JP4258092B2 - Image processing apparatus and image processing method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing apparatus and an image processing method for processing various images, and more particularly to an image processing apparatus suitable for filtering processing using convolution calculation and an image processing method for performing such image processing.
[0002]
[Prior art]
Radiation images such as X-ray images are often used for disease diagnosis, etc. In order to obtain this X-ray image, X-rays transmitted through the subject are irradiated onto a phosphor layer (phosphor screen), thereby Conventionally, so-called radiographs, in which visible light is generated and developed by irradiating a film using a silver salt with the visible light in the same manner as ordinary photographs, have been used.
[0003]
However, in recent years, a method for taking out an image directly from a phosphor layer without using a film coated with silver salt has been devised. In this method, the radiation that has passed through the subject is absorbed by the phosphor, and then the phosphor is excited by light or thermal energy, for example, so that the radiation energy accumulated by the phosphor is absorbed as fluorescence. There is a method of obtaining image data by radiating and photoelectrically converting the fluorescence.
[0004]
Specifically, for example, US Pat. No. 3,859,527 and JP-A-55-12144 disclose a radiation image conversion method using a stimulable phosphor and using visible light or infrared light as stimulating light. This method uses a radiation image conversion panel in which a photostimulable phosphor layer is formed on a support. The radiation transmitted through the subject is applied to the photostimulable phosphor layer of the conversion panel, and the radiation of each part of the subject is detected. Radiation energy corresponding to the transmittance is accumulated to form a latent image, and then the stimulated layer is scanned with stimulated excitation light to radiate the accumulated radiation energy and convert it into light. This optical signal is photoelectrically converted to obtain radiation image data.
[0005]
The radiographic image data obtained in this way is displayed as it is or after being subjected to image processing and output to a silver salt film, a CRT or the like for visualization or filing in an electronic filing device. In the image processing, gradation processing is performed for the purpose of making the density of a region of interest (a region including an image portion necessary for diagnosis) in a reproduced image constant, and for outputting a shadow of a human body structure or lesion more easily. And image processing such as spatial frequency processing.
[0006]
In addition to radiation images, in various fields, various types of image processing are performed on image data and then images are reproduced or displayed in various fields. Then, by performing enhancement processing as the image data, the outline and the like are clarified.
[0007]
As such image processing, a non-sharp image in which only low frequency components of image data are extracted is created, and the image data in which high-frequency components are emphasized is obtained by subtracting this non-sharp image from the image data of the original image. It has been known.
[0008]
In addition, the image data of the original image is converted into a frequency component such as Fourier transform, and after emphasizing a predetermined frequency in the frequency component, inverse conversion is performed to obtain image data with desired enhancement. Are known.
[0009]
Further, by converting the image into a multi-resolution space, the image is decomposed into a plurality of frequency bands, and an enhancement process is performed on at least one of the frequency band images decomposed into a plurality of frequencies, Image processing that obtains processed image data by inversely transforming an image in a frequency band that has been subjected to enhancement processing and an image in another frequency band is also known.
[0010]
As an image processing technique effective for enhancing a desired frequency band, an image processing technique using a Laplacian pyramid method has been proposed as described in JP-A-6-301766. It was.
[0011]
[Problems to be solved by the invention]
However, in the case of the above Laplacian pyramid processing, in order to obtain an output signal, the entire image or the entire image that has already been filtered is filtered, and an operation based on the processed image data is performed. To obtain output image data.
[0012]
For this reason, even when an output pixel value for one pixel of interest is obtained, it is necessary to process all the signals of the entire image, and much processing time and memory are required.
[0013]
In addition, since one calculation performs filter processing with the same function on the entire image data, it is difficult to perform processing depending on the area in the image (processing only for a specific area).
[0014]
The present invention has been made in view of the above problems, and an image processing apparatus capable of easily obtaining a converted signal value of a target pixel without performing conversion on all image data, and It is to realize an image processing method.
[0015]
Further, the present invention has been made in view of the above problems, and when converting all the image data, the converted image data is not required to have a memory equal to the size of the image data. An image processing apparatus and an image processing method that can be obtained are realized.
[0016]
[Means for Solving the Problems]
In the present invention, when image processing is performed on image data of an image composed of a plurality of pixels, the pixel value of the pixel of interest in the original image and the pixel values of neighboring pixels and a predetermined function between the pixel value The value of the pixel of interest after conversion is determined by the calculated value based on the convolution operation. Thereby, it is possible to realize an image processing apparatus and an image processing method capable of easily obtaining the converted signal value of the target pixel without performing conversion on all the image data. In addition, an image processing apparatus and an image processing method that can obtain converted image data without requiring a memory equal to the size of the image data when converting all the image data can be realized.
[0017]
That is, in the case where the converted value of the target pixel is obtained based on the weighted average value obtained by changing the mask size, the target pixel is averaged (weighted) using a plurality of size masks. Processing, obtaining the difference value of the average values of masks of close sizes in adjacent ranges, performing conversion processing on this difference value, and adding the converted difference value to the pixel value of the target pixel, thereby processing the processed signal value Get. In this way, by performing all processing in units of pixels, it is possible to save the memory used and improve the processing speed.
[0018]
Specifically, it is as described below.
The invention according to claim 1 is an image processing apparatus comprising image processing means for performing image processing on image data of a medical image composed of a plurality of pixels photographed by radiation irradiation, wherein the image processing means , The pixel value of the pixel of interest and the pixel values of its neighboring pixels about A given function by The value of the pixel of interest after conversion is determined by a calculated value based on a plurality of convolution operations, and the predetermined function used for the plurality of convolution operations is The average value calculation ranges are different from each other, and a plurality of functions for obtaining weighted average values by Gaussian distribution or uniform distribution in the average value calculation range. An image processing apparatus characterized by that.
[0024]
Claim 2 In the described invention, the image processing means is , Pixel image of interest Elementary The average value calculation range is adjacent to the value function By adding the converted value obtained by converting the difference value of the calculated value of the convolution operation based on the predetermined conversion The value of the target pixel after conversion Claims obtained 1 The image processing apparatus described in the above.
[0025]
Claim 3 In the described invention, the image processing means is , Widest average value calculation range function The average value calculation range is adjacent to the average value obtained by function By adding the converted value obtained by converting the difference value of the calculated value of the convolution operation based on the predetermined conversion The value of the target pixel after conversion Claims obtained 1 The image processing apparatus described in the above.
[0026]
Claim 4 The present invention is characterized in that the predetermined conversion for the difference value is given by a function that is nonlinear with respect to the input value. 2 Or claims 3 The image processing apparatus according to any one of the above.
[0027]
Claim 5 The invention described in claim 1 is characterized in that the non-linear function has at least a part where the absolute value of the output value is smaller than the absolute value of the input value. 4 It is an image processing apparatus of description.
[0028]
Claim 6 The invention described in claim 1, wherein the non-linear function varies depending on an average value calculation range in which a difference is obtained. 4 Or claims 5 The image processing apparatus according to any one of the above.
[0029]
Claim 7 The invention described in claim 1 is characterized in that the non-linear function takes a smaller output value near the input value 0 as the average value calculation range obtained by calculating the difference is wider. 4 To claims 6 The image processing apparatus according to any one of the above.
[0030]
Claim 8 In the described invention, the image processing means performs a calculation based on a plurality of convolution operations of the target pixel and its neighboring pixels and a predetermined function, and has a sufficient memory area to store the value, The converted values for a plurality of pixels of interest are determined by repeatedly using the region. 7 The image processing apparatus according to any one of the above.
[0032]
Claim 9 The invention described in claim 1, wherein the predetermined function used for the plurality of convolution operations is different depending on image data. 8 The image processing apparatus according to any one of the above.
[0033]
Claim 1 0 The described invention is an image processing method for performing image processing on image data of a medical image composed of a plurality of pixels photographed by radiation irradiation, wherein the pixel value of a target pixel and the pixel values of its neighboring pixels A plurality of average value calculation ranges are different from each other, and a weighted average value by a Gaussian distribution or uniform distribution in the average value calculation range Functions by The value of the target pixel after conversion is determined by the calculated value based on multiple convolution operations. , An image processing method characterized by this.
[0035]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. First, image processing for each pixel using a plurality of masks, which is a characteristic part of the embodiment of the present invention, will be described, and then the entire radiation image processing apparatus to which the image processing of the present embodiment is applied. The configuration will be described, and then the overall operation of the radiation image processing apparatus will be described.
[0036]
<Details of image processing>
FIG. 1 is an explanatory view schematically showing image processing using a mask according to an embodiment of the present invention. FIG. 2 is a schematic diagram showing the state of the mask and the image data. Note that the image processing described below is performed by an image processing unit of an apparatus described later.
[0037]
That is, the present invention relates to various image processing (contour emphasis by frequency processing, etc.) performed on image data before reproducing or displaying the obtained image data in the field of various images such as radiographic images.
[0038]
In FIG. 1 and FIG. 2, mask A (A pixel × A pixel), mask B (B pixel × B pixel), mask C (C pixel × C pixel) are used as masks of different sizes with respect to the target pixel a. A mask D (D pixel × D pixel) is arranged.
[0039]
These masks A to D are functions having a plurality of shapes used for a plurality of convolution operations for the same target pixel, and are configured by a function family having different average value calculation ranges in a distribution for obtaining a weighted average value. It is a thing. The distribution is a Gaussian distribution or a uniform distribution.
[0040]
First, in the image data in FIG. 2, for example, the upper left end pixel is selected as the target pixel (S1 in FIG. 3). Then, using a plurality of masks having different sizes (here, mask A to mask D), a calculated value (weighted average value) by each mask is obtained (S2 in FIG. 3). Specifically, calculation of a non-sharp image using a Gaussian mask is performed for each pixel using each mask.
[0041]
Then, the difference between the calculated values of the masks having the closest sizes is obtained (S3 in FIG. 3). In the case of FIG. 1, the difference d_AB between the calculated value of the mask A and the calculated value of the mask B, the difference d_BC between the calculated value of the mask B and the calculated value of the mask C, the calculated value of the mask C and the calculated value of the mask D The difference d_CD is obtained.
[0042]
Next, predetermined conversion is performed on each difference (S4 in FIG. 3). As the predetermined transformation, for example, a nonlinear transformation given by a nonlinear function with respect to an input value as shown in FIG. 4 is applicable.
[0043]
The difference d_AB is subjected to nonlinear transformation to generate a nonlinear difference d′ _AB, the difference d_BC is nonlinearly transformed to generate a nonlinear difference d′ _BC, and the difference d_CD is subjected to nonlinear transformation to obtain a nonlinear difference d′ _CD. Generate. This nonlinear function has at least a part where the absolute value of the output value is smaller than the absolute value of the input value. The nonlinear function varies depending on the average value calculation range for which the difference is obtained. The non-linear function takes a smaller output value near the input value 0 as the average value calculation range for which the difference is obtained is wider (the mask is larger) (see FIG. 4).
[0044]
In the above case, when there is a sudden change in the pixel value in the original image, the difference value (d_AB,...) Of the above calculation also takes a very large value, and the magnitude of the value is generally The difference value obtained from a large mask size has a characteristic that it is large. If these values are added to the target pixel without conversion, a very large change in the pixel value occurs compared to the original pixel, causing artifacts. Therefore, by using a conversion function as shown in FIG. 4, the pixel value is more strongly suppressed as the value obtained from the difference in mask size becomes larger, and the artifact can be eliminated.
[0045]
Then, each of the above nonlinear transformation differences is added. This addition result corresponds to the high-frequency component Sh for the pixel a. This high frequency component Sh is emphasized by a predetermined coefficient β to obtain βSh (FIG. 3 S6). As the predetermined coefficient β, as shown in FIG. 5, the emphasis above a predetermined signal value is made stronger than the predetermined signal value, thereby suppressing the emphasis degree of a low-concentration region where noise is conspicuous and high. It is desirable because sufficient emphasis in the density part can be achieved.
[0046]
Further, it is desirable that the coefficient β has a shape as shown in FIG. That is, it changes almost linearly between the reference values T1, T2. As shown by the solid lines in FIG. 20, when values “A” and “B” are set and low luminance is emphasized, β from the reference value T1 to the value “A” is maximized, and the value “ B "to the reference value T2 is minimized. In addition, from the value “A” to the value “B”, β changes almost linearly. When emphasizing high luminance, as shown by a broken line, β from the reference value T1 to the value “A” is minimized and maximized from the value “B” to the reference value T2. In addition, from the value “A” to the value “B”, β changes almost linearly. Although not shown, when medium luminance is emphasized, β of values “A” to “B” is maximized. By doing in this way, the emphasis degree of arbitrary luminance parts can be controlled.
[0047]
The enhanced high-frequency component βSh obtained as described above is added to the original pixel a to obtain a processed pixel a + βSh (S7 in FIG. 3). In this way, with respect to the pixel a, a processed pixel a + βSh is obtained by calculation using a plurality of masks, addition, and subtraction. That is, in this embodiment, it is possible to obtain the converted signal value of the pixel of interest without converting all the image data.
[0048]
The calculation related to the pixels as described above is sequentially performed for each pixel with respect to all pixels of the image data or all pixels in a desired region (NO in step S8 in FIG. 3 → S1 to S7), thereby converting all the image data. Also when performing the above, it is possible to obtain the converted image data without requiring a memory equal to the size of the image data. Therefore, effects such as saving of working memory capacity and shortening of calculation time can be obtained.
[0049]
In the above description, the image processing using the four masks having different sizes of the mask A to the mask D has been described. However, more masks can be used. Note that the number of masks having different sizes can be further increased. In order to sufficiently emphasize other regions without causing artifacts even for very large signal changes, it is desirable to suppress the difference obtained from a relatively large mask size, and at least four masks. It is desirable to perform processing according to size.
[0050]
In addition, since the image processing of this embodiment can be performed for each pixel, it is also suitable for processing only a specific area in the image data, and the entire conventional image data is processed. Compared with the one to be performed, the effect of shortening the calculation time is remarkably obtained.
[0051]
In addition, since the image processing of this embodiment can be performed for each pixel, it is also suitable for usage in which the processing content is changed for each of several areas in the image data.
[0052]
For example, by obtaining an appropriate threshold value by using a known discriminant analysis method for the image of the foot side as shown in FIG. 6 (a), the bone portion (FIG. 6 (c)) and other regions (soft tissue) : Disassemble into FIG. 6 (b)).
[0053]
For each image thus decomposed, the above-described image processing is performed using only a plurality of masks having a relatively small mask size only on the bone portion (FIG. 6C), and other portions such as soft tissue. For (FIG. 6B), the above-described image processing is performed using a plurality of masks having a relatively large mask size. In this way, processing depending on the characteristics of the region is possible, and frequency processing using a plurality of masks and frequency characteristics can be different between the soft region and the bone region. Here, there are two areas, but it is also possible to perform image processing by changing the mask size into more areas.
[0054]
By changing the mask size according to the region in this way, the bone portion processed with a relatively small mask size is emphasized with a relatively high frequency component in the image, and only minute components such as trabeculae are sharp. Thus, in the soft part region processed with a relatively large mask size, relatively low frequency components in the image such as muscle and adipose tissue are emphasized, and large components such as muscle and adipose tissue become clear.
[0055]
It is also possible to execute different image processing by changing the mask size in accordance with the type of image.
<Overall configuration of radiation image processing apparatus>
FIG. 8 is a system configuration diagram showing an overall configuration of a radiation image processing apparatus to which the image processing of the present embodiment is applied. The radiation generator 30 is controlled by the controller 10, and the radiation emitted from the radiation generator 30 is applied to the imaging panel mounted on the front surface of the radiation image reader 40 through the subject 5.
[0056]
FIG. 9 shows a configuration of a radiation image reader 40 using an FPD (Flat Panel Display). The imaging panel 41 has a substrate having a thickness sufficient to obtain a predetermined rigidity, and a detection element 412-(1,1) that outputs an electrical signal on the substrate according to the dose of irradiated radiation. ) To 412- (m, n) are two-dimensionally arranged in a matrix. Further, the scanning lines 415-1 to 415-m and the signal lines 416-1 to 416-n are disposed so as to be orthogonal, for example.
[0057]
The scanning lines 415-1 to 415-m of the imaging panel 41 are connected to the scanning driving unit 44. When the readout signal RS is supplied from the scan driver 44 to one of the scanning lines 415-1 to 415-m (p is any value of 1 to m), the scanning line 415 is supplied. Electric signals SV-1 to SV-n corresponding to the radiation dose irradiated from the detection element connected to -p are output to the image data generation circuit 46 via the signal lines 416-1 to 416-n. Supplied.
[0058]
This detection element 412 may be any element that outputs an electrical signal corresponding to the dose of irradiated radiation. For example, when a detection element is formed using a photoconductive layer in which electron-hole pairs are generated and change in resistance when irradiated with radiation, the detection element is in accordance with the amount of radiation generated in the photoconductive layer. A quantity of charge is stored in the charge storage capacitor, and the charge stored in the charge storage capacitor is supplied to the image data generation circuit 46 as an electrical signal. It is desirable that the photoconductive layer has a high dark resistance value, such as amorphous selenium, lead oxide, cadmium sulfide, mercuric iodide, or an organic material exhibiting photoconductivity (photoconductivity to which an X-ray absorption compound is added. In particular, amorphous selenium is desirable.
[0059]
When the detection element 412 is formed using, for example, a scintillator that generates fluorescence when irradiated with radiation, an image signal is generated by generating an electrical signal based on the fluorescence intensity generated by the scintillator with a photodiode. The circuit 46 may be supplied.
[0060]
As the imaging panel 41 using such a configuration, as disclosed in JP-A-9-90048, X-rays are absorbed by a phosphor layer such as an intensifying screen to generate fluorescence, and the intensity of the fluorescence. Is detected by a photodetector such as a photodiode provided for each pixel. As another fluorescence detecting means, there is a method using a CCD or a C-MOS sensor.
[0061]
In particular, in the imaging panel (FPD) of the method disclosed in the above Japanese Patent Laid-Open No. 6-342098, since the X-ray dose is directly converted into the charge amount for each pixel, the sharpness degradation in the FPD is small and the sharpness is excellent. Since an image is obtained, the effects of the X-ray image recording system and the X-ray image recording method are great and preferable.
[0062]
Furthermore, the imaging panel 41 can be configured as shown in FIGS. FIG. 10 is a front view of a radiographic image detector that can be used as the imaging panel 41, and shows an example in which the radiographic image detector is composed of a plurality of units. The dotted line in FIG. 10 is a line of the grid 50 of the radiation image detector, but is actually hidden from the protective member and the X-ray scintillator and cannot be seen from the front. FIG. 10 shows an example of 6 × 6 = 36 units, but the number is not limited to this.
[0063]
FIG. 11 is a schematic diagram of a longitudinal section of the radiation image detector. The radiation image detector is configured by arranging an X-ray scintillator 51, a lens array 52, and area sensors 54 corresponding to the lenses 53 of the lens array 52 in this order. The X-ray scintillator 51 is protected by a protection member 55. Each lens 53 of the lens array 52 is supported by a lens support member 58, and a transparent member 56 is disposed between the X-ray scintillator 51 and the lens array 52. The area sensor 54 is supported by an area sensor support member 57.
[0064]
The shape, thickness, ray path, etc. of the components of the radiation image detector are not accurate. The grid 50 does not directly touch the X-ray scintillator 51 but abuts against the transparent member 56, thereby preventing the grid 50 from hitting the X-ray scintillator 51 and scratching it, and the boundary line of the grid 50 to Prevents missing parts.
[0065]
Since the area sensors 54 corresponding to the respective lenses 53 of the lens array 52 are arranged in this order, the spatial resolution is high, the image quality is high, the thickness is small, the size is small, and the weight is light.
[0066]
The X-ray scintillator 51 emits visible light upon exposure to X-rays such as gadolinium oxysulfide and cesium iodide, and the X-ray scintillator 51 emits visible light upon exposure to X-rays, resulting in high spatial resolution and high image quality. is there.
[0067]
The lens array 52 is formed from a lens group composed of a combination of two or more different lenses 53, has high spatial resolution, high image quality, and can be made thin. If the imaging magnification of the lens 53 is from 1 / 1.5 to 1/20, and the imaging magnification is greater than 1 / 1.5, the area sensor becomes too large to be disposed, and if it is less than 1/20, X The distance from the line scintillator 51 to the lens increases, and the thickness of the radiation image detector increases.
[0068]
A clear image can be obtained by using a solid-state imaging device such as a CCD or C-MOS sensor as the area sensor 54.
The image data generation circuit 46 sequentially selects the electrical signal SV supplied based on an output control signal SC from a read control circuit 48 described later, and converts it into digital image data DT. The image data DT is supplied to the reading control circuit 48.
[0069]
The reading control circuit 48 is connected to the controller 10 and generates a scanning control signal RC and an output control signal SC based on the control signal CTD supplied from the controller 10. The scanning control signal RC is supplied to the scanning drive unit 44, and the readout signal RS is supplied to the scanning lines 415-1 to 415-m based on the scanning control signal RC.
[0070]
The output control signal SC is supplied to the image data generation circuit 46. For example, when the imaging panel 41 is configured by (m × n) detection elements 412 as described above by the scanning control signal RC and the output control signal SC from the reading control circuit 48, the detection element 412− Assuming that data based on the electric signal SV from (1,1) to 412- (m, n) is data DP (1,1) to DP (m, n), data DP (1,1), DP (1 , 2),... DP (1, n), DP (2,1),..., DP (m, n) in this order, and the image data DT is generated by the image data generation circuit 46. To the reading control circuit 48. Further, the reading control circuit 48 also performs processing for sending the image data DT to the controller 10.
[0071]
Image data DT obtained by the radiation image reader 40 is supplied to the controller 10 via the reading control circuit 48. If the image data obtained by logarithmic conversion processing is supplied when the image data obtained by the radiation image reader 40 is supplied to the controller 10, the processing of the image data in the controller 10 can be simplified.
[0072]
Further, the radiation image reader is not limited to the one using FPD, and may be one using a stimulable phosphor. FIG. 12 shows a configuration when a radiation image reader 60 using a stimulable phosphor is used. In the conversion panel 61 irradiated with radiation, a stimulable phosphor layer is stimulated on the support. It is provided by vapor deposition of photosensitive phosphor or application of stimulable phosphor coating. This photostimulable phosphor layer is shielded or covered with a protective member in order to block adverse effects and damages caused by the environment.
[0073]
A light beam generating unit (gas laser, solid state laser, semiconductor laser, etc.) 62 generates a light beam whose emission intensity is controlled. This light beam reaches the scanning unit 63 via various optical systems, is deflected by the scanning unit 63, further deflects the optical path by the reflecting mirror 64, and is guided to the conversion panel 61 as the excitation excitation scanning light. .
[0074]
A condensing end made of an optical fiber or a sheet-like light guide member of the condensing body 65 is disposed in the vicinity of the conversion panel 61 where the excitation light is scanned, and scanning of the light beam from the light beam generating unit 62 is performed. Thus, the stimulated light emission having the light emission intensity proportional to the latent image energy generated in the conversion panel 61 is received.
[0075]
The filter 66 passes only light in the stimulated emission wavelength region from the light introduced from the light collector 65, and the light that has passed through the filter 66 is incident on the photomultiplier 67.
[0076]
The photomultiplier 67 generates a current signal corresponding to incident light by photoelectric conversion. This current signal is supplied to the current / voltage conversion unit 70 and converted into a voltage signal. Further, the voltage signal is amplified by the amplifying unit 71 and then converted to digital image data DT by the A / D converting unit 72. Here, a logarithmic conversion amplification unit (log amplifier) is used as the amplification unit 71. The image data DT is sequentially subjected to image processing in the image processing apparatus 80, and the image data DTC after image processing is transmitted to the printer 83 via the interface 82.
[0077]
A CPU (Central Processing Unit) 81 is for controlling image processing in the image processing device 80. In the image processing device 80, various image processing (for example, spatial frequency processing, dynamic range processing) is performed on the image data DT. Compression, gradation processing, enlargement / reduction processing, movement, rotation, statistical processing, etc.) to generate image data DTC in a form suitable for diagnosis.
[0078]
This image data DTC is supplied to the printer 83, and a hard copy of the radiation image of each part of the human body can be obtained from the printer 83. A monitor such as a CRT may be connected to the interface 82, and a storage device (filing system) that can store image data of a plurality of radiation images may be connected.
[0079]
The read controller 75 also adjusts the light beam intensity of the light beam generator 62, adjusts the gain of the photomultiplier 67 by adjusting the power supply voltage of the high voltage power supply 76 for the photomultiplier, and adjusts the current / voltage converter 70 and the amplifier 71. Gain adjustment and adjustment of the input dynamic range of the A / D converter 72 are performed, and the reading gain is adjusted comprehensively.
[0080]
The image data DT obtained from the A / D conversion unit 72 is supplied to the controller 10 and controls the operation of the reading control unit 75 by a control signal CTD from the controller 10.
[0081]
The radiation image reader may irradiate light from a light source such as a laser or a fluorescent lamp onto a silver salt film on which a radiation image is recorded, and photoelectrically convert the transmitted light of the silver salt film to generate image data. Good. Moreover, the structure which converts radiation energy directly into an electrical signal using a radiation quantum counting type detector, and produces | generates image data may be sufficient.
[0082]
Next, the configuration of the controller 10 is shown in FIG. The CPU 11 for controlling the operation of the controller 10 is connected to the system bus 12 and the image bus 13 and to the input interface 17. The operation of the CPU 11 for controlling the operation of the controller 10 is controlled based on a control program stored in the memory 14.
[0083]
A display control unit 15, a frame memory control unit 16, an output interface 18, a shooting control unit 19, a disk control unit 20, and the like are connected to the system bus 12 and the image bus 13. Is controlled, and image data is transferred between the units via the image bus 13.
[0084]
A frame memory 21 is connected to the frame memory control unit 16, and image data obtained by the radiation image reader 40 is stored via the imaging control unit 19 and the frame memory control unit 16. The image data stored in the frame memory 21 is read and supplied to the display control unit 15 and the disk control unit 20. The frame memory 21 may store the image data supplied from the radiation image reader 40 after being processed by the CPU 11.
[0085]
An image display device 22 is connected to the display control unit 15, and a radiographic image based on the image data supplied to the display control unit 15 is displayed on the screen of the image display device 22. Here, when the number of display pixels of the image display device 22 is smaller than the number of pixels of the radiation image reader 40, the entire captured image can be displayed on the screen by reading out the image data. Further, if image data in a region corresponding to the number of display pixels of the image display device 22 is read, a captured image at a desired position can be displayed in detail.
[0086]
When image data is supplied from the frame memory 21 to the disk control unit 20, for example, the image data is continuously read out and written into the FIFO memory in the disk control unit 20, and then sequentially recorded in the disk device 23. The
[0087]
Further, the image data read from the frame memory 21 and the image data read from the disk device 23 can be supplied to the external device 100 via the output interface 18.
[0088]
In the image processing unit 26, irradiation field recognition processing, region of interest setting, normalization processing, gradation processing, and the like of the image data DT supplied from the radiation image reader 40 via the imaging control unit 19 are performed. Further, frequency enhancement processing, dynamic range compression processing, or the like may be performed. In addition, image processing etc. can also be performed as the structure which CPU11 serves as the image processing part 26. FIG.
[0089]
Therefore, the image processing unit 26 constitutes image processing means for executing image processing using the convolution operation in the claims.
An input device 27 such as a keyboard is connected to the input interface 17. By operating the input device 27, management information such as information for identifying image data obtained by photographing and information relating to photographing is input.
[0090]
As the external device 100 connected to the output interface 18, a scanning laser exposure apparatus called a laser imager is used. In this scanning laser exposure apparatus, the intensity of a laser beam is modulated by image data, and after exposure to a conventional silver halide photographic light-sensitive material or thermal phenomenon silver halide photographic light-sensitive material, an appropriate development process is performed, thereby performing radiographic image processing. A hard copy can be obtained.
[0091]
Although the image data supplied from the radiation image reader 40 is stored in the frame memory 21, the supplied image data may be stored after being processed by the CPU 11. The disk device 23 stores image data stored in the frame memory 21, that is, image data supplied from the radiation image reader 40 and image data obtained by processing the image data by the CPU 11 together with management information. be able to.
[0092]
<Operation of radiation image processing apparatus>
Next, the operation of the above radiation image processing apparatus will be described. When obtaining a radiographic image of the subject 5, the subject 5 is assumed to be positioned between the radiation generator 30 and the imaging panel 41 of the radiographic image reader 40, and the radiation emitted from the radiation generator 30 is the subject 5. And the radiation that has passed through the subject 5 enters the imaging panel 41. The same applies to the case where the radiation image reader 60 is used instead of the radiation image reader 40. In the following description, the radiation image reader 40 is used, and the description when the radiation image reader 60 is used is omitted. To do.
[0093]
Management information indicating identification of the subject 5 to be photographed and information relating to photographing is input to the controller 10 using the input device 27. Input of management information using the input device 27 is performed by operating a keyboard, using a magnetic card, a bar code, HIS (in-hospital information system: network-based information management), or the like.
[0094]
This management information includes, for example, an ID number, name, date of birth, sex, imaging date and time, imaging site and imaging position (for example, which part of the human body was irradiated with radiation from which direction), imaging method (simple imaging, contrast imaging) Imaging, tomographic imaging, magnified imaging, etc.), imaging conditions (tube voltage, tube current, irradiation time, presence / absence of use of scattered radiation removal grid, etc.).
[0095]
The shooting date and time can be automatically obtained from the CPU 11 by using a clock function built in the CPU 11. It should be noted that the management information to be input may be only related to the subject to be photographed at that time. A series of management information is input in advance, and the subjects are photographed in the order of input, or the management that has been input as necessary. Information may be read and used.
[0096]
When the power switch of the radiation image reader 40 is turned on, the imaging panel 41 is initialized by the reading control circuit 48 and the scanning drive unit 44 of the radiation image reader 40 based on the control signal CTD from the controller 10. Is called. This initialization is for obtaining a correct electrical signal corresponding to the radiation dose emitted from the imaging panel 41.
[0097]
When the initialization of the imaging panel 41 in the radiation image reader 40 is completed, the radiation from the radiation generator 30 can be irradiated. Here, when a switch for irradiating radiation is provided in the radiation generator 30, when this switch is operated, radiation is emitted from the radiation generator 30 toward the subject 5 for a predetermined time, and A signal DFS indicating the start of radiation irradiation and a signal DFE indicating the end of irradiation are supplied to the controller 10.
[0098]
At this time, the radiation amount of the radiation applied to the imaging panel 41 of the radiation image reader 40 is modulated by the subject 5 because the degree of radiation absorption by the subject 5 is different. In the detection elements 412-(1,1) to 412-(m, n) of the imaging panel 41, an electrical signal based on the radiation modulated by the subject 5 is generated.
[0099]
Next, in the controller 10, a predetermined time after the signal DFS is supplied, for example, when the irradiation time of the radiation is about 0.1 seconds, a time longer than this irradiation time (for example, about 1 second), or Immediately after the signal DFE is supplied, the control signal CTD is supplied to the reading control circuit 48 of the radiation image reader 40 in order to start generation of the image data DT by the radiation image reader 40.
[0100]
On the other hand, when a switch for irradiating radiation is provided in the controller 10, when this switch is operated, an irradiation start signal CST for starting irradiation of radiation is transmitted via the imaging control unit 19 to the radiation generator. 30, radiation is emitted from the radiation generator 30 toward the subject 5 for a predetermined time. This irradiation time is set based on management information, for example.
[0101]
Next, in the controller 10, a predetermined time after outputting the irradiation start signal CST, a control signal CTD for starting generation of image data by the radiation image reader 40 is sent to the reading control circuit 48 of the radiation image reader 40. Supply. The controller 10 supplies the radiographic image reader 40 with a control signal CTD for starting generation of image data by the radiographic image reader 40 after detecting the end of radiation irradiation by the radiation generator 30. It is good. In this case, generation of image data during radiation irradiation can be prevented.
[0102]
In the reading control circuit 48 of the radiation image reader 40, the scanning control signal RC and the output control signal SC are generated based on the control signal CTD for starting the generation of the image data supplied from the controller 10. The scanning control signal RC is supplied to the scanning drive unit 44 and the output control signal SC is supplied to the image data generation circuit 46, and the image data DT obtained from the image data generation circuit 46 is supplied to the reading control circuit 48. The The image data DT is sent to the controller 10 by the reading control circuit 48.
[0103]
The image data DT supplied to the controller 10 is stored in the frame memory 21 via the imaging control unit 19, the frame memory control unit 16, and the like. A radiographic image can be displayed on the image display device 22 by using the image data stored in the frame memory 21. Further, the image data stored in the frame memory 21 is processed by the image processing unit 26 and supplied to the display control unit 15, or the image data subjected to the image processing is stored in the frame memory 21, and the frame memory 21 By supplying the stored image data to the display control unit 15, the brightness, contrast, sharpness, and the like are adjusted, and a radiation image suitable for diagnosis or the like can be displayed. In addition, by supplying image data that has undergone image processing to the external device 100, a hard copy of a radiation image suitable for diagnosis or the like can be obtained.
[0104]
In the image processing unit 26, even if the level distribution of the image data output from the imaging panel 41 varies due to the difference in radiation dose, the normalization of the image data DT is always performed so that a stable radiation image can be obtained. Processing is performed. Even if the distribution of the level of the image data fluctuates, gradation processing is performed on the normalized image data DTreg, which is image data after normalization processing, in order to obtain a density and contrast radiation image suitable for diagnosis and the like. Done. Further, the image processing unit 26 reduces the contrast of the fine structure portion of the subject in the frequency enhancement processing for controlling the sharpness of the normalized radiation image with respect to the normalized image data DTreg and the entire radiation image having a wide dynamic range. It is also possible to perform a dynamic range compression process so that the density is within the easy-to-see density range.
[0105]
<Contents of image processing>
In the radiographic image processing apparatus according to the present embodiment, irradiation field recognition processing and ROI recognition are performed prior to the above normalization processing, gradation processing, and dynamic range compression processing. That is, the image processing procedure of the present embodiment is as shown in FIG.
[0106]
As shown in FIG. 7, the processing procedure of this embodiment is as follows.
・ Irradiation field recognition processing,
・ ROI recognition processing,
-Frequency processing or image processing for each pixel using multiple masks of different sizes,
・ Dynamic range compression processing,
・ Tone processing,
It is like this. Hereinafter, the processing procedure of the present embodiment will be described in this order.
[0107]
(1) Irradiation field recognition processing:
By the way, when taking a radiographic image, for example, in order to prevent radiation from being applied to a portion not required for diagnosis, or to a portion not required for diagnosis, radiation scattered in this portion is used for diagnosis. In order to prevent the resolution from being reduced by being incident on a required portion, a radiation non-transparent material such as a lead plate is installed in a part of the subject 5 or the radiation generator 30 so that the radiation field to the subject 5 is irradiated. Irradiation field restriction is performed to limit the above.
[0108]
When this irradiation field stop is performed, if level conversion processing and subsequent gradation processing are performed using the image data of the irradiation field area and the irradiation field area, the image data in the irradiation field is determined by the image data of the irradiation field area. Image processing of a part required for diagnosis is not performed properly. For this reason, the image processing unit 26 performs irradiation field recognition processing for determining the irradiation field inner region and the irradiation field outer region.
[0109]
In the irradiation field recognition processing, for example, a method disclosed in Japanese Patent Application Laid-Open No. 63-259538 is used, and a line from a predetermined position P on the imaging surface toward the end side of the imaging surface as shown in FIG. For example, differentiation processing is performed using the above image data. As shown in FIG. 14B, the differential signal Sd obtained by this differentiation process has a signal level that is high at the irradiation field edge portion. Therefore, the signal level of the differential signal Sd is determined to determine one irradiation field edge candidate point. EP1 is required. A plurality of irradiation field edge candidate points EP1 to EPk are obtained by performing the process of obtaining the irradiation field edge candidate points radially about a predetermined position on the imaging surface. An irradiation field edge portion is obtained by connecting adjacent edge candidate points of the plurality of irradiation field edge candidate points EP1 to EPk obtained in this way with straight lines or curves.
[0110]
Moreover, the method shown by Unexamined-Japanese-Patent No. 5-7579 can also be used. In this method, when the imaging surface is divided into a plurality of small areas, the radiation dose of the radiation is reduced substantially uniformly in the small areas outside the irradiation field where radiation irradiation is blocked by the irradiation field stop. Becomes smaller. In addition, in a small region within the irradiation field, the radiation value is modulated by the subject, so that the dispersion value is higher than that outside the irradiation field. Further, in a small region including the irradiation field edge portion, the portion with the smallest radiation dose and the portion with the radiation dose modulated by the subject coexist, so the dispersion value is the highest. From this, the small area including the irradiation field edge portion is determined by the dispersion value.
[0111]
Moreover, the method shown by Unexamined-Japanese-Patent No. 7-181609 can also be used. In this method, image data is rotated about a predetermined center of rotation, and rotation is performed until the boundary line of the irradiation field becomes parallel to the coordinate axis of the orthogonal coordinates set on the image by the parallel state detection means. When the state is detected, the linear equation of the boundary before rotation is calculated by the linear equation calculation means based on the rotation angle and the distance from the rotation center to the boundary line. Thereafter, by determining a region surrounded by a plurality of boundary lines from a linear equation, the region of the irradiation field can be determined. When the irradiation field edge portion is a curve, for example, one boundary point is extracted based on the image data by the boundary point extracting means, and the next boundary point is extracted from the boundary candidate point group around this boundary point. Similarly, by sequentially extracting boundary points from the boundary candidate point group around the boundary points, it is possible to determine whether the irradiation field edge portion is a curve.
[0112]
(2) ROI recognition processing:
When the irradiation field recognition is performed, the image processing unit 26 converts the distribution of the image data DT from the radiation image reader into a desired level distribution, and then the level distribution of the image data DT from the radiation image reader. A region (hereinafter referred to as “region of interest” or “ROI (Region Of Interest)”) is determined. By determining a representative value from the image data in the ROI set here and converting the representative value to a desired level, image data having a desired level distribution can be obtained.
[0113]
For example, the setting of the ROI to the rib cage in the chest front process is set according to the following procedure. First, the left and right lines are determined by the following S1 to S3.
S1: Find vertical projections (cumulative values in one direction of data) of the upper and lower parts of the image data and the part of the image data excluding the outside of the irradiation field (FIG. 15A, FIG. 15). b)).
S2: A point where the signal value has a minimum value (referred to as Pc) in a range of 1/3 of the central portion (1/3 * x to 2/3 * x in FIG. 15) from the obtained vertical projection. The middle line column (Xc).
S3: The vertical projection values obtained from the 1/3 column (2/3 * x, 1/3 * x in FIG. 15) of the left and right images to the outside (left and right direction) of the image. A point below the threshold value (Tl, Tr) is searched, and the first point is taken as the left end / right end (Xl, Xr) of the lung field. As the threshold value, the maximum value (Plx, Prx) of the projection value is updated from the column of Pc and 1/3 of the entire image toward the outside (left-right direction) of the image,
Tl = ((k1-1) * Plx + Pc) / k1
Tr = ((k2-1) * Prx + Pc) / k2
And Here, k1 and k2 are constants.
[0114]
Next, the upper and lower lines are determined by S4 and S5.
S4: Take a horizontal projection in the section determined in the above step (FIG. 15C).
S5: The horizontal direction obtained from the 1/4 and 1/2 lines (1/4 * y and 1/2 * y in FIG. 15) of the entire image from the upper and lower sides toward the outside (vertical direction) of the image A point whose projection value is less than or equal to the threshold value is searched for, and the first point is set as the upper and lower ends (Yt, Yb) of the right lung field. As the threshold, each of the whole image
The maximum value (Ptx, Pbx) of the projection value in the range of 1/4 * y to 1/2 * y and 1/2 * y to 4/5 * y and the line of the maximum value outside the image (vertical direction) Using the minimum projection value (Ptn, Pbn) in the range of
Tt = ((k3-1) * Ptx + Ptn) / k3
Tb = ((k4-1) * Pbx + Pbn) / k4
And Here, k3 and k4 are constants.
[0115]
Further, the parameters k1 to k4 used for obtaining the threshold value by the above formula are obtained empirically.
The setting of the ROI is not limited to the case where the profile is analyzed and set as described above. For example, as shown in JP-A-5-7578, the image data of each pixel and the discriminant analysis method are used. Threshold values determined by the above are compared, and an identification code is added to each pixel based on the comparison result. Labeling is performed for each pixel group having consecutive identification codes indicating that the threshold value is greater than or equal to the threshold value. The field area is extracted, and the ROI can be set so as to include the lung field and the subdiaphragm area on the basis of the extracted lung field area.
[0116]
Further, as disclosed in JP-A-62-26047, a lung field region is recognized by detecting a lung field contour using a boundary point tracking method, and the lung field and the subdiaphragm are based on the recognized lung field region. The ROI may be set so as to include the region.
[0117]
Furthermore, since it is generally performed that the most important part for diagnosis is taken as the center of the irradiation field, a region such as a circle or a rectangle is set at the center of the irradiation field region and the ROI is set. You can also
[0118]
(3) Frequency processing and dynamic range compression processing:
Further, frequency processing (frequency enhancement processing) and dynamic range compression processing are executed as image processing. In the frequency enhancement process, a frequency enhancement process for controlling the sharpness of the normalized radiation image is performed. In the dynamic range compression process, control is performed so that the density range is easy to see by the compression process.
[0119]
In the frequency enhancement process, for example, the function F is determined by a method shown in Japanese Patent Publication Nos. 62-62373 and 62-62376 in order to control the sharpness by the non-sharp mask processing shown in the following equation.
[0120]
Sout = Sorg + F (Sorg−Sus)
Sout is image data after processing, Sorg is image data before frequency enhancement processing, and Sus is unsharp data obtained by processing the image data before frequency enhancement processing by averaging processing or the like.
[0121]
In this frequency enhancement process, for example, F (Sorg−Sus) is set to β × (Sorg−Sus), and β (enhancement coefficient) is changed substantially linearly between the reference values T1 and T2 as shown in FIG. . As shown by the solid lines in FIG. 20, when values “A” and “B” are set and low luminance is emphasized, β from the reference value T1 to the value “A” is maximized, and the value “ B "to the reference value T2 is minimized. In addition, from the value “A” to the value “B”, β changes almost linearly. When emphasizing high luminance, as shown by a broken line, β from the reference value T1 to the value “A” is minimized and maximized from the value “B” to the reference value T2. In addition, from the value “A” to the value “B”, β changes almost linearly. Although not shown, when medium luminance is emphasized, β of values “A” to “B” is maximized. As described above, in the frequency enhancement processing, the sharpness of an arbitrary luminance portion can be controlled by the function F.
[0122]
Further, the frequency enhancement processing method is not limited to the non-sharp mask processing, and a technique such as a multi-resolution method disclosed in Japanese Patent Laid-Open No. 9-44645 may be used. In the frequency enhancement process, the frequency band to be enhanced and the degree of enhancement are set based on the imaging region, the imaging position, the imaging conditions, the imaging method, etc., as in the selection of the basic gradation curve in the gradation processing. .
[0123]
In the dynamic range compression processing, the function G is determined by the method disclosed in Japanese Patent Publication No. 266318 in order to perform control to make the density range easy to see by the compression processing shown in the following equation.
[0124]
Stb = Sorg + G (Sus)
Note that Stb is image data after processing, Sorg is image data before dynamic range compression processing, and Sus is unsharp data obtained by processing image data before dynamic range compression processing by averaging processing or the like.
[0125]
Here, as shown in FIG. 21A, G (Sus) has a characteristic that G (Sus) increases when the unsharp data Sus becomes smaller than the level “La”. The image data Sorg shown in FIG. 21B is image data Stb in which the dynamic range on the low density side is compressed as shown in FIG. 21C. Further, as shown in FIG. 21 (d), when G (Sus) has a characteristic that G (Sus) decreases when the unsharp data Sus becomes larger than the level “Lb”, The density is low, and the dynamic range on the high density side of the image data Sorg shown in FIG. 21B is compressed as shown in FIG. In the dynamic range compression processing, the correction frequency band and the degree of correction are set based on the imaging region, the imaging posture, the imaging conditions, the imaging method, and the like.
[0126]
Here, the reference values T1 and T2 and the values “A” and “B” or the levels “La” and “Lb”, which are processing conditions in the frequency enhancement processing and dynamic range compression processing, are representative value D1 and D2 determination methods. It is calculated | required by the same method.
[0127]
For the frequency processing at this stage, the method of the embodiment of the present invention already described with reference to FIGS. 1 to 6 can be used. That is, by performing arithmetic processing using a plurality of masks for every pixel of the image data sequentially for each pixel, a memory equal to the size of the image data is required when converting all of the image data. It is possible to obtain the converted image data without doing so. Therefore, effects such as saving of working memory capacity and shortening of calculation time can be obtained.
[0128]
(4) Tone processing:
First, representative values D1 and D2 are set from a cumulative histogram of image data in the ROI. The representative values D1 and D2 are set as image data levels at which the cumulative histogram has a predetermined ratio m1 and m2.
[0129]
When the representative values D1 and D2 are set, the normal values for converting the representative values D1 and D2 into desired reference values S1 and S2 as shown in FIG. 16 with reference to a normalization lookup table provided in advance. Processing is performed. Here, the characteristic curve CC indicates the level of a signal output in accordance with the radiation dose of the radiation applied to the imaging panel 41. The normalization processing lookup table is generated by calculation using an inverse function of the function indicating the characteristic curve CC of the imaging panel 41. Of course, normalization processing may be performed by arithmetic processing without using the normalization processing lookup table.
[0130]
As shown in FIG. 17, by this normalization processing, even if the radiation doses Ra to Rb are lower than the doses R1 to R2 at which the image data of the desired reference values S1 to S2 can be obtained, the desired reference Since the image data of the values S1 to S2 can be obtained, the exposure amount of the subject can be reduced, and at the same time, the variation in the signal distribution due to the difference in the body shape of the subject can be corrected.
[0131]
Next, gradation processing is performed using normalized image data DTreg obtained by normalization processing. In the gradation processing, for example, a gradation conversion curve as shown in FIG. 18 is used, and normalized image data DTreg is output with reference values S1 and S2 of normalized image data DTreg as parameter values S1 ′ and S2 ′. Converted to image data DTout. The levels S1 ′ and S2 ′ correspond to predetermined luminance or photographic density in the output image.
[0132]
The above tone conversion curve is preferably a continuous function over the entire signal region of the normalized image data DTreg, and its differential function is also preferably continuous. Moreover, it is preferable that the sign of the differential coefficient is constant over the entire signal region.
[0133]
In addition, since the preferable gradation conversion curve shape and levels S1 ′ and S2 ′ differ depending on the imaging region, imaging position, imaging conditions, imaging method, etc., the gradation conversion curve may be created for each image each time. For example, as shown in Japanese Patent Publication No. 5-26138, a plurality of basic gradation conversion curves are stored in advance, and one of the basic gradation conversion curves is read and rotated and translated. Thus, a desired gradation conversion curve can be easily obtained. The image processing unit 26 is provided with a gradation processing lookup table corresponding to a plurality of basic gradation curves, and is obtained by referring to the gradation processing lookup table based on the normalized image data DTreg. By correcting the image data according to the rotation and translation of the basic gradation conversion curve, output image data DTout subjected to gradation conversion can be obtained. In the gradation conversion process, not only the two reference values S1 and S2 but also one reference value or three or more reference values may be used.
[0134]
Here, the selection of the basic gradation curve and the rotation and translation of the basic gradation curve are performed based on the imaging region, imaging position, imaging conditions, imaging method, and the like. When these pieces of information are input as management information using the input device 27, the basic gradation curve can be easily selected and the rotation direction of the basic gradation curve by using this management information. And the amount of translation can be determined. Further, the levels of the reference values S1 and S2 may be changed based on the imaging region, the imaging posture, the imaging conditions, and the imaging method.
[0135]
Further, the selection of the basic gradation curve and the rotation or translation of the basic gradation curve may be performed based on information on the type of image display device and the type of external device for image output. This is because the preferred gradation may differ depending on the image output method.
[0136]
<Other embodiments>
In addition, it is not necessary to perform each process in the above (1)-(4) using all the information of a radiographic image, It is desirable to use the image reduced by the thinning process. In this case, since the amount of data is reduced, the processing speed can be improved and the memory capacity can be saved. In this case, it is preferable to use image data that has been thinned so that the effective pixel size of the reduced image by thinning is 0.4 mm to 10.0 mm, preferably 1.0 mm to 6.0 mm.
[0137]
In the above description, the example in which the image processing of the present embodiment is applied to a radiation image has been described. However, the image processing of the present embodiment can be applied to various images other than the radiation image. is there.
[0138]
【The invention's effect】
As described above in detail, according to the present invention, when image processing is performed on image data of an image composed of a plurality of pixels, the pixel value of the target pixel and the pixel values of its neighboring pixels and a predetermined function are used. The value of the pixel of interest after conversion is determined by a calculated value based on a plurality of convolution operations. Thereby, it is possible to realize an image processing apparatus and an image processing method capable of easily obtaining the converted signal value of the target pixel without performing conversion on all the image data. In addition, an image processing apparatus and an image processing method that can obtain converted image data without requiring a memory equal to the size of the image data when converting all the image data can be realized.
[0139]
That is, in the case where the converted value of the target pixel is obtained based on the weighted average value obtained by changing the mask size, the target pixel is averaged (weighted) using a plurality of size masks. Processing is performed to obtain a difference value between the average values of the masks having similar sizes, a conversion process is performed on the difference value, and the converted difference value is added to the pixel value of the target pixel to obtain a processed signal value. In this way, by performing all processing in units of pixels, it is possible to save the memory used and improve the processing speed.
[0140]
In addition, by applying to a medical image and further to a radiographic image, it is possible to provide an image that is very easy to recognize a lesion.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram showing an image processing procedure of a radiation image processing apparatus of the present invention.
FIG. 2 is an explanatory diagram showing an image processing procedure of the radiation image processing apparatus of the present invention.
FIG. 3 is a flowchart showing an image processing procedure of the radiation image processing apparatus of the present invention.
FIG. 4 is an explanatory diagram showing image processing of the radiation image processing apparatus of the present invention.
FIG. 5 is an explanatory diagram showing image processing of the radiation image processing apparatus of the present invention.
FIG. 6 is an explanatory view showing image processing of the radiation image processing apparatus of the present invention.
FIG. 7 is an explanatory diagram showing an image processing procedure of the radiation image processing apparatus of the present invention.
FIG. 8 is a configuration diagram showing a configuration of a radiographic image processing apparatus according to the present embodiment;
FIG. 9 is a configuration diagram showing a configuration of a radiation image reader according to the present embodiment.
FIG. 10 is an explanatory diagram for explaining a reading unit used in the present embodiment.
FIG. 11 is an explanatory diagram for explaining a reading unit used in the present embodiment;
FIG. 12 is a configuration diagram showing a configuration using another radiation image reader according to the present embodiment.
FIG. 13 is a configuration diagram showing a configuration of a controller according to the present embodiment.
FIG. 14 is an explanatory diagram for explaining irradiation field recognition processing according to the present embodiment example;
FIG. 15 is an explanatory diagram for explaining recognition of ROI in the present embodiment example;
FIG. 16 is an explanatory diagram for explaining level conversion in the present embodiment;
FIG. 17 is an explanatory diagram showing normalization processing in the present embodiment example;
FIG. 18 is an explanatory diagram showing gradation conversion characteristics in the present embodiment.
FIG. 19 is an explanatory diagram showing a relationship between an emphasis coefficient and image data in the present embodiment.
FIG. 20 is an explanatory diagram illustrating a relationship between an enhancement coefficient and image data in the present embodiment.
FIG. 21 is an explanatory diagram for describing dynamic range compression processing in the present embodiment;
[Explanation of symbols]
10 Controller
26 Image processing unit
30 Radiation generator
40,60 Radiation image reader
41 Imaging panel
44 Scanning drive unit
46 Image data generation circuit
48 Reading control circuit
61 Conversion panel
62 Light beam generator
63 Scanning section
64 Reflector
65 light collector
66 Filter
67 Photomultiplier
70 Current / voltage converter
75 Reading control unit

Claims (10)

An image processing apparatus comprising image processing means for performing image processing on image data of a medical image composed of a plurality of pixels photographed by radiation irradiation,
The image processing means determines the value of the pixel of interest after conversion based on a calculated value based on a plurality of convolution operations by a predetermined function for the pixel value of the pixel of interest and the pixel value of its neighboring pixels,
Said plurality of convolution predetermined function used for calculation is different in the average value calculation range to each other, there are multiple functions der obtaining a weighted average value by a Gaussian distribution or uniform distribution in the averaging range,
An image processing apparatus. Wherein the image processing means, the target pixel after conversion by adding the conversion value of the difference value was converted based on a predetermined conversion convolution calculated value of the function averaging range adjacent picture element value of the pixel of interest the image processing apparatus according to claim 1, in obtaining the values, characterized in that. The image processing means adds the converted conversion value based on the average value averaging range converts the difference value of the convolution computation of a function of a predetermined adjacent to determined by the average value calculating range widest function obtaining a value of the target pixel after conversion by the image processing apparatus according to claim 1, characterized in that. The predetermined conversion with respect to the difference value, the image processing apparatus according to claim 2 or claim 3 provided by a non-linear function, it is characterized with respect to the input value. The image processing apparatus according to claim 4, wherein the non-linear function has at least a part where an absolute value of an output value is smaller than an absolute value of an input value. The image processing apparatus according to claim 4 or claim 5 wherein the non-linear function is different from the average value calculation range determined difference, and wherein the. The nonlinear function as averaging range generates differential is wide, take a small output value in the vicinity of the input value of 0, that the image processing apparatus according to any one of claims 4 to 6, characterized in. The image processing means performs a calculation based on the pixel of interest and its neighboring pixels and a plurality of convolution operations of a predetermined function, has a sufficient memory area to store the value, and repeatedly uses the area the image processing apparatus according to any one of claims 1 to 7 to determine a value after conversion to a plurality of the pixel of interest, characterized in that by. The image processing apparatus according to any one of claims 1 to 8 wherein the plurality of convolution predetermined function used for calculation varies by the image data, it is characterized. An image processing method for performing image processing on image data of a medical image composed of a plurality of pixels photographed by radiation irradiation,
For the pixel value of the pixel of interest and the pixel values of the neighboring pixels, the average value calculation range is different from each other, and multiple convolution operations are performed by a plurality of functions for obtaining a weighted average value by Gaussian distribution or uniform distribution in the average value calculation range the calculated value determining the value of the target pixel after the conversion based on,
An image processing method.

Images