. 2 shows a functional configuration example of the console 30. As shown in FIG. 2, the console 30 includes a CPU 31, an operation unit 32, a display unit 33, a first communication unit 34, a second communication unit 35, a barcode reader 36, a RAM 37, a storage unit 38, and an HDD (Hard Disk Drive). 39 and the like, and each part is connected by a bus N3.

  The CPU 31 reads out various programs stored in the storage unit 38, develops them in a work area formed in the RAM 37, and centrally controls the operation of each unit by executing processing according to the program.

  The operation unit 32 includes a keyboard having cursor keys, numeric input keys, various function keys, and the like, and a pointing device such as a mouse, and sends an instruction signal input to the CPU 31 by key operation or mouse operation on the keyboard. Output. The operation unit 32 may include a touch panel on the display screen of the display unit 33. In this case, the operation unit 32 outputs an instruction signal input via the touch panel to the CPU 31.

  The display unit 33 is configured by a monitor such as an LCD (Liquid Crystal Display) or a CRT, and displays an input instruction, data, or the like from the operation unit 32 in accordance with an instruction of a display signal input from the CPU 31. In the present embodiment, the display unit 33 outputs various reports generated in a quality evaluation process, which will be described later, on the screen in accordance with an instruction of a display signal input from the CPU 31.

  The first communication unit 34 includes a LAN adapter, a router, a TA (Terminal Adapter), and the like, and controls communication with each device connected to the communication network N1. That is, the first communication unit 34 receives image data obtained by the reading device 20 via the communication network N1, and further receives the device ID of the reading device 20 via the communication network N1.

  The second communication unit 35 includes a LAN adapter, a router, a TA (Terminal Adapter), and the like, and controls communication with the service center 50 connected to the communication network N2. In other words, the second communication unit 35 outputs the generated various reports to the service center 50, and stores the various reports stored in the service center 50 via the communication network N2.

  The barcode reader 36 reads the cassette ID represented by the barcode attached on the cassette 12 and outputs it to the CPU 31.

  The RAM 37 forms a work area for temporarily storing various programs, input or output data, parameters, and the like that can be executed by the CPU 31 read from the storage unit 38 in various processes that are executed and controlled by the CPU 31.

  The storage unit 38 is configured by a nonvolatile semiconductor memory or the like, and stores various programs executed by the CPU 31 and data necessary for execution of processing by the program or data such as processing results. These various programs are stored in the form of readable program codes, and the CPU 31 sequentially executes operations according to the program codes.

  An HDD (Hard Disc Drive) 39 stores and stores the general image data transmitted from the reading device 20 in association with the imaging order information. Further, the HDD 39 includes the image data of the radiographic image obtained by photographing the QC phantom transmitted from the reading device 20, the device ID of the device that performed the reading, the cassette ID used for photographing, and the QC image data. QC identification information indicating, reading time (including date) information and the like are stored and stored in association with each other.

The HDD 39 has a QC shooting management information file 391, a general shooting management information file 392, and an image file storage area 393.
The QC shooting management information file 391 stores one record in association with the management number, device ID, cassette ID, QC identification information, reading time information, processed flag (default is OFF), and image ID for each QC phantom image data. As a file to be stored.
The general photographing management information file 392 is a file that stores the registered photographing order information and the image ID of image data photographed in accordance with the photographing order information in association with each other.
The image file storage area 393 is an area provided for storing and saving image data received from the reading device 20.

  The image data transmitted from the reading device 20 is an image file in DICOM format, and accompanying information including an image ID is attached. In the present embodiment, the image data is acquired based on the image ID using information related to the image data stored in the QC shooting management information file 391 and the shooting order related to the image data stored in the general shooting management information file 392. Associated with information.

The storage capacity of the HDD 39 needs to be prepared with a certain margin.
Usually, after imaging is performed and image data is received from the reading device 20 or the like, interpretation by a doctor is started within one week at the latest, or a quality evaluation process described later is executed. Accordingly, the storage capacity must be at least one week after the image data is received.

  For example, if the image data handled by the console 30 is 2000 pixels × 2500 pixels per image and about 100 images are generated per day, the HDD has a capacity of 40 GB (the storage capacity for image storage is 20 GB). 2000 images can be stored for about 20 days.

  In the radiation image reading system A, the QC tool 301 is mounted on at least one console 30. The QC tool 301 is an arithmetic program for quality evaluation, and the CPU 31 of the console 30 on which the QC tool 301 is mounted realizes quality evaluation processing by software processing in cooperation with the QC tool 301. The console 30 equipped with the QC tool 301 has a QC evaluation result file 394 that stores various reports generated by the quality evaluation process. Hereinafter, the console 30 on which the QC tool 301 is mounted will be described as a console 30a.

  The console 30a that executes the quality evaluation process by the QC tool 301 includes an image acquisition unit that acquires a radiographic image (image data) obtained by photographing a QC phantom (test simulation subject model) in which a plurality of evaluation members are included, and the image First binarization means for binarizing data to generate a first binarized image, and a continuous area having a predetermined signal value from the first binarized image as a first binarized area First extracting means for extracting, first feature amount calculating means for calculating the feature amount of each first binarized region, feature amount of each first binarized region, and each predetermined first And the first evaluation member corresponding to each of the first binarized regions according to the comparison result, and each of the first evaluation members specified by the first evaluation member is identified. The evaluation calculation corresponding to the first evaluation member is executed for one binarized area. If the comparison result by the first calculating means and the specifying means does not include a predetermined feature value of each first evaluation member corresponding to the feature value of any of the first binarized areas, abnormality occurrence information ( It is a quality evaluation device that functions as anomaly information generation means for generating an error report.

  Furthermore, when executing the quality evaluation process, the console 30a extracts an image of an area having a predetermined positional relationship from the acquired image data with reference to any area where the first evaluation member is specified. An extraction unit, a second binarization unit that binarizes the extracted image to generate a second binarized image, and a second binarization region that extracts a second binarized region from the second binarized image. 2 extraction means, second feature quantity calculation means for calculating the feature quantity of the second binarized area, feature quantity of the second binarization area, and a predetermined feature quantity of the second evaluation member And a determination means for determining the presence / absence of a second evaluation member corresponding to each second binarized area according to the comparison result, and each corresponding second evaluation member identified 2nd calculation which performs evaluation calculation corresponding to the 2nd evaluation member to 2 binarization fields Stage, to function as.

FIG. 3 shows an example of the QC phantom P.
The QC phantom P has a linearity and display dimensional accuracy of the reading device depending on a material that shields and absorbs radiation having different sizes or shapes between two transparent substrates such as acrylic plates. This is a quality evaluation subject in which evaluation members for image quality evaluation items such as sharpness, reading laser scanning property, and low contrast resolution are enclosed.

  As evaluation members, metal disks 1a to 1c for display dimension accuracy evaluation, edge pattern 2 for sharpness evaluation, step wedges 3a to 3d for linearity evaluation, jitters 4a and 4b for reading laser scanning performance evaluation, Burger phantoms 5a to 5j for low contrast resolution evaluation are used. In the present embodiment, the metal discs 1a to 1c, the edge pattern 2, the step wedges 3a to 3d, and the jitters 4a and 4b are used as the first evaluation member, and the burger phantom 5a has a lower contrast than the first evaluation member. ˜5j is a second evaluation member.

  The metal disks 1a to 1c are made of circular copper plates, and are arranged at three corners of the QC phantom P as shown in FIG. The image data of the metal disks 1a to 1c is used for quantitative evaluation of the display dimensional accuracy of the reading device.

  The edge pattern 2 is formed of a tungsten plate on which an edge portion having a sharp edge is formed, and the edge portion is arranged in a slightly inclined state with respect to the main scanning direction and the sub-scanning direction of reading by the reading device 20. ing. The image data of the edge pattern 2 is used for quantitative evaluation of the sharpness (MTF; Modulation Transfer Function) of the reading device.

  The step wedges 3a to 3d are calibrated with four rectangular copper plates having different thicknesses from each other step by step. The step wedges 3a to 3d shown in FIG. 3 are gradually reduced in thickness from the step wedge 3a toward the step wedge 3d, and are step wedges of radiation transmission amount. The image data of the step wedges 3a to 3d is used for quantitative evaluation of the linearity and dynamic range of the reading device.

  The jitters 4a and 4b are formed of a rectangular copper plate on which straight edge portions are formed, and one of the long sides of the jitters 4a and 4b is parallel to the main scanning direction of reading by the reading device 20. The other long side is arranged so as to be parallel to the sub-scanning direction. The image data of the jitters 4a and 4b is used for evaluating the scanning performance (scanning accuracy) of the reading laser in each scanning direction.

  The burger phantoms 5a to 5j are made of acrylic cylinders having a predetermined diameter and different heights (thicknesses). For example, as shown in FIG. They are arranged side by side so that the (thickness) gradually decreases (thin) and the diameter decreases. The images of the burger phantoms 5a to 5j are used for visual evaluation or quantitative evaluation of the low contrast resolution of the reading device.

  Note that the QC phantom that can be used in the present embodiment is not limited to the QC phantom P shown in FIG. 3, and may be anything that holds a plurality of types of evaluation members having different sizes and shapes.

  The imager 40 is a device that records an image on a recording medium such as a film based on the image data received from the console 30 and outputs a hard copy reproduced as a visible image. Examples of the imager 40 include a photothermal silver salt type imager that records a latent image by laser exposure on a transmission type recording medium (film) based on image data, and visualizes the latent image by thermal phenomenon processing, or an inkjet An imager or the like that records on a reflective recording medium (paper medium, sticker, or the like) by a method or a laser method can be used.

  The service center 50 is a device for accumulating various reports input from the console 30 via the communication network N2, and controls the entire device according to a program and executes various processings, a CPU (Central Processing Unit), various programs and data. Is stored in a non-volatile memory based on instructions from a control unit composed of a ROM (Read Only Memory) and a RAM (Random Access Memory), an HDD 51 for storing various reports, an operation unit, and an operation unit A display unit for displaying various reports, a communication unit for controlling data transmission / reception with the console 30 on the communication network N2, and the like are provided.

Next, the operation of the quality evaluation process in the radiation image reading system A will be described.
First, an image acquisition process for acquiring image data used for the quality evaluation process when the combination of the cassette 12 and the reading device 20 is a quality evaluation target will be described.

  Here, the quality of the image data obtained in the radiation image reading system A depends on a stimulable phosphor sheet for recording a radiation image, a recording medium such as an FPD, and a reading device for reading the image data from the recording medium. That is, when the cassette 12 that has been photographed is loaded in the reading device 20 and reading is performed, the quality of the obtained image data is determined by the combination of the cassette 12 and the reading device 20 to be used.

  The image acquisition process is executed by the reading device 20 and the console 30. The processing on the reading device 20 side is realized by software processing in cooperation with the CPU of the reading device 20 and a program stored in the ROM, and the processing on the console 30 side is a program stored in the CPU 31 and the storage unit 38. It is a process realized by software processing in cooperation with.

  In the console 30, a shooting menu screen is displayed on the display unit 33 in accordance with an operation from the operation unit 32 by an operator (such as a shooting engineer), and a QC menu is selected from the shooting menu screen. Then, the barcode of the cassette 12 to be evaluated is read from the barcode reader 36 and the cassette ID is input. Information indicating that the input cassette ID is an image for quality evaluation processing is registered, and its own console ID and the like are associated with each other and notified (transmitted) to all the reading devices 20 connected to the communication network N1. The

  In the reading device 20, when a cassette ID or the like is notified from the console 30 via the communication unit, the notified console ID, cassette ID, or the like is associated with each other and stored in the RAM. An operator attaches a jig to the cassette 12 for quality evaluation registered on the console 30 or places it in a predetermined position, and fixes the QC phantom at a predetermined position to perform radiography. Is called.

When the cassette 12 is loaded into any of the reading devices 20 to be evaluated, the cassette ID of the cassette 12 loaded by the barcode reader of the reading device 20 is read and associated with the cassette ID read in the RAM. The stored console ID and the like are retrieved.
Then, according to the information obtained by the search, the QC image recorded in the loaded cassette 12 is read, and the obtained image data includes an image ID for identifying the image, its own device ID, The cassette ID, reading time information, and the like of the read cassette 12 are attached as supplementary information, and are transmitted to the console 30 (console 30 of the console ID obtained by the search) that registered the cassette 12 via the communication unit.

  When the console 30 receives the image data and the accompanying information from the reading device 20 via the first communication unit 34, the associated information is stored in association with the received image data, and the process ends.

  In the image acquisition process described above, if a cassette defined for calibration is used as the cassette 12, the image data for evaluating the quality variation and the like of the reading device 20 with the reading device 20 as a quality evaluation target is used. Can be acquired. Further, if the reading device 20 determined for calibration is used as a device for reading the cassette 12, the cassette 12 is set as a quality evaluation target, and image data for evaluating the quality variation of the cassette 12 can be acquired. it can.

  With reference to FIG. 4, the quality evaluation process performed in the console 30a in which the QC tool 301 is mounted will be described. FIG. 4 shows a flowchart of the quality evaluation process. The quality evaluation process is a process realized by a software process in cooperation with the CPU 31 and the QC tool 301 and the like when an execution of the quality evaluation process is instructed by a predetermined operation of the operation unit 32.

First, QC phantom image data photographed from the reading device 20 is acquired via the first communication unit 34 (step S1).
That is, step S1 corresponds to an image acquisition step of acquiring image data as a radiographic image obtained by photographing a QC phantom in which a plurality of evaluation members corresponding to each image quality evaluation item is included.

  From the image data acquired in step S1, image data (hereinafter referred to as QC image data) of an area in which a QC phantom that is a quality evaluation target is copied is extracted (step S2). Note that the region extracted in step S2 can be set in advance because the QC phantom is photographed with a fixed position at the time of photographing.

  The signal value of the extracted QC image data is converted, and it is determined whether or not the QC image data can be extracted from the image data (step S3). If the QC image data cannot be extracted (step S3; No), the QC An error report (abnormality occurrence information) indicating that the image data could not be extracted is generated, the error report is stored in the QC evaluation result file 394 (step S4), and this process ends.

  When the QC image data can be extracted (step S3; Yes), a histogram is created from the QC image data subjected to signal value conversion (step S5). A threshold is determined based on the created histogram (step S6), the QC image data is binarized using the determined threshold as a reference, and a first binarized image is generated (step S7).

That is, one of two binary levels is assigned to each pixel based on whether the pixel value of each pixel exceeds a threshold pixel value.
For example, when the threshold pixel value is 145, 1 is set for a pixel having a pixel value of 200 (a pixel value darker than the threshold), while the pixel value is 100 (a pixel value brighter than the threshold). 0 is set for the pixel. Thereby, the pixels in the QC image data are divided into two areas, that is, an area where the evaluation member does not exist (pixel area set to 1) and an area where the evaluation member exists (pixel area set to 0). Can be classified into areas.

  Accordingly, steps S2 to S7 correspond to a first binarization step for generating a first binarized image based on image data as a radiographic image.

  The pixels set to 0 in the first binarized image are counted, and the total number of pixels set to 0, that is, the total area of the pixels classified into the region where the evaluation member exists is calculated. The calculated total area is compared with the predetermined planar total area (reference area) of the metal discs 1a to 1c, the edge pattern 2, the step wedges 3a to 3d, and the jitters 4a and 4b. It is determined whether the total area is within a predetermined range of the reference area (step S8).

  In step S8, the total planar area of the burger phantoms 5a to 5j is not added to the reference area because the burger phantoms 5a to 5j have a much lower contrast than the first evaluation member. This is because the pixels corresponding to ˜5j are set to 1 when the QC image data is binarized in step S7, that is, it is determined that there is no evaluation member.

  When the calculated total area is not within the predetermined range of the reference area (step S8; No), an error report indicating that the QC image data is not appropriate is generated, and the error report is stored in the QC evaluation result file 394 (step S4) This process is terminated.

  By comparing the total area calculated in step S8 with the reference area, it is determined whether or not the binarization process has succeeded, that is, the evaluation member to be detected (metal disks 1a to 1c, edge pattern 2, step It can be determined whether or not the areas of the wedges 3a to 3d and the jitters 4a and 4b) have been detected. For example, when the calculated total area is smaller than a predetermined range of the reference area, it is determined that any of the evaluation members to be detected is not detected, and the QC phantom is not installed properly or the QC phantom is damaged. Is determined to have occurred. If the calculated total area is larger than the predetermined range of the reference area, it is determined that a foreign object is reflected in the QC image data or that the QC phantom is broken. As described above, when the calculated total area is outside the predetermined range of the reference area, the QC image data is not data that can be subjected to normal quality evaluation. The result is generated as an error report.

  When the calculated total area is within the predetermined range of the reference area (step S8; Yes), a plurality of regions in which pixels set to 0 scattered in the first binarized image are connected to adjacent pixels are displayed. A labeling process for extracting and labeling as the first binarized region is executed (step S9).

FIG. 5 shows a flowchart of the labeling process. The labeling process is a process realized by software processing in cooperation with the CPU 31 and the QC tool 301 and the like.
In the labeling process, first, a first binarized image is scanned, and pixels set to 0 that are not labeled (for example, a number for identifying each first binarized area) are detected (step S31; Yes), a label is attached to the pixel (step S32). Then, it is determined whether or not there is a pixel set to 0 among the pixels adjacent to the labeled pixel (step S33). If there is an adjacent pixel set to 0 (step S33; Yes), the same label is attached to the adjacent pixels (step S34). Further, paying attention to the labeled pixel, the presence or absence of a pixel set to 0 among the adjacent pixels is similarly determined and labeling is executed. By repeating the process until there is no pixel to be assigned the same label (step S35; Yes), a group of pixels having the same label, that is, one first binarized region is formed. Is done.
Thereafter, by repeating the same process, a plurality of first binarized areas with different labels are extracted for each set of pixels set to 0 extracted by the binarization process.

  Therefore, steps S8 and S9 correspond to a first extraction step of extracting a continuous area having a predetermined signal value from the generated first binarized image as the first binarized area.

  After step S9, the feature quantity of the first binarized area corresponding to each label is calculated for each label, and the barycentric coordinate is calculated from the feature quantity (step S10). The feature amount is the size and shape of the first binarized region corresponding to each label, and the shape is a perimeter, circularity, aspect ratio, or the like.

First, an area is calculated as the size of the first binarized region. The area is obtained by counting the total number of pixels with the same label. Next, the perimeter is calculated. The perimeter is obtained by counting the number of pixels located on the outermost side of the first binarized area. Then, a circularity (a value normalized to 0 to 1) indicating a circularity is calculated based on the calculated area and perimeter. The degree of circularity e is S as the area and L as the perimeter.
e = 4πS / L ²
It is calculated by the following formula. The circularity indicates, for example, a value of 1.0 for a circle, 0.79 for a square, and a value less than 0.79 for a rectangle.
Also, the aspect ratio is calculated. Furthermore, the barycentric coordinates of the first binarized area are calculated based on the calculated feature amount.

  Accordingly, step S10 corresponds to a first feature quantity calculation step of calculating the feature quantity of each extracted first binarized region.

  After the feature amount and the barycentric coordinate for each label are calculated (after step S10), the feature amount for each label and the first evaluation member (metal disks 1a to 1c, edge pattern 2, step wedges 3a to 3d, jitter) 4a and 4b) are compared with feature values set in advance (step S11).

For example, the relationship between the feature quantity and the evaluation member is as shown in Table 1. In step S <b> 11, it is preferable not to compare all feature amounts, but to compare specific feature amounts with an evaluation member.

  The presence / absence of a label having a feature amount within a predetermined range of the feature amount set in advance for each evaluation member is determined, and it is determined whether or not there is a label corresponding to all the evaluation members (step S12). In the determination in step S12, it is preferable to set an appropriate predetermined range because the feature amount of the label is affected by the threshold value when binarized.

  When there is no label corresponding to any of the evaluation members, that is, when the labels corresponding to all the evaluation members are not discriminated (step S12; No), an error report indicating that the QC image data is not appropriate is generated. Then, the error report is stored in the QC evaluation result file 394 (step S4), and this process is terminated.

  Therefore, Step S12; No, Step S4 is an error report as abnormality occurrence information when there is no feature value of each first evaluation member determined in advance corresponding to the feature value of any of the first binarized areas. This corresponds to an abnormality information generation step for generating.

  When labels corresponding to all the evaluation members are determined, that is, when an evaluation member corresponding to each label is specified (step S12; Yes), the barycentric coordinates calculated in step S10 are the centers of the evaluation members. Set as coordinates. Thus, the position of each evaluation member is specified. Based on the center coordinates of each evaluation member, analysis center coordinates serving as a reference when evaluating each evaluation member are calculated (step S13).

  Therefore, Steps S11 and S12 are the calculated feature values of the respective first binarized regions and the predetermined first evaluation members (the metal disks 1a to 1c, the edge pattern 2, the step wedge 3a to 3d and jitters 4a and 4b) are compared with each other, and a first evaluation member corresponding to each first binarized region is specified according to the comparison result.

  FIG. 6A shows an example of the positional relationship between the center coordinates of the edge pattern 2, the analysis center coordinates, and the analysis region, and FIG. FIG. 6C shows an example of the positional relationship between the center coordinates of the jitters 4a and 4b, the analysis center coordinates, and the analysis region. The center coordinates are shown as black circles, the analysis center coordinates are black stars, and the analysis area is surrounded by a broken line.

  As shown in FIGS. 6A to 6C, the center coordinates and the analysis center coordinates of each evaluation member are different from each other, and a range determined in advance based on the analysis center coordinates is an analysis region. Yes. As for the positional relationship between the center coordinates and the analysis center coordinates, since the size and arrangement method (tilt) of the evaluation member are predetermined, the analysis center coordinates are calculated based on the center coordinates by determining the center coordinates. can do.

Next, processing for position recognition of the burger phantoms 5a to 5j, which are evaluation members with low contrast resolution, is executed.
The burger phantoms 5a to 5j have a low contrast with the peripheral region due to the nature of the material and image quality evaluation items, and it is difficult to recognize the position simultaneously with the metallic first evaluation member. Therefore, the position of the burger phantoms 5a to 5j is recognized separately from the first evaluation member.

  First, among the first evaluation members whose central coordinates are set in step S13, image data of a predetermined position and a predetermined region (small region) from the central coordinates of any evaluation member is obtained in step S2. Extracted from the extracted QC image data (step S14).

  In step S14, for example, the positional relationship between the metal disc 1a and the burger phantom 5a is less likely to be largely displaced on the QC phantom. Therefore, image data of a preset region centered on the position where the burger phantom 5a should exist from the metal disk 1a and including the planar area of the burger phantom 5a is extracted from the QC image data extracted in step S2. .

  Thus, by extracting the image data of the small area, only the image corresponding to the burger phantom and the image data around the burger phantom are extracted to the small area, that is, corresponding to the first evaluation member. Image data not including the image data to be extracted is extracted. Therefore, when binarizing, it is possible to set a threshold at which a burger phantom can be extracted.

  Therefore, step S14 is a region extraction step of extracting an image of a region having a predetermined positional relationship from the QC image data as the acquired radiographic image with reference to any region where the first evaluation member is specified. Equivalent to.

  A histogram is created from the image data of the small area extracted in step S14 (step S15). A threshold is determined based on the created histogram (step S16), and the image data of the small area is binarized using the determined threshold as a reference to generate a second binarized image (step S17).

  Accordingly, steps S15 to S17 correspond to a second binarization process in which the extracted image data of the small region is binarized to generate a second binarized image.

  Then, a labeling process is performed in which a region in which pixels set to 0 existing in the second binarized image are connected is extracted as a second binarized region to perform labeling (step S18). Accordingly, step S18 corresponds to a second extraction step of extracting the second binarized area from the generated second binarized image.

  After step S18, the feature amount of the second binarized area corresponding to the label is calculated, and the barycentric coordinate is calculated from the feature amount (step S19). Accordingly, step S19 corresponds to a second feature amount calculation step of calculating the feature amount of the second binarized area.

  After the feature amount of the label and the barycentric coordinates are calculated (after step S19), the feature amount of the label and the preset feature amount of the burger phantom are compared with each other (step S20).

  It is determined whether the feature amount of the label is within a predetermined range of the feature amount of the burger phantom as the second evaluation member set in advance (step S21). In the determination in step S21, it is preferable to set an appropriate predetermined range because the feature amount of the label is affected by the threshold value when binarized.

  Accordingly, in steps S20 and S21, the feature value of the second binarized area and the feature value of the second evaluation member set in advance are compared with each other, and each second binary value is determined according to the comparison result. This corresponds to a determination step of determining the presence or absence of the second evaluation member corresponding to the valued area.

  If the feature quantity of the label is not within the predetermined range of the preset feature quantity of the burger phantom (step S21; No), an error report indicating that the QC image data is not appropriate is generated, and the error report is subjected to QC evaluation. The result file 394 is stored (step S4), and this process is terminated.

  When the feature quantity of the label is within a predetermined range of the feature quantity of the burger phantom, that is, when the second binarized area to which the label is attached is specified as the area where the burger phantom exists (step S21; Yes) ), The center-of-gravity coordinates calculated in step S19 are set as the center coordinates of the burger phantom. As a result, the position of the burger phantom is specified. Then, based on the center coordinates of the burger phantom, analysis center coordinates serving as a reference when the burger phantom is evaluated are calculated (step S22).

And based on the analysis center coordinate of each evaluation member calculated in step S13 and step S22, the analysis area used for each evaluation is extracted (step S23). Calculation of each image quality evaluation item (QC calculation) is executed for each extracted analysis region (step S24).
The image quality evaluation items executed in step S24 are display dimension accuracy evaluation, sharpness evaluation, linearity evaluation, reading laser scanning performance evaluation, and low contrast resolution evaluation.

The display dimensional accuracy evaluation is performed using the metal disks 1a to 1c, and the display dimensional accuracy of the reading device 20 in the main scanning direction and the sub-scanning direction, that is, a predetermined distance between the metal disks 1a to 1c. The error (evaluation result value) is calculated.
For example, based on the calculated center coordinates of the metal disks 1a to 1c, first, the distance between the metal disks 1a to 1b on the QC image data is calculated, and the obtained value and the actual QC phantom are calculated. An evaluation result value (%) is calculated by a predetermined arithmetic expression based on the measured value of the distance (stored in advance) between the metal disks 1a and 1b obtained by measurement. Further, the distance between the metal disks 1b to 1c on the QC image data is calculated, and the obtained value and the actually measured value of the distance between the metal disks 1b to 1c obtained by actually measuring the QC phantom P, Based on the above, the evaluation result value is calculated by a predetermined arithmetic expression. Note that the distance between two points on the QC image data is represented by the number of pixels between the two points of the image data × read pixel size.

In the sharpness evaluation, the edge pattern 2 is used to calculate an MTF that is an index of the sharpness in the main scanning direction and the sub-scanning direction of the reading apparatus.
This MFT has a line spread function obtained by differentiating the image of the analysis area of the edge portion in the main scanning direction and the image of the analysis area of the edge portion in the sub-scanning direction extracted based on the analysis center position of the edge pattern 2. Calculated by Fourier transform.

In the linearity evaluation, the linearity of the reading apparatus is calculated and evaluated using the step wedges 3a to 3d.
For example, the average value of the signal value of each pixel of the image data in the area corresponding to each of the step wedges 3a to 3d on the QC image data is calculated, and the transmitted dose converted from the average value and the step wedges 3a to 3d. The linearity is calculated by evaluating the relationship with the theoretical penetration dose.

In the reading laser scanning property evaluation, the amount of pixel shift in the main scanning direction and the sub scanning direction of the reading device is evaluated using the jitters 4a and 4b.
As for the pixel shift amount, for example, a least square error straight line is calculated from the edge portion of the jitter 4a on the QC image data, a pixel shift amount from the theoretical value of the minimum square error straight line of the jitter 4b is calculated, and the QC image A least square error line is calculated from the edge portion of the jitter 4b on the data, and a pixel shift amount from the theoretical value of the least square error line of the jitter 4b is calculated.

In the low contrast resolution evaluation, the low contrast resolution of the reading apparatus is evaluated using the burger phantoms 5a to 5j.
For example, the contrast-to-noise ratio is calculated based on the contrast and noise amount between the average signal value in the burger phantom and the average signal value in the surrounding area, and the low contrast resolution is evaluated based on the calculated contrast-to-noise ratio. Is done.

  Accordingly, in steps S23 and S24, the first evaluation member (metal discs 1a to 1c, edge pattern 2, step wedges 3a to 3d, jitter 4a and 4b) is determined based on the feature value of the first binarized region. For each identified first binarized area, a first calculation step for executing an evaluation calculation corresponding to the first evaluation member and a second evaluation member (burger phantoms 5a to 5j) are specified. This corresponds to a second calculation step of executing an evaluation calculation corresponding to the second evaluation member for each second binarized region.

  After the QC calculation of each image quality evaluation item in step S24, an evaluation result value calculated for each image quality evaluation item is compared with an evaluation reference value predetermined for each image quality evaluation item, and for each image quality evaluation item in advance. It is checked against the set allowable range, and pass / fail is determined for each image quality evaluation item that is acceptable if it is within the allowable range and rejected if it is outside the allowable range (step S25).

  An evaluation report indicating the result determined in step S25 for the QC image data on which this process has been executed is generated, the evaluation report is stored in the QC evaluation result file 394 (step S26), and the process ends.

  The error report or the evaluation report stored in the QC evaluation result file 394 is displayed on the display unit 33 in response to an operation signal from the operation unit 32, or appropriately sent to the service center 50 via the second communication unit 35. Or sent.

  Reference position information for identifying the position of each evaluation member does not need to be held in the QC phantom P, and even if the position of the evaluation member in the QC phantom P is changed, new reference position information is again created. The QC phantom P can be evaluated even if there is a variation in the position of each evaluation member for each QC phantom P. Therefore, the QC phantom P can be evaluated. Regardless of the difference in physical characteristics of each, it is possible to improve the efficiency of quality evaluation.

  Furthermore, since the region of the image for specifying the evaluation member can be limited and the evaluation member can be specified from the image in the region, the second evaluation member as an evaluation member having a lower contrast than the first evaluation member. (Burger Phantom) can be specified, and evaluation suitable for the second evaluation member can be performed.

  In addition, an error report can be generated, and an evaluation operation is not performed on the QC phantom P that cannot be identified due to damage, so that an erroneous evaluation result is not generated.

  Further, the present invention is not limited to the contents of the above embodiment, and can be appropriately changed without departing from the gist of the present invention.

1 is an overall configuration diagram of a radiation image reading system. It is a figure which shows the function structural example of a console. 5 is a diagram illustrating an example of a QC phantom P. FIG. It is a flowchart of a quality evaluation process. It is a flowchart of a labeling process. FIG. 6A shows an example of the positional relationship between the center coordinates of the edge pattern, the analysis center coordinates, and the analysis region. FIG. 6B shows an example of the positional relationship between the center coordinates of the step wedge, the analysis center coordinates, and the analysis region. FIG. 6C is a diagram illustrating an example of the positional relationship between the jitter center coordinates, the analysis center coordinates, and the analysis region.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1a, 1b Metal disc 2 Edge pattern 3a-3d Step wedge 4a, 4b Jitter 5a-5j Burger phantom 10 Imaging device 11 Radiation source 12 Cassette 20 Reading device 30, 30a Console 31 CPU
32 Operation unit 33 Display unit 34 First communication unit 35 Second communication unit 36 Bar code reader 37 RAM
38 Storage unit 39 HDD
301 QC Tool 391 QC Shooting Management Information File 392 General Shooting Management Information File 393 Image File Storage Area 394 QC Evaluation Result File 40 Imager 50 Service Center 51 HDD
A Radiation image reading system M Subject N1, N2 Communication network N3 Bus P QC phantom

Claims (9)

An image acquisition step of acquiring a radiographic image obtained by photographing a QC phantom in which a first evaluation member having a size or shape corresponding to an image quality evaluation item is included;
A first binarization step of binarizing the radiation image acquired by the image acquisition step to generate a first binarized image;
A first extraction step of extracting a continuous region having a predetermined signal value as a first binarized region from the first binarized image generated by the first binarization step;
A first feature amount calculation step of calculating a feature amount of the first binarized region extracted by the first extraction step;
A first calculation step of executing, for the first binarized region, an evaluation calculation associated in advance based on the feature amount of the first binarized region calculated by the first feature amount calculating step. When,
Quality evaluation method for radiation images including Abnormal information generation for generating abnormality occurrence information when the comparison result of the specific step does not include the characteristic value of the predetermined first evaluation member corresponding to the characteristic value of any of the first binarized regions Process,
The quality evaluation method of the radiographic image of Claim 1 containing these. The feature amount is the size and shape of each region,
The shape is circularity, perimeter length or aspect ratio for each region,
The quality evaluation method of the radiographic image of Claim 1 or 2. The QC phantom includes a second evaluation member,
A region extracting step of extracting an image of a region in a predetermined positional relationship from the radiographic image acquired by the image acquiring step with reference to any region in which the corresponding first evaluation member is specified by the specifying step; ,
A second binarization step of binarizing the image extracted by the region extraction step to generate a second binarized image;
A second extraction step of extracting a second binarized region from the second binarized image generated by the second binarization step;
A second feature amount calculation step of calculating a feature amount of the second binarized region extracted by the second extraction step;
The feature quantity of the second binarized region calculated by the second feature quantity calculation step is compared with the predetermined feature quantity of the second evaluation member, and according to the comparison result A determination step of determining the presence or absence of the second evaluation member corresponding to the second binarized region;
A second calculation step of executing an evaluation calculation corresponding to the second evaluation member for the second binarized region in which the second evaluation member corresponding to the determination step is specified;
The quality evaluation method of the radiographic image as described in any one of Claim 1 to 3 containing these. Image acquisition means for acquiring a radiographic image obtained by photographing a QC phantom in which a first evaluation member having a size or shape corresponding to an image quality evaluation item is included;
First binarization means for binarizing the radiation image acquired by the image acquisition means to generate a first binarized image;
First extraction means for extracting a continuous area having a predetermined signal value from the first binarized image generated by the first binarization means as a first binarized area;
First feature amount calculating means for calculating a feature amount of the first binarized area extracted by the first extracting means;
First calculation means for executing, on the first binarized area, an evaluation calculation associated in advance based on the feature value of the first binarized area calculated by the first feature value calculating means. When,
A quality evaluation apparatus comprising: Abnormal information generation for generating abnormality occurrence information when the comparison result by the specifying means has no characteristic value of the predetermined first evaluation member corresponding to the characteristic value of any of the first binarized regions Means,
The quality evaluation apparatus according to claim 5, comprising: The feature amount is the size and shape of each region,
The shape is circularity, perimeter length or aspect ratio for each region,
The quality evaluation apparatus according to claim 5 or 6. The QC phantom includes a second evaluation member,
A region extracting unit that extracts an image of a region in a predetermined positional relationship from the radiation image acquired by the image acquiring unit with reference to any region in which the corresponding first evaluation member is specified by the specifying unit; ,
Second binarization means for binarizing the image extracted by the region extraction means to generate a second binarized image;
Second extraction means for extracting a second binarized area from the second binarized image generated by the second binarization means;
Second feature amount calculation means for calculating a feature amount of the second binarized area extracted by the second extraction means;
The feature quantity of the second binarized area calculated by the second feature quantity calculation means is compared with the predetermined feature quantity of the second evaluation member, and according to the comparison result Determination means for determining the presence or absence of the second evaluation member corresponding to the second binarization region;
Second calculation means for executing an evaluation calculation corresponding to the second evaluation member on the second binarized region in which the second evaluation member corresponding to the determination means is specified;
The quality evaluation apparatus according to any one of claims 5 to 7. Computer
Image acquisition means for acquiring a radiation image obtained by photographing a QC phantom in which a first evaluation member having a size or shape corresponding to an image quality evaluation item is included;
First binarization means for binarizing the radiation image acquired by the image acquisition means to generate a first binarized image;
First extraction means for extracting a continuous area having a predetermined signal value from the first binarized image generated by the first binarization means as a first binarized area;
First feature amount calculating means for calculating a feature amount of the first binarized area extracted by the first extracting means;
First calculation means for executing, on the first binarized area, an evaluation calculation associated in advance based on the feature value of the first binarized area calculated by the first feature value calculating means. ,
Program to function as.

Images