JPH04156532A - Radiation image reading condition and/or image processing condition determining method and device - Google Patents

Abstract

PURPOSE:To resolve various defects due to displacement and efficiently determine conditions with high precision by inputting the information on the center position of an object in an image to a neural network, and outputting the reading condition and/or image processing condition in consideration of the center position of the object when inputting the first image signal from this network and outputting the reading condition and/or image processing condition. CONSTITUTION:A preliminary read image signal is obtained from an accumulative phosphor sheet recording X-ray images with a preliminary reading means and inputted to a computer system. The computer system inputs the preliminary read image signal to a neural network 3 and inputs it to a means 2 determining the center position of an object in an image, and the reading condition for primary reading is determined from the image signal in consideration of the center position of the object in the network 3. The image signal obtained by a primary reading means from the accumulative phosphor sheet completed with preliminary reading is again inputted to the computer system, proper image processing is applied to the image signal, and the X-ray image based on this image signal is reproduced and displayed.

Description

Detailed Description of the Invention (Industrial Application Field) The present invention is based on an image signal representing a radiographic image, reading conditions for obtaining an image signal, image processing conditions for performing image processing on an image signal, etc. The present invention relates to a method and apparatus for determining desired radiation image reading conditions and/or image processing conditions.

(Prior Art) It is practiced in various fields to read a recorded radiation image to obtain an image signal, perform appropriate image processing on the image signal, and then reproduce and record the image. For example, an X-ray image is recorded using an X-ray film with a low gamma value designed to be compatible with later image processing, and the X-ray image is read from the film on which it is recorded and converted into an electrical signal. By converting this electrical signal (image signal) into a visible image after performing image processing on it and reproducing it as a visible image in a copy photograph, etc., a reproduced image with good image quality performance such as contrast, sharpness, and graininess can be obtained. (Refer to Japanese Patent Publication No. 61-5193).

In addition, the applicant has proposed radiation (X-rays, α-rays).

When irradiated with β rays, γ rays, electron beams, ultraviolet rays, etc., a portion of this radiation energy is accumulated, and then when irradiated with excitation light such as visible light, stimulable fluorescence exhibits stimulated luminescence depending on the accumulated energy. Radiographic image information of a subject such as a human body is partially recorded on a sheet of stimulable phosphor using a stimulable phosphor, and this stimulable phosphor sheet is scanned with excitation light such as a laser beam. The resulting stimulated luminescent light is read photoelectrically to obtain an image signal, and based on this image data, a radiation image of the subject can be recorded on a recording material such as a photographic light-sensitive material, a CRT, etc. A radiation image recording and reproducing system that outputs visual images has already been proposed (Japanese Patent Laid-Open No. 5
No. 5-12429, No. 5B-11395.

No. 55-163472, No. 56-104645, No. 55-116340, etc.).

This system has the practical advantage of being able to record images over a much wider range of radiation exposure compared to conventional radiographic systems using silver halide photography. In other words, in a stimulable phosphor, it is recognized that the amount of emitted light that is stimulated to emit light due to excitation after accumulation is proportional to the amount of radiation exposure over an extremely wide range.
Therefore, even if the amount of radiation exposure varies considerably due to various imaging conditions, the amount of stimulated luminescence emitted from the stimulable phosphor sheet can be read by the photoelectric conversion means by setting the reading gain to an appropriate value. By converting the radiation image into an electric signal and using this electric signal to output the radiation image as a visible image to a recording material such as a photographic light-sensitive material or a display device such as a CRT, a radiation image that is not affected by fluctuations in radiation exposure amount can be obtained. be able to.

In the above system, the stimulable phosphor sheet is scanned in advance with a low-level light beam before obtaining an image signal by reading it under optimal reading conditions depending on the dose of radiation applied to the stimulable phosphor sheet. Pre-reading is performed to read the outline of the radiation image recorded on this sheet, the pre-read image signal obtained by this pre-reading is analyzed, and then the sheet is irradiated with a high-level light beam and scanned, and this radiation image is There is also a system configured to perform actual reading in which image signals are obtained by reading under optimal reading conditions.

Here, reading conditions are a general term for various conditions that affect the relationship between the amount of stimulated luminescence light and the output of the reading device during reading, such as reading gain, scale factor, or , the power of excitation light during reading, etc.

Also, the high level and low level of the light beam are, respectively.
If the energy of the light beam irradiated per unit area of the sheet or the energy of stimulated luminescence light emitted from the sheet depends on the wavelength of the light beam (has a wavelength sensitivity distribution), It refers to the weighting energy after weighting the energy of the light beam irradiated per unit area of the sheet with the wavelength sensitivity.As a method of changing the level of the light beam, light beams of different wavelengths are used. methods to use, methods to change the intensity of the light beam itself emitted from a laser light source, methods to change the intensity of the light beam by inserting or removing an ND filter, etc. on the optical path of the light beam, methods to change the beam diameter of the light beam. Various known methods can be used, such as a method of changing the scanning density and a method of changing the scanning speed.

In addition, regardless of whether the system performs this pre-reading or the system that does not, the obtained image signal (including the pre-read image signal) is analyzed and the optimal image processing conditions are determined when applying image processing to the image signal. Some systems let you decide. The term "image processing conditions" as used herein is a general term for various conditions when performing processing on an image signal that affects the gradation, sensitivity, etc. of a reproduced image based on the image signal. The method of determining the optimal image processing conditions based on this image signal is not limited to systems using stimulable phosphor sheets, and for example, image signals are obtained from radiation images recorded on recording sheets such as conventional X-ray films. It is also applied to the system.

Reading conditions and/or image processing conditions (hereinafter referred to as reading conditions etc.) are based on the image signal (including the pre-read image signal).

), the algorithm is determined in advance from the results of statistically processing a large number of radiographic images (for example, see JP-A-80-1.85944 and JP-A-61-280183). .

As one of the conventionally employed algorithms, a method is known in which a histogram of an image signal is obtained and reading conditions etc. are determined based on this histogram.

The method of determining reading conditions, etc. based on this histogram is to find both the maximum and minimum values of the range required as image information from the histogram of the image signal, and then use For example, how to determine reading conditions so that information can be read accurately in book reading (
Find only the maximum value from the histogram (see Japanese Patent Laid-Open No. 60-156055), subtract a predetermined value from the maximum value, set it as the minimum value, and set the range between this maximum value and the minimum value as necessary. Method for determining the range of image information (Japanese Patent Laid-Open No. 80-1
(Refer to Publication No. 85944), or find only the minimum value from the histogram, set the value obtained by adding a predetermined value to the minimum value as the maximum value, and set the range between the minimum value and the maximum value as the range of necessary image information. (Japanese Unexamined Patent Publication No. 61-28016
In addition, there are methods using a differential histogram (see Japanese Patent Application No. 62-67302), methods using cumulative histograms (see Japanese Patent Application Laid-open No. 170730/1988), and methods that use a histogram to create multiple small regions based on discrimination criteria. There are many known methods of determining reading conditions, etc., such as a method of dividing the image into two (see Japanese Patent Application No. 62-96718).

However, in the above-mentioned method using histogram analysis to determine reading conditions etc. based on the histogram of the image signal, various characteristic values are calculated by local analysis using threshold processing, so local features of the histogram are emphasized. If you do too much, you may get incorrect results.

Therefore, it may be possible to apply a method using neural network software that has appeared in recent years to this field.

This neural network is developed by inputting information (teacher signal) on whether the output signal outputted when a certain input signal is given is a correct signal or an incorrect signal. It is equipped with an error back propagation learning (pack propagation) function that corrects the weights of the connections between the nodes (the weights of the synaptic connections). This can sometimes increase the probability of outputting the correct answer.

By using this neural network, it is possible to determine the above-mentioned reading conditions and the like by inputting image data of a radiation image.

That is, the image data of the radiation image is input to the neural network, the reading conditions, etc. are outputted, and the neural network is made to repeatedly 'learn' in advance, so that correct reading conditions, etc. can be gradually determined. I can do it.

(Problem to be Solved by the Invention) However, in the method of determining reading conditions etc. using the above-mentioned neural network, the given image data is received as is and judgments are made regarding the entire image data. The position of is standard (2nd A
(Fig. 2B), the same weight is given to the so-called omitted portion where there is no subject in the image, and calculations are performed for judgment, resulting in incorrect results. Sometimes I end up giving out. In addition, since learning is repeated by inputting image data as is, if a large number of images with such positional shifts are input, a huge amount of time will be required for learning, which is difficult to implement realistically. There is a problem.

Therefore, the present invention provides a method for determining radiation image reading conditions and/or image processing conditions efficiently and with high accuracy by eliminating various drawbacks associated with positional deviations as described above when using a neural network. The purpose of the invention is to provide a device and a device.

(Means for Solving the Problems) One of the present inventions uses the above-mentioned stimulable phosphor sheet,
It is used in a system that performs pre-reading, and the method for determining radiation image reading conditions and/or image processing conditions is to irradiate excitation light onto a stimulable phosphor sheet on which a radiation image of a subject has been recorded, and to release the stimulable phosphor. Based on a first image signal representing the radiation image obtained by reading the stimulated luminescence light emitted from the sheet, the stimulable phosphor sheet is irradiated with excitation light again and emitted from the stimulable phosphor sheet. A radiation source for determining reading conditions for reading the stimulated luminescence light obtained to obtain a second image signal representing the radiation image and/or image processing conditions for performing image processing on the obtained second image signal. In the method for determining image reading conditions and/or image processing conditions, when the first image signal is input to a binary network and the neural network outputs the reading conditions and/or the image processing conditions, the neural Information regarding the center position of the object in the image is input to a network, and the neural network outputs the reading condition and/or the image processing condition from the image signal in consideration of the center position of the object. This is a characteristic feature.

The apparatus for determining radiation image reading conditions and/or image processing conditions according to the present invention, which implements the above method, irradiates excitation light onto a stimulable phosphor sheet on which a radiation image of a subject is recorded. Based on the first image signal representing the radiation image obtained by reading the emitted stimulated luminescence light, the stimulable phosphor sheet is irradiated with excitation light again, and the excitation light is emitted from the stimulable phosphor sheet. Radiation image reading for determining reading conditions when reading stimulated luminescence light to obtain a second image signal representing the radiation image and/or image processing conditions when performing image processing on the obtained second image signal. A condition and/or image processing condition determining device, comprising: means for determining and outputting the center position of a subject in the image from the first image signal; and an output of the subject center position determining means and the first image signal. and a neural network that outputs the reading conditions and/or the image processing conditions based on the first image signal while taking into account the center position of the subject.

Another method according to the present invention is to determine image processing conditions not limited to stimulable phosphor sheets.

In other words, this radiation image processing condition determination method determines image processing conditions for performing image processing on an image signal based on an image signal representing a radiation image of a subject. When inputting the image processing conditions to a network and outputting the image processing conditions from the neural network, information regarding the center position of the subject in the image is input to the neural network, and in the neural network,
The present invention is characterized in that the image processing condition is output from the image signal in consideration of the center position of the subject.

A device for implementing this is a radiation image processing condition determination device that determines image processing conditions for performing image processing on an image signal based on an image signal representing a radiation image, and a radiation image processing condition determination device that determines image processing conditions for performing image processing on the image signal, means for determining and outputting the center position of a subject; and inputting the output of the subject center position determining means and the image signal, and determining the image processing conditions based on the image signal while taking the center position of the subject into consideration. It is characterized by consisting of a neural network that outputs.

(Operations and Effects of the Invention) The method for determining radiation image reading conditions and/or image processing conditions according to the present invention includes: inputting an image signal to a neural network and outputting reading conditions and/or image processing conditions from the neural network; Information regarding the center position of a subject in an image is input to a neural network, and the neural network outputs reading conditions and/or image processing conditions from the image signal, taking into account the center position of the subject. Even if there is a shift in the position of the image, the neural network can always accurately capture the position of the subject and perform calculations for judgment on image data similar to the standard image data of the subject, so it is efficient and highly efficient. Radiation 1 image reading conditions and/or image processing conditions can be determined with precision.

Note that as a method for determining the center position of a subject in an image, for example, the method described in Japanese Patent Laid-Open No. 2-28782 can be adopted.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the basic concept of the present invention.

That is, an apparatus for carrying out the radiation image reading condition and/or image processing condition determination method of the present invention includes means 2 for determining and outputting the center position of the subject in the image from the image signal 1 of the radiation image of the subject; a neural network 3 that inputs the output of the subject center position determining means 2 and the image signal 2, and outputs the reading condition and/or the image processing condition based on the image signal while taking the center position of the subject into consideration; It is equipped with the following.

As a result, when the image signal 1 is input to the neural network 3 and the neural network 3 outputs the reading conditions and/or the image processing conditions, the secondary network 3 is provided with information about the subject in the image. Information regarding the center position may be input, and the neural network 3 may output the reading conditions and/or the image processing conditions from the image signal in consideration of the center position of the subject.

Next, an example of an X-ray image reading apparatus incorporating a computer system to which the radiation image reading condition and/or image processing condition determining method according to the present invention is applied will be described in detail.

FIG. 3 is a perspective view showing an example of a computer system including an example of an X-ray image reading device and an example of the radiation image reading condition and/or image processing condition determining device of the present invention. This system uses the aforementioned stimulable phosphor sheet and performs pre-reading.

In an X-ray imaging device (not shown), an X-ray image of a subject is accumulated and recorded on a stimulable phosphor sheet.

The stimulable phosphor sheet 11 on which this X-ray image is recorded is
First, the sheet 11 is set at a predetermined position 00 by scanning with a weak light beam to emit only a portion of the radiation energy accumulated in the sheet 11. The stimulable phosphor sheet 11 set at a predetermined position is conveyed (sub-scanned) in the direction of arrow Y by a sheet conveying means 13 such as an endless belt driven by a motor 12. On the other hand, a weak light beam 15 emitted from a laser light source 14 is reflected and deflected by a rotating polygon mirror 16 that is driven by a motor 23 and rotates at high speed in the direction of the arrow, passes through a focusing lens 17 such as an fθ lens, and then passes through a mirror 18 to is incident on the sheet 11 in the sub-scanning direction (arrow Y
Main scanning is performed in the direction of arrow X, which is substantially perpendicular to the main direction. From the part of the sheet 11 irradiated with this light beam 15, stimulated luminescence light 1 of a light amount corresponding to the radiographic image information stored and recorded is emitted.
9 is emitted, and this stimulated luminescence light 19 is guided by a light guide 20 and passed through a photomultiplier (photomultiplier tube) 2.
1 is photoelectrically detected. The light guide 20 is made by molding a light guide material such as an acrylic plate, and the linear entrance end surface 20a is a stimulable phosphor sheet.]
The light receiving surface of the photomultiplier 21 is coupled to an annular output end surface 20b extending along the main scanning line on the photomultiplier 1. The stimulated luminescent light 19 that has entered the light guide 20 from the incident end surface 20a travels through the interior of the light guide 20 through repeated total reflection, and then travels through the light guide 20 through the exit end surface 20a.
The photomultiplier 21 emits the light and receives the light from the photomultiplier 21, and the photomultiplier 21 converts the amount of stimulated luminescence light 19 representing a radiation image into an electrical signal.

The analog output signal S output from the photomultiplier 21 is logarithmically amplified by a logarithmic amplifier 26 and digitized by an A/D converter 27 to obtain a pre-read image signal Sp. The signal level of this pre-read image signal Sp is
It is proportional to the logarithm of the amount of stimulated luminescence light emitted from each pixel of 1.

In the above-mentioned pre-reading, the reading conditions, such as the voltage value applied to the photomultiplier 2I and the amplification factor of the logarithmic amplifier 26, are set so that the radiation energy accumulated in the stimulable phosphor sheet 1 can be read over a wide range. It is determined.

The obtained pre-read image signal Sp is sent to the computer system 4
It is input to 0. This computer system 40 includes an example of the radiation image reading condition and/or image processing condition determination device of the present invention, and includes a main body 41 having a built-in CPU and internal memory, and a floppy disk inserted as auxiliary memory. Drive unit 4 that is driven
2. A keyboard 43 for the operator to input necessary instructions to the computer system 40. and a CRT display 44 for displaying necessary information. - Within this computer system 40, division patterns and irradiation fields are recognized as necessary based on the input pre-read image signal Sp, and then, based on the pre-read image signal Sp, a neural network performs reading during main reading. Conditions, namely sensitivity Sk and latitude Gp during main reading
is determined, and the determined sensitivity Sk and latitude c
For example, the voltage value applied to the photomultiplier 21', the amplification factor of the logarithmic amplifier 26', etc. are controlled according to p. At this time, information on the subject center position is input to the neural network together with the pre-read image signal Sp, and the neural network outputs the sensitivity Sk and latitude Gp from the pre-read image signal Sp in consideration of the subject center position. be done.

Here, the latitude Gp corresponds to the light intensity ratio of the most intense stimulated luminescence light to the weakest stimulated luminescence light that is converted into an image signal during main reading, and the sensitivity Sk corresponds to the ratio of the light amount of the most intense stimulated luminescence light to the weakest stimulated luminescence light that is converted into an image signal during main reading. This refers to the photoelectric conversion rate that determines the level of image signal generated from stimulated luminescence light.

The stimulable phosphor sheet 11' for which pre-reading has been completed is set at a predetermined position in the main reading means 100', and the sheet 11' is scanned by a light beam 15' that is stronger than the light beam used for the above-mentioned pre-reading. An image signal is obtained according to the reading conditions determined by the above, but since the structure of the main reading means 100' is different from the structure of the above-mentioned pre-reading means 100, the components corresponding to each component of the pre-reading means 100 are , the number used in the pre-reading means 100 is shown with a dash attached, and the explanation will be omitted.

The image signal S0 obtained by being digitized by the A/D converter 27' is sent to the computer system 40 again.
is input. Appropriate image processing is performed on the image signal SΩ within the combination coater system 40, and the image signal subjected to this image processing is sent to a reproducing device (not shown), and the reproducing device reproduces an X-ray image based on the second image signal. Is displayed.

In the computer system 40, the pre-read image signal Sp is input to the neural network 3, and at the same time, the pre-read image signal Sp is input to the means 2 for determining the center position of the object in the image, and the pre-read image signal Sp is input to the neural network 3. In step 3, reading conditions for actual reading are determined from the image signal in consideration of the center position of the subject.

The method for determining the center position of the subject is disclosed in Japanese Unexamined Patent Publication No. 2-2
The method disclosed in No. 8782 is effective.

That is, based on the image signal obtained from each pixel on a recording sheet such as a stimulable phosphor sheet or photographic film on which a radiation image including a subject image is recorded, the image signal value corresponding to each pixel or its reciprocal is determined. Find the center of gravity of the recording sheet by weighting each corresponding pixel with
Let it be an image point within the subject image.

Alternatively, after obtaining an image signal in the same manner as above, based on this image signal, when the image signal value corresponding to each pixel or the reciprocal of this image signal value is associated with each corresponding pixel, recording is performed. For each of the two different directions on the sheet, calculate the cumulative distribution in which the image signal value or its reciprocal is accumulated and plotted in each of the above directions, and for each of these cumulative distributions, calculate the value to approximately half the maximum cumulative value. It is also possible to obtain coordinate points in each corresponding direction and use the positions on the recording sheet determined by these coordinate points as image points in the subject image.

Further, in determining whether to use the image signal value or its reciprocal in the above two methods, the following method can be used. That is, after obtaining an image signal, based on this image signal, a first representative value representing the image signal value corresponding to the peripheral area of the recording sheet and an image signal corresponding to the entire or approximately central area of the recording sheet are determined. a second representative value that represents the value, compare the magnitude of the first representative value and the second representative value, and select either the image signal value or its reciprocal in accordance with the comparison result. I can do it.

Hereinafter, a method for causing the neural network to repeatedly perform learning and output correct reading conditions will be described in detail.

FIG. 4 is a diagram showing an example of a neural network equipped with an error backpropagation learning (pack propagation) function. Error backpropagation learning (pack propagation) is, as mentioned above, by comparing the output of the neural network with the correct answer (teacher signal), and sequentially calculating connection weights (synaptic connection weights) from the output side to the input side. It is a “learning algorithm” that corrects the

As shown in the figure, the first layer (input layer), second layer (middle layer), and third layer (output layer) of this neural network are composed of 11 units, n2 units, and 2 units, respectively. Each signal F1. input to the first layer (input layer). F2.・
...IFIII is the pre-read image signal sp corresponding to each pixel of the X-ray image, and the two outputs y:+F2 from the third layer (output layer) are used for the sensitivity and contrast during main reading, respectively. This is a corresponding signal. The i-th unit of the layer is ui, each input to the unit 7 is X:, each output is Y
7 s u? to :Wkll the weight of the connection to l
It is assumed that each unit 7 has the same characteristic function. At this time, input X:, output y of each unit 7:
is X knee 8w7-”1 ・y knee ・(4) Y+
-f (X+) ... (5)
becomes. Each unit u configuring the input layer: (i=
], 2. −． nH) Each input F1, F2 to
．． -,.

F++1 is not weighted and is directly applied to each unit u:
(+=1, 2..., nl). The input 01 signals Fヱ, F2. ... +Ffil is weighted by the weight Wk k + l of each connection and the final output Y? The reading conditions (sensitivity and contrast) for actual reading are determined.

Here, a method of determining the weight w L k + l of each of the above-mentioned connections will be explained. First, the initial value of the weight Wkli+1 of each connection is given by a random number. At this time, even if the input F, ~il changes to the maximum, the output Y: +
It is preferable to limit the range of the random number so that Y2 is a value within a predetermined range or a value close to this.

A number of X-ray images or stimulable phosphor sheets with known optimal reading conditions are read as described above to obtain a pre-read image signal sp, thereby obtaining the above 01 inputs F1.
+F2,...,Fn) are calculated. This 01
The inputs Fl, F2 + . . . , Fo, are input to the neural network shown in FIG. 4, and each unit U
, output 7 can be monitored.

Once each output y7 is determined, the final output is )'l
l y2 and the teacher signal (sensitivity y1 and contrast 1
, ρ1 El −(J'+−y+)2 ...(6)1
3 Ez-(F2 F2)2...(7) is obtained. These squared errors E1. The weight of each connection w::" is set as follows so that each of E2 is minimized.
will be corrected. In addition, the following describes the output of y:, y2
About 3'? Since it is the same as , it is omitted here.

To minimize the squared error E1, the weight w::'I of each connection is modified as 20E1 is a function of W77'', where η is a coefficient called a learning coefficient.

Here, and from equation (4), x: 1-Σ 1i++ , y, -(4)'
Therefore, equation (9) becomes as follows.

Here, from equation (6), if we transform equation (11) using equation (5), then from equation (3), t' (x) -f (x) (1-f (x )
) −(13), so f' (X+ )−Y: ・ (I Y?) ・
...(14).

If we put -2 in equation (10) and substitute equations (12) and (14) into equation (10), we get -(V knee 7)・yl・(13/?)・y.

...(15) Substituting this equation (15) into equation (8), W? : -W
? ? -η・(y: d)・y:, (ly:)
・yl...(16). This (1B)
According to the formula, W? ? (i-1, 2,...

nl) are modified.

Next, since , substituting equations (4) and (5) into equation (17), we obtain f' (X?) −y”+ ・(13/: ) −
(19), so this equation (19) and (12),
Substituting equation (14) into equation (18), ・y,・ (1y+) ・W4. ...(20)(
Put -1 in equation 10) and change equation (20) to (lO)
Substituting into the formula, we get: ・7 tō・ (1-y?)・W''1 ・yl...(21
) Substituting this equation (21) into equation (8), replacing k-1, we get 3goWl , -Wl , -η・(y+ y+)
・yl・ (1-yo)・yl ・(1-y+)・y
;・W4. ...(
22), and Wl r (1
-1, 2,...

nx) is substituted into this equation (22), and Wl''(i''1
,2゜...,nl ;j-1,L...,nl
) will be corrected.

Theoretically, by using equations (16) and (22) and setting the learning coefficient η sufficiently small and increasing the number of learnings sufficiently, the weight w of each connection can be set to a predetermined value. Although it is possible to converge, it is not realistic to make the learning coefficient η too small because it slows down the progress of learning.On the other hand, the learning coefficient η
If is set to a large value, learning may oscillate (the weights of the connections described above may not converge to a predetermined value). Therefore, in reality, we add an inertia term as shown in the following equation to the amount of correction of the weight of the connection to suppress the vibration.

The learning coefficient η is set to a somewhat large value.

(e.g., D.E., Rumelhart, G.E., Hi
nton and R1゜W11th ass: Learn
ing 1internal representative
onsbyerrorpropagationlnPa
ral]elDistributedProcess
ing, Volume IJ, L, McClell! a
nd, D, E.

Rumelhart and The PDP
Re5search Group, HITPress,
(Refer to L986b J) ΔWf”r” (t+1)−α・ΔW? : ””(
t) Tens ΔW:;'' (t) represents the amount of correction obtained by subtracting the weight of the connection before correction W::'' from the connection weight after correction V/7:'' in the second learning. , α are coefficients called inertia terms.

For example, α-0, 9η-0 as the inertia term α and the learning coefficient η.
, 25 to modify (learning) the weight Wi of each connection: “1”
For example, 200,000 times, and then the weight of each connection W
? :" is fixed at the final value. At the end of this learning, the two outputs V:+YH become signals that correctly represent the sensitivity and contrast during actual reading, respectively.

After the learning is completed, the histogram is obtained from the pre-read image signal representing the X-ray image at the time of pre-reading, and this is input to the neural network shown in FIG. 4, and the resulting output y? Reading conditions (sensitivity and contrast) for main reading for +3'2 or above X-ray images
It becomes a signal representing. Since this signal is obtained after learning as described in L, it accurately represents the reading conditions for actual reading.

It should be noted that the neural network described above is not limited to a three-layer structure, and it goes without saying that it may have even more layers. Furthermore, it goes without saying that each layer can be configured with any number of units depending on the number of pixels of the input pre-read image signal sp, the accuracy of the required reading conditions, etc.

In the above method of determining reading conditions etc. using only a neural network, the given image data is accepted as is and judgments are made on the entire image data, so the position of the subject on the screen is standard (see Figure 2A). )
If there is a large deviation from the image (Figure 2B), the so-called omitted portion where there is no subject at all is given the same weight as the subject part of the image and calculations are performed for judgment.
It may give incorrect results. In addition, since learning is repeated by inputting image data as is, if a large number of images with such positional shifts are input, it will be necessary to spend a huge amount of time on learning, which is difficult to implement realistically. .

Therefore, in the present invention, when using a neural network, the image signal is sent to the neural network so that the radiation image reading conditions can be determined efficiently and with high accuracy even if there is a positional deviation as described above. When inputting and outputting the reading conditions from the neural network, information regarding the center position of the subject in the image is input to the neural network, and the neural network calculates the image signal by taking the center position of the subject into account. The entire object is moved as a whole according to the deviation of the center position, and the object center position of the image signal when the neural network learns matches the object center position of the input image signal, and then the reading conditions are set. This outputs the following.

Thereby, even if there is a positional shift as described above, the radiation image reading conditions can be determined efficiently and with high accuracy.

The voltage applied to the photomultiplier 21' of the main reading means 100', the amplification factor of the amplifier 26', etc. are controlled according to the reading conditions determined by the neural network as described above, and the main reading is performed according to the controlled conditions. It can be done.

In the above embodiment, the pre-reading means 1.00 and the main reading means 10
0' are configured separately, but as mentioned above, the configuration of the pre-reading means 100 and the main reading means 100' is between the two, so the pre-reading means 100 and the main reading means 100' are integrated. May also be used for both purposes. In this case, after performing the pre-reading, the stimulable phosphor number [1] may be returned, and the main reading may be performed by scanning again.

If the pre-reading means and the main reading means are used, is it necessary to switch the intensity of the light beam between the pre-reading and the main reading? Various methods can be used, such as a method of switching the intensity itself and a method of using a filter.

Furthermore, in the above embodiment, a method for obtaining reading conditions for book reading using the computer system 40 was explained.
The computer system 40 can also determine image processing conditions when performing image processing on the image signal S0.

That is, the method of determining reading conditions by the computer 40 using a histogram and a neural network can also be applied to determining image processing conditions when performing various types of image processing on an image signal.

In this case, the actual reading is performed under predetermined reading conditions regardless of the pre-read image signal Sp, and the computer system 40 may determine the image processing conditions based on the pre-read image signal Sp. Further, the computer system 40 may determine both the reading conditions and the image processing conditions.

Further, in the above embodiment, the present invention is applied to a radiation image reading method that performs pre-reading, but the present invention is also applicable to a radiation image reading method that performs reading equivalent to main reading without pre-reading.

In this case, image processing conditions are determined by histogram analysis from an image signal obtained by reading an appropriate method, and a computer system comprising a neural network adds correction to this to determine appropriate image processing conditions.

Furthermore, although the above embodiment for determining image processing conditions is based on reading images recorded on a stimulable phosphor sheet, the present invention is applicable not only to radiation images recorded on a stimulable phosphor sheet. It goes without saying that the present invention can also be applied to cases in which image processing is performed on signals obtained by reading images such as medical images recorded on conventional X-ray film using an appropriate method.

The optimal image processing conditions thus determined are input to an image processing device, and the image processing device performs image processing, such as gradation processing, on the input image signal under the optimal image processing conditions. .

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the basic configuration of the method for determining image reading conditions and/or image processing conditions of the present invention. FIGS. 2A and 2B are two images in which the subject center position is shifted from the screen. By way of example, FIG. 3 shows an X
FIG. 4 is a perspective view showing an example of a line image reading device. FIG. 4 is a diagram showing an example of a neural network used in the method of the present invention. 1... Image signal 2... Subject center determining means 3... Neural network 11, 11'... Stimulative phosphor sheet 19, 19'
... Stimulated luminescent light 21, 21'... Photo multiplier 26, 26'
...Logarithmic amplifiers 27, 27'...A/D converter 40...Computer system 100'...Main reading means Sp...Pre-read image signal

Claims (4)

[Claims] (1) A first image representing the radiation image obtained by irradiating a stimulable phosphor sheet on which a radiation image of a subject is recorded with excitation light and reading the stimulated luminescence light emitted from the stimulable phosphor sheet. Based on the image signal, the stimulable phosphor sheet is irradiated with excitation light again and the stimulated luminescence light emitted from the stimulable phosphor sheet is read to obtain a second image signal representing the radiation image. In a method for determining radiation image reading conditions and/or image processing conditions for determining reading conditions and/or image processing conditions for performing image processing on the obtained second image signal, the first image signal is applied to a neural network. When inputting the reading conditions and/or the image processing conditions from the neural network, input information regarding the center position of the object in the image to the neural network, and input the information regarding the center position of the object in the image to the neural network. A method for determining radiation image reading conditions and/or image processing conditions, characterized in that the reading conditions and/or the image processing conditions are output from the image signals in consideration of the position. (2) A first image representing the radiation image obtained by irradiating a stimulable phosphor sheet on which a radiation image of the subject is recorded with excitation light and reading the stimulated luminescence light emitted from the stimulable phosphor sheet. Based on the image signal, the stimulable phosphor sheet is irradiated with excitation light again and the stimulated luminescence light emitted from the stimulable phosphor sheet is read to obtain a second image signal representing the radiation image. In a radiation image reading condition and/or image processing condition determination device for determining reading conditions and/or image processing conditions for performing image processing on the obtained second image signal, means for determining and outputting the center position of a subject; inputting the output of the subject center position determining means and the first image signal; A radiation image reading condition and/or image processing condition determination device comprising: a neural network that outputs the reading condition and/or the image processing condition based on the neural network. (3) In a radiation image processing condition determination method for determining image processing conditions for performing image processing on an image signal based on an image signal representing a radiation image of a subject, the image signal is input to a neural network, and the neural network When outputting the image processing conditions from the network, information regarding the center position of the object in the image is input to the neural network, and the neural network converts the image signal from the image signal in consideration of the center position of the object. A radiation image processing condition determination method characterized by outputting processing conditions. (4) In a radiation image processing condition determination device that determines image processing conditions for performing image processing on an image signal based on an image signal representing a radiation image, determining a center position of a subject in the image from the image signal. and a neural network that inputs the output of the object center position determining means and the image signal, and outputs the image processing condition based on the image signal while considering the center position of the object. A radiation image processing condition determination device characterized by: