JP2010075368A - Apparatus, method, and program of electronic endoscope - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a normal image or a diagnostic image where it is easy to detect whether a site operated by zooming to an image reader is a lesion or not. <P>SOLUTION: An endoscope apparatus forms a normal image showing a biological mucous and a diagnostic image as a spectroscopic image in a narrow area showing the biological mucous, and the endoscope apparatus includes a switch 21 which can receive an instruction for zooming to the normal image or the diagnostic image at a predetermined magnification, a zooming part (anyone of a zooming mechanism 5, a signal processing circuit (B) 26, and a signal processing circuit (C) 33) which zooms according to the instruction for zooming received by the switch 21 at the predetermined magnification, an image processing circuit 27 changing a plurality of image processing conditions according to the predetermined magnification and operating the image processing based on the plurality of changed image processing conditions to the normal image or the diagnostic image, and a monitor 34 displaying the normal image or the diagnostic image processed by image processing. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an electronic endoscope apparatus, method, and program, and more particularly to an electronic endoscope apparatus that creates a narrow-band spectral image of a biological mucosa and diagnoses the biological mucosa.

  Conventionally, as an electronic endoscope apparatus using a solid-state image sensor, a normal image representing a visible light region image mainly by imaging a visible light wavelength region and a plurality of types of narrow light transmitting light only in a narrow wavelength band. A method of imaging a biological mucous membrane of a digestive organ (for example, stomach) through a band-pass filter to obtain a plurality of types of narrow-band spectral images of the biological mucosa and generating a diagnostic image by combining these narrow-band spectral images (Narrow Band Imaging-NBl) is known.

  Such an apparatus is equipped with a rotary filter that combines three types of narrow-band bandpass filters that transmit light in different wavelength ranges, and obtains a plurality of types of narrow-band spectral images by performing imaging in a frame sequential manner. A plurality of types of narrow-band spectroscopic images (diagnostics) showing the biological mucous membrane by imaging the biological mucosa while sequentially irradiating the biological mucous membrane with illumination light that has been dispersed through the three types of narrow-band bandpass filters similar to the above And so on that are also considered.

  The biological mucous membrane diagnostic image obtained by synthesizing a plurality of types of narrow-band spectral images obtained in this way can express the fine structure of the biological mucous membrane that has not been obtained conventionally.

  In such a system (Narrow Band Imaging-NBl), in order to make it easy for the reader to make a diagnosis, appropriate parameters for each tissue can be displayed appropriately so that any biological tissue can be displayed. And performing image processing on a diagnostic image signal using parameters corresponding to each tissue has been proposed (see Patent Document 1).

  On the other hand, the present invention relates to an electronic endoscope apparatus that takes a normal color image (also referred to as a normal image) in a plane simultaneous manner by arranging RGB mosaic filters composed of a plurality of types of wideband bandpass filters on a solid-state image sensor. A method of creating a diagnostic image as described above by acquiring an image equivalent to the narrow-band spectral image obtained by using the narrow-band bandpass filter by arithmetic processing based on a normal image obtained by imaging a biological mucous membrane Has been proposed.

  This method performs a principal component analysis for estimating the spectral reflection intensity distribution of the biological mucosa using a large number of measurement data related to the spectral reflection intensity distribution of the biological mucosa in the visible wavelength range. It has been proposed by finding that the three principal components of the third principal component can substantially restore the spectral reflection intensity distribution over the entire visible wavelength range of the biological mucous membrane. According to this restoration technique, spectral reflection estimation matrix data obtained in advance using a large number of measurement data relating to the spectral reflection intensity distribution of the biological mucous membrane and imaging through normal RGB mosaic filters corresponding to the three principal components are used. Obtained by synthesizing narrowband spectral images in a plurality of wavelength regions obtained by spectral image estimation calculation with the normal image obtained in step 1, and synthesizing the narrowband spectral images obtained through the narrowband bandpass filter. An image equivalent to the diagnostic image can be obtained in a pseudo manner (see Patent Document 2).

Also, an electronic endoscope apparatus that displays a normal image has been proposed that automatically changes the brightness of the normal image in accordance with the zoom magnification (see Patent Document 3).
JP 2006-341078 A JP 2003-93336 A JP 2001-37711 A

  However, the conventional electronic endoscope apparatus has various problems as described below. As in the device described in Patent Document 1, in order to make it easy for an interpreter to diagnose, an appropriate parameter is stored for each tissue so that an appropriate diagnostic image signal can be displayed in any living tissue. In a case where image processing is performed on a diagnostic image signal using parameters corresponding to each tissue, it is necessary to set a plurality of parameters in advance for each specific tissue, and the setting becomes complicated There is.

  In general, a target tissue on an image on which an image interpreter performs zoom processing is often a target tissue (for example, a site having a high possibility of a lesion).

  However, if there is no function for automatically changing the image processing conditions for an image on which zoom processing is performed as in the device described in Patent Document 2, there is a problem that it is difficult for an image interpreter to determine the tissue of interest.

  Further, as in the apparatus described in Patent Document 3, if only the brightness in the normal image is automatically changed according to the zoom magnification, there is a problem that it is difficult for the radiogram interpreter to accurately determine the target tissue.

  Therefore, in the electronic endoscope apparatus that creates the normal image and the diagnostic image, the plurality of image processing conditions are changed according to the predetermined magnification value, and the normal image or the It is desired to control the diagnostic image so as to perform image processing.

  Accordingly, the present invention has been made in view of the above circumstances, and creates a normal image and a desired diagnostic image, and determines whether a site to be subjected to zoom processing is a lesion or the like for an image interpreter. It is an object of the present invention to provide an electronic endoscope apparatus, method, and program for displaying a normal image or a diagnostic image that can be easily discriminated.

  The electronic endoscope apparatus of the present invention generates a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucosa, and the normal image or the diagnostic image is In response to the zoom instruction receiving unit capable of receiving an instruction to perform zoom processing at a predetermined magnification value, and the instruction to perform zoom processing received by the zoom instruction receiving unit, for a normal image or a diagnostic image, A zooming unit that performs zoom processing at a predetermined magnification value, and a plurality of image processing conditions are changed according to the predetermined magnification value, and a normal image or a diagnostic image is changed based on the changed plurality of image processing conditions. An image processing unit that performs image processing and a display unit that displays a normal image subjected to image processing or a diagnostic image subjected to image processing are provided.

  “Normal image” refers to a color image representing a visible light region image by imaging the wavelength region of visible light.

  The “diagnostic image” refers to an image composed of a narrowband spectral image used for diagnosis. For example, it may be an image obtained by spectral image estimation calculation based on a wavelength set selected from a plurality of types of wavelength sets, or a plurality of types of narrowband spectral images may be obtained and these narrowband spectral images may be obtained. It may be an image generated by using the Narrow Band Imaging-NBl method.

  The “zooming unit” refers to a unit that performs an enlargement process or a reduction process by an optical zoom process by changing a focal length of a lens of a zoom mechanism unit by controlling a zoom drive circuit. Further, not only the optical zoom described above but also a signal processing circuit may perform enlargement processing or reduction processing by electronic zoom processing.

  The “plurality of image processing conditions” relates to a correction amount of at least two of brightness enhancement processing, structure enhancement processing, and color gradation processing. Furthermore, it may have a display setting for either the normal image or the diagnostic image.

  The electronic endoscope apparatus of the present invention further includes a database unit that stores a plurality of image processing conditions corresponding to each predetermined magnification value in advance, and the image processing unit stores the predetermined magnification value stored in the database unit. Image processing may be performed based on a plurality of corresponding image processing conditions.

  The electronic endoscope display method of the present invention generates a normal image representing a biological mucosa and a diagnostic image which is a narrow-band spectral image representing the biological mucosa, and has a predetermined magnification with respect to the normal image or the diagnostic image. In response to an instruction to perform zoom processing with a value, in response to the received instruction to perform zoom processing, perform zoom processing with a predetermined magnification value, change a plurality of image processing conditions according to the predetermined magnification value, The normal image or the diagnostic image is subjected to image processing based on a plurality of changed image processing conditions, and the normal image subjected to the image processing or the diagnostic image subjected to the image processing is displayed.

  The electronic endoscope display program of the present invention generates a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucous membrane. A function for receiving an instruction to perform zoom processing at a predetermined magnification value, a function for performing zoom processing at a predetermined magnification value in response to the received instruction to perform zoom processing, and a plurality of functions according to a predetermined magnification value The image processing conditions are changed, and a normal image or diagnostic image is subjected to image processing based on the changed plurality of image processing conditions, and the normal image or image processing subjected to image processing is performed. This is for causing a computer to realize a function of displaying a diagnostic image.

  According to the electronic endoscope apparatus, method, and program of the present invention, in response to an instruction to perform zoom processing, zoom processing is performed at a predetermined magnification value, and a plurality of image processing conditions are set according to the predetermined magnification value. Change and display the normal image or diagnostic image subjected to image processing based on the changed multiple image processing conditions, and display the normal image subjected to image processing or the diagnostic image subjected to image processing Thus, it is possible to provide a normal image or a diagnostic image that can easily determine whether or not the tissue subjected to the zoom process is a lesion or the like to the image interpreter.

  In addition, according to the electronic endoscope apparatus, method, and program of the present invention, the electronic endoscope apparatus, method, and program further include a database unit that stores in advance a plurality of image processing conditions corresponding to each predetermined magnification value. Whether the tissue subjected to zoom processing is a lesion or the like without performing complicated operations by the image interpreter by performing image processing based on a plurality of image processing conditions corresponding to each predetermined magnification value stored in It is possible to provide a normal image or a diagnostic image that can be easily discriminated.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 shows a basic configuration of an electronic endoscope apparatus according to an embodiment of the present invention.

  The electronic endoscope apparatus of the present invention generates a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucosa, and the normal image or the diagnostic image is In response to the switch 21 (zoom instruction reception unit) capable of receiving an instruction to perform zoom processing at a predetermined magnification value and the instruction to perform zoom processing received by the switch 21 (zoom instruction reception unit), a normal image Alternatively, a zoom mechanism unit 5 (zooming unit) that performs zoom processing with respect to a diagnostic image at a predetermined magnification value, and a plurality of image processing conditions are changed according to the predetermined magnification value, so that a normal image or a diagnostic image In contrast, an image processing circuit 27 (image processing unit) that performs image processing based on a plurality of changed image processing conditions, and a normal image that has undergone image processing or a diagnostic image that has undergone image processing are displayed. Monitor 34 (display unit) such as those provided with.

  The electronic endoscope apparatus according to the present invention further includes a database 46 (database unit) that stores a plurality of image processing conditions corresponding to each predetermined magnification value in advance, and the image processing circuit 27 is stored in the database 46. Image processing is performed based on a plurality of image processing conditions corresponding to each stored predetermined magnification value.

  As shown in the figure, the electronic endoscope apparatus includes a scope 10 inserted into a body cavity of a subject and a processor device 12 connected to the scope 10. A light source device 14 that emits white light is disposed in the processor device 12. An illumination window 22 is provided at the distal end of the scope 10, and the other end of the light guide 23 whose one end is connected to the light source device 14 faces the illumination window 22.

  The light source device 14 includes a lamp 14a that emits white light, a lighting drive circuit 14b that lights the lamp 14a, a diaphragm 14c disposed on the front side of the lamp 14a, and a diaphragm driver 14d that opens and closes the diaphragm 14c. Has been. An optical system for allowing white light emitted from the lamp 14a to enter the light guide 23 is provided between the lamp 14a and the light guide 23, but these are not shown. In addition, this type of light source device may be arranged in a separate part from the processor device 12.

  The distal end of the scope 10 is provided with a zoom mechanism 5 that is a combined imaging system that enables zoom control, and a CCD 15 that is a solid-state imaging device. As the CCD 15, for example, a primary color filter having RGB color filters on the imaging surface is attached.

  The zoom mechanism unit 5 includes a zoom lens, and the zoom lens as an optical control unit can be moved in the front-rear direction of the optical axis by the zoom drive circuit 6, and when the zoom lens moves, The magnification of the image can be changed by a predetermined magnification value.

  The predetermined magnification value is transmitted from the zoom drive circuit 6 to the image processing circuit 27 from the microcomputer 20 via the microcomputer 35.

  The zoom drive circuit 6 is composed of an actuator, and for example, a piezoelectric actuator, an ultrasonic actuator, a motor, or the like can be used.

  Connected to the CCD 15 is a CCD drive circuit 16 that forms a drive pulse based on a synchronizing signal, and also CDS / AGC (correlated double sampling / automatic) that samples and amplifies an image (video) signal output from the CCD 15. A gain control circuit 17 is connected. The CDS / AGC circuit 17 is connected to an A / D converter 18 for digitizing the analog output. Further, in the scope 10, a microcomputer 20 that controls various circuits provided therein and performs communication control with the processor device 12 is disposed.

  Further, near the base of the scope 10, there is provided a push-type switch 21 that is connected to the microcomputer 20 and switches the display mode.

  On the other hand, the processor device 12 is provided with a DSP (digital signal processor) 24 that can perform various kinds of image processing on the digitized image signal.

  The DSP 24 generates a Y / C signal composed of a luminance (Y) signal and a color difference [C (R−Y, B−Y)] signal from the R, G, and B three-color image signals output from the CCD 15. The DSP 24 is connected to a signal processing circuit (A) 25 that performs I / P conversion, noise removal, and the like.

  Connected to the signal processing circuit (A) 25 are a signal processing circuit (B) 26 for forming a normal image signal for display and a first color conversion circuit 28 for generating a spectral image signal to be described later. .

  The DSP 24 may be arranged on the scope 10 side.

  The signal processing circuit (B) 26 generates a normal image signal for display and outputs the normal image signal to the image processing circuit 27. In addition, it is possible to execute an electronic zoom process that performs an enlargement process or a reduction process on a normal image signal.

  Further, the signal processing circuit (C) 33 generates a diagnostic image signal for display and outputs the diagnostic image signal to the image processing circuit 27. In addition, it is possible to execute an electronic zoom process for performing an enlargement process or a reduction process on the diagnostic image signal.

  The image processing circuit 27 is connected to a monitor 34 made up of, for example, a liquid crystal display device or a CRT, and an image recording device 45.

  In the image processing circuit 27, when the normal image display mode is selected, the normal image signal output from the signal processing circuit (B) 26 is output to the monitor 34 and the image recording device 45, and the diagnostic image display mode is output. Is selected, a normal image signal output from the signal processing circuit (B) 26 and a diagnostic image output from the signal processing circuit (C) 33 are generated in the normal image signal. The data is output to the monitor 34 and the image recording device 45.

The image processing circuit 27 performs various image processing such as brightness enhancement processing, structure enhancement processing, color gradation processing, mirror image processing, mask generation, character generation, and color adjustment on the normal image or the diagnostic image. Hereinafter, the generation of the normal image and the diagnostic image will be mainly described.

  The first color conversion circuit 28 converts the Y / C signal output from the signal processing circuit (A) 25 into a three-color image signal of R, G, and B. On the subsequent stage side of the first color conversion circuit 28, corresponding to the selected wavelength regions λ1, λ2, and λ3 read out from the matrix (spectral) data for generating the estimated spectral image signal stored in advance. The first color that performs matrix operation on the three-color image signals R, G, and B using the parameters to output estimated spectral image signals λ1s, λ2s, and λ3s for the selected wavelengths λ1, λ2, and λ3. The spatial conversion processing circuit 29 and the estimated spectral image signals λ1s, λ2s, and λ3s are amplified using the gain value for each image signal input by the user, and pseudo color spectral image signals λ1t, λ2t, and λ3t are output. The average luminance of all the signals of the second color space conversion processing circuit 30 and the pseudo color spectral image signals λ1t, λ2t, and λ3t is the Y / C signal output from the DSP 24, that is, all the signals of the normal image signal. average The luminance adjustment circuit that adjusts the luminance of the pseudo color spectral image signals λ1t, λ2t, and λ3t so as to be substantially equal to the degree, and outputs the pseudo color spectral image signals λ1t ′, λ2t ′, and λ3t ′ that have undergone the luminance adjustment. 31 and the pseudo color spectral image signals λ1t ′, λ2t ′, and λ3t ′ are input to the R, G, and B channels for processing corresponding to the RGB signals, respectively, and the input signals are converted into Y / C signals. A second color conversion circuit 32, a signal processing circuit (C) 33, and an image processing circuit 27 for conversion are sequentially connected in this order.

  In the processor device 12, communication with the scope 10 is performed, each circuit in the device 12 is controlled, and matrix (spectral) data for forming an estimated spectral image signal is stored in the color space. A microcomputer 35 having a function of inputting to the conversion processing circuit 29 is provided. The memory 36 stores matrix data for forming an estimated spectral image signal based on a wavelength set, a gain set, and an RGB signal for forming a diagnostic image.

  As wavelength sets of λ1, λ2, and λ3, for example, as shown in FIG. 2, wavelength range standard set a of 400, 500, and 600 (nm, the same applies hereinafter), and blood vessels of 470, 500, and 670 for depicting blood vessels. B1 set b, blood vessel B2 set c of 475, 510, 685 also for depicting blood vessels, tissue E1 set d of 440, 480, 520 for depicting specific tissues, and 480 for depicting specific tissues as well 510,580 tissue E2 set b, 400,430,475 hemoglobin set f for depicting the difference between oxyhemoglobin and deoxyhemoglobin, 415,450,500 blood for depicting the difference between blood and carotene -Carotene set g, 8 of 420, 550, 600 blood-cytoplasm set h to depict the difference between blood and cytoplasm Long set is stored. Further, gain standard sets of 1, 1, 1 are stored as gain sets of e1, e2, e3.

Matrix data is stored as a table. In the present embodiment, an example of matrix data stored in the memory 36 is as shown in Table 1 below.

The matrix data in Table 1 includes 61 wavelength range parameters (coefficient sets) p1 to p61 obtained by dividing a wavelength range of 400 nm to 700 nm at 5 nm intervals, and parameters P1 to P3 for normal image formation. Each of the parameters p1 to p61 includes coefficients k _pr , k _pg , and k _pb (p = 1 to 61) for matrix calculation.

Then, in the first color space conversion processing circuit 29, a matrix operation represented by the following equation is performed by the coefficients k _pr , k _pg , k _pb and the RGB signal output from the first color conversion circuit 28, and the estimated spectral image Signals λ1s, λ2s, and λ3s are formed.

That is, when, for example, 500 nm, 620 nm, and 650 nm are selected as the wavelength ranges λ1, λ2, and λ3 constituting the spectroscopic image, the coefficients (k _pr , k _pg , k _pb ) are selected from among the 61 parameters in Table 1. , The coefficient of the parameter p21 corresponding to the central wavelength of 500 nm (−0.00119, 0.002346, 0.0016), the coefficient of the parameter p45 corresponding to the central wavelength of 620 nm (0.004022, 0.000068, −0.00097), and the coefficient of the parameter p51 corresponding to the central wavelength of 650 nm The matrix calculation is performed using (0.005152, -0.00192, 0.000088).

  In addition to the memory 36, a monitor 34, a keyboard-type input unit 41, an image recording controller 42, and the scope 20 microcomputer 20 are connected to the microcomputer 35.

  FIG. 3 is a diagram illustrating an example of a screen 47 displayed on the monitor 34 when the diagnostic image display mode is selected. On the screen 47 of the monitor 34, a normal image 48, a diagnostic image 49, a wavelength region display small screen 51 for displaying, selecting or setting a wavelength region for forming the diagnostic image, and a gain of each wavelength region are displayed. Then, a small gain display screen 52 for selecting or setting is displayed.

  The wavelength range display small screen 51 displays, for example, a set selection switch 53 for selecting a wavelength set of a to h, and wavelength display switches for displaying the wavelength ranges λ1, λ2, and λ3 and setting them by manual input. 54a to 54c are displayed. The gain display small screen 52 displays a set selection switch 55 for selecting a standard gain set (1, 1, 1) and gains e1, e2, e3, and is used for manual input setting. Gain display switches 56a to 56c are displayed. The switches displayed on these small screens are operated by a keyboard-type input unit 41.

  The switch 21 includes an up / down lock lever for locking the up / down bending state of the scope 10, an up / down angle knob for bending the tip of the scope 10 up / down, a left / right angle knob for bending the tip of the scope 10 left / right, and the scope 10. Left / right lock knob for locking the left / right curved state, an RC switch for transmitting an image to the image recording device 45 connected to the electronic endoscope of the present embodiment, an FR switch for capturing a normal image or a diagnostic image, It has an MM switch that enables zoom processing and allows selection of a normal image display mode and a diagnostic image display mode, a suction button that performs suction, an air / water supply button that performs air / water supply, and the like.

  Hereinafter, the operation of the electronic endoscope apparatus of the present embodiment having the above-described configuration will be described.

  First, the operation in the normal image display mode will be described. In the normal image display mode, a normal color image is formed. When these images are formed, the light source device 14 shown in FIG. 1 is driven, and white light emitted therefrom enters the light guide 23 through the stop 14c, and the light guide 23 arranged in the scope 10 is used. The object to be observed is irradiated with white light emitted from the tip. Then, the CCD 15 driven by the CCD driving circuit 16 images the object to be observed and outputs an imaging signal. This image signal is amplified by correlated double sampling and automatic gain control in the CDS / AGC circuit 17 and then A / D converted by the A / D converter 18 and input to the DSP 24 of the processor unit 12 as an RGB image signal. Is done. In the DSP 24, color conversion processing is performed on the RGB image signal that is the three-color image signal output from the scope 10, and the above-described Y / C signal, that is, the normal image signal is formed. The Y / C signal (normal image signal) output from the DSP 24 is subjected to I / P conversion and noise removal in the signal processing circuit (A) 25, and is input to the signal processing circuit (B) 26 for normal display. An image signal is generated (in some cases, enlargement processing or reduction processing using electronic zoom is performed) and output to the image processing circuit 27. In the image processing circuit 27, brightness enhancement processing, structure enhancement processing, color gradation processing, mirror image processing, mask generation, character generation, color adjustment, etc. are performed on the normal image signal output from the signal processing circuit (B) 26. Various signal processings are performed and output to the monitor 34 and the image recording device 45.

  Next, the operation in the diagnostic image display mode will be described. When the apparatus is operating in the normal image display mode, when the user presses a switch 21 (MM switch described later) provided in the scope 10, the operation mode changes from the normal image display mode to the normal image and the diagnostic image. Is switched to the diagnostic image display mode for displaying both images.

  When the apparatus is operating in the diagnostic image display mode, when the user presses a switch 21 (MM switch described later) provided in the scope 10, the operation mode is changed from the diagnostic image display mode to the normal image. Switch to normal image display mode.

  The diagnostic image display mode may display only the diagnostic image.

  In the diagnostic image display mode, the spectral color image forming operation is performed in parallel with the normal color image forming operation described above. Hereinafter, spectral color image formation will be described. As described above, the Y / C signal (normal image signal) output from the DSP 24 is input to the signal processing circuit (B) 26 via the signal processing circuit (A) 25 to form a normal color image. . At the same time, the Y / C signal output from the DSP 24 is input to the first color conversion circuit 28 via the signal processing circuit (A) 25, where it is converted into an RGB signal. The RGB signals are supplied to the first color space conversion processing circuit 29, and the first color space conversion processing circuit 29 performs matrix calculation for forming an estimated spectral image based on the RGB signals and matrix data.

  Hereinafter, this calculation will be described. The first color space conversion processing circuit 29 uses the matrix data stored in the memory 36 described above to perform the matrix operation of the above equation 1 for forming a diagnostic image on the RGB signals. Three wavelength ranges of λ1, λ2, and λ3 are set by operating the input unit 41, and the microcomputer 35 sets parameters corresponding to these three selected wavelength ranges from the memory 36 from the matrix data stored in the memory 36. These are read out and input to the first color space conversion processing circuit 29.

For example, when the wavelengths 500 nm, 620 nm, and 650 nm are selected as the three wavelength ranges λ1, λ2, and λ3, the coefficients of the parameters p21, p45, and p51 of Table 1 corresponding to the respective wavelengths are used, and the RGB signals are used. Estimated spectral image signals λ1s, λ2s, and λ3s are formed by the matrix calculation of the following equation (2).

  Thereafter, the estimated spectral image signals λ1s, λ2s, and λ3s are supplied to the second color space conversion processing circuit 30. The second color space conversion processing circuit 30 uses the estimated spectral image signals λ1s, λ2s, and λ3s and the matrix indicating the gain values of the estimated spectral image signals (λ1s, λ2s, and λ3s) to form a pseudo color spectral image. Matrix calculation is performed.

  Hereinafter, this calculation will be described. A gain value for each estimated spectral image signal (λ1s, λ2s, λ3s) is set by operating the input unit 43. The microcomputer 35 generates a 1 × 3 matrix corresponding to these three gain values and outputs it to the second color space conversion processing circuit 30.

For example, when e1, e2, and e3 are selected as gain values for the three wavelength regions λ1, λ2, and λ3, pseudo color spectroscopy is performed from the estimated spectral image signal (λ1s, λ2s, λ3s) by the following matrix calculation. Image signals λ1t, λ2t, and λ3t are formed.

  Thereafter, the pseudo color spectral image signals λ1t, λ2t, and λ3t are supplied to the luminance adjustment circuit 31. In the luminance adjustment circuit 31, the average luminance of all the pseudo color spectral image signals λ1t, λ2t, and λ3t is substantially equal to the Y / C signal output from the DSP 24, that is, the average luminance of all the normal image signals. As described above, the luminance adjustment for adjusting the luminance of the pseudo color spectral image signals λ1t, λ2t, and λ3t is performed.

Hereinafter, this brightness adjustment will be described. The microcomputer 35 first calculates the Y / C signal related to one field output from the DSP 24, that is, the average value (normal Yav) of the Y (luminance) signal of the normal image signal for one field. In addition, when the pseudo color spectral image signals λ1t, λ2t, and λ3t for one field are generated, all the pseudo color spectral image signals for one field are converted to Y / C signal Y (luminance). ) Calculate the signal value by the following equation.

The microcomputer 35 calculates an average value (spectral Yav) of all Y (luminance) signal values of the pseudo color spectral image signal for one field, and obtains a ratio with the normal Yav. When normal Yav = α · spectral Yav, pseudo color spectral image signals λ1t ′, λ2t ′, and λ3t ′ that have undergone luminance adjustment are calculated by the following calculation.

  Thereafter, the pseudo-color spectral image signals λ1t ′, λ2t ′, and λ3t ′ are input to the second color conversion circuit 32 as three-color image signals of Rs, Gs, and Bs, respectively. In the second color conversion circuit 32, Rs, Gs , Bs three-color image signals are converted into Y / C signals (Y, Rs-Y, Bs-Y), and the Y / C signals, that is, spectral image signals are subjected to signal processing by the signal processing circuit (C) 33. And input to the image processing circuit 27. In the image processing circuit 27, the normal image signal output from the signal processing circuit (B) 26 and the spectral image signal output from the signal processing circuit (C) 33 are generated and output to the monitor 34 and the image recording device 45. .

  The diagnostic image displayed on the monitor 34 based on the spectral image signal is composed of color components in the wavelength range as shown in FIGS. That is, FIG. 4 is a conceptual diagram in which three wavelength regions λ1, λ2, and λ3 for forming a diagnostic image are superimposed on the spectral sensitivity characteristics R, G, and B of the color filter of the primary color CCD 15, and FIG. It is the conceptual diagram which piled up three wavelength range (lambda) 1, (lambda) 2, and (lambda) 3 on the reflection spectrum of the biological body. Spectral image signals λ1s, λ2s, and λ3s based on the parameters p21, p45, and p51 exemplified above are color signals in a wavelength range of about ± 10 nm with 500 nm, 620 nm, and 650 nm as center wavelengths as shown in FIG. Thus, a diagnostic image (moving image or still image) composed of a combination of colors in these three wavelength ranges is displayed.

  Next, display, selection, and setting of the wavelength ranges λ1, λ2, λ3 and gains e1, e2, e3 will be described. After the electronic endoscope apparatus is shipped from the factory, when the apparatus is first turned on and the apparatus is started, the above-mentioned wavelength range standard set a (400, 500, 600) and gain standard set (1, 1, 1) are set by the microcomputer 35. Selected. When the diagnostic image display mode is selected, the selected wavelength range (400, 500, 600) is displayed by the wavelength display switches 54a to 54c, and the gain standard set (1, 1, 1). Is displayed on the gain display switches 56a to 56c. The first color space conversion processing circuit 29 performs the above-described matrix calculation for the wavelength range (400, 500, 600) to form the amplification circuit estimated spectral image signals λ1s, λ2s, λ3s. Further, the second color space conversion processing circuit 30 performs the above-described calculation using the gain (1, 1, 1), and forms the pseudo color spectral image signals λ1t, λ2t, λ3t.

  Further, an interpreter such as a clinician can select the other wavelength sets b to h in order and arbitrarily by clicking the set selection switch 53 using the keyboard-type input unit 41. Further, the wavelength range can be set to an arbitrary value by moving the positions of the wavelength display switches 54a to 54c to the left and right by operating the keyboard-type input unit 41. Similarly, the gain can be set to an arbitrary value by moving the positions of the gain display switches 56a to 56c to the left and right. The parameters corresponding to the wavelength ranges λ1, λ2, and λ3 of the selected wavelength set are read from the memory 36 by the microcomputer 35, and these parameters are input to the first color space conversion processing circuit 29. The first color space conversion processing circuit 29 performs the above-described matrix operation using the input parameters to form estimated spectral image signals λ1s, λ2s, and λ3s. The selected gain set is input to the second color space conversion processing circuit 30. The second color space conversion processing circuit 30 performs the above-described calculation using the input parameters to form pseudo color spectral image signals λ1t, λ2t, and λ3t. The luminance adjustment unit 31 performs the above-described calculation on the pseudo color spectral image signals λ1t, λ2t, and λ3t, and calculates pseudo color spectral image signals λ1t ′, λ2t ′, and λ3t ′ that have undergone luminance adjustment.

  As the wavelength set, in addition to the wavelength set as described above, another set is prepared according to the request of the operator of the apparatus, and the set is stored in the memory 36 so that it can be appropriately used. Good.

  In the present embodiment, the output of the image processing circuit 27 is also input to an image recording device 45 (for example, a printer) in addition to the monitor 34, and an image recording controller controlled by the microcomputer 35. When 42 gives an image recording instruction to the image recording device 45, a hard copy of a normal image or a diagnostic image of the scene designated by the instruction is output from the image recording device 45.

  Hereinafter, a method for performing image processing on a normal image of the present invention when the normal image display mode is set will be described with reference to the flowchart shown in FIG.

  First, the zoom drive circuit 6 receives a magnification value received by a switch 21 (MM switch) capable of receiving an instruction to perform zoom processing (for example, a magnification value indicating how much to enlarge, a reduction value) If so, a magnification value indicating the degree of reduction) is received (ST1).

  Next, the zoom drive circuit 6 controls the zoom mechanism 5 so as to perform zoom processing according to the received magnification value. Thereby, the zoom mechanism unit 5 performs zoom processing (ST2). Thereafter, the signal processing circuit (B) 26 transmits the zoomed normal image and the image processing circuit 27.

  The magnification value is transmitted from the zoom drive circuit 6 to the image processing circuit 27 from the microcomputer 20 via the microcomputer 35.

  When the magnification value is lower than a predetermined value (threshold value) (ST3; YES), the image processing circuit 27 performs image processing on the normal image based on the image processing condition A recorded in the database 46 ( ST4).

  When the magnification value is higher than a predetermined value (threshold value) (ST3; NO), the image processing circuit 27 performs image processing on the normal image based on the image processing condition B recorded in the database 46 ( ST5).

  The image processing based on the image processing conditions performed by the image processing circuit 27 may be performed by the signal processing circuit (B) 26.

  Here, the image processing condition A and the image processing condition B will be described.

  The image processing circuit 27 sets the brightness (brightness change) correction amount strongly as the image processing condition A as shown in FIG. The contrast (color gradation enhancement) correction amount is set to normal (default value), and the sharpness (structure enhancement) correction amount is set to normal (default value).

  Further, when the magnification value to be reduced is higher than a predetermined value, the image processing circuit 27 sets the brightness (brightness change) correction amount to normal (default value) as the image processing condition A, and contrast (color gradation emphasis). ) The correction amount may be set strongly, and the sharpness (structure enhancement) correction amount may be set strongly.

  On the other hand, when the magnification value to be enlarged is higher than a predetermined value, the image processing circuit 27 sets the brightness (brightness change) correction amount to normal (default value) as the image processing condition B as shown in FIG. The contrast (color gradation enhancement) correction amount is set strongly, and the sharpness (structure enhancement) correction amount is set strongly.

  Further, the image processing circuit 27 sets the brightness (brightness change) correction amount strongly as the image processing condition B when the magnification value to be reduced is low or the magnification to be reduced is equal, and the contrast (color gradation) The enhancement amount may be set to normal (default value), and the sharpness (structure enhancement) correction amount may be set to normal (default value).

  The monitor 34 (display unit) displays a normal image subjected to image processing based on either the image processing condition A or the image processing condition B (ST6).

  The image processing condition A and the image processing condition B may have display settings for changing from the current display mode to either the normal image display mode or the diagnostic image mode. When the magnification value to be enlarged is higher than the predetermined value, the normal image display mode can be automatically changed to the diagnostic display mode. Conversely, the display mode may be automatically changed.

  Further, when the magnification value to be reduced from the predetermined value is high, it is possible to automatically change from the normal image display mode to the diagnostic display mode. Conversely, the display mode may be automatically changed.

  Hereinafter, a method for performing image processing on the diagnostic image of the present invention when the diagnostic image display mode is set will be described with reference to the flowchart shown in FIG.

  First, the zoom drive circuit 6 receives a magnification value received by a switch 21 (MM switch) capable of receiving an instruction to perform zoom processing (for example, a magnification value indicating how much to enlarge, a reduction value) If so, a magnification value indicating how much the image is reduced is received (ST11).

  Next, the zoom drive circuit 6 controls the zoom mechanism 5 so as to perform zoom processing according to the received magnification value. Thereby, the zoom mechanism unit 5 performs zoom processing (ST12). Thereafter, the signal processing circuit (C) 33 transmits the diagnostic image subjected to the zoom process to the image processing circuit 27. The magnification value is transmitted from the zoom drive circuit 6 to the image processing circuit 27 from the microcomputer 20 via the microcomputer 35.

If the magnification value is lower than a predetermined value (threshold value) (ST13; YES), the image processing circuit 27 performs image processing on the diagnostic image based on the image processing condition A recorded in the database 46. (ST14).

  When the magnification value is higher than a predetermined value (threshold value) (ST13; NO), the image processing circuit 27 performs image processing on the diagnostic image based on the image processing condition B recorded in the database 46. (ST15).

  The image processing based on the image processing conditions performed by the image processing circuit 27 may be performed by the signal processing circuit (C) 33.

  When the magnification value is lower than the threshold value, the purpose is to distinguish between a normal part of the biological mucosa and an abnormal part of the biological mucous membrane (for example, cancer, gastric ulcer, etc.), and thus highlights bright, fine color change to a distant view. It is preferable to set such image processing conditions. On the other hand, if the magnification value is higher than the threshold value, it is necessary to observe the detailed state of the part suspected of being abnormal, and image processing conditions that emphasize fine blood vessels, reduce redness of the mucous membrane, are preferred. Therefore, it is preferable to set the image processing conditions as follows.

  For example, when the magnification value to be magnified is lower than a predetermined value (the magnification includes the same magnification), the image processing circuit 27 sets the RGB value, R is 520 nm, G as the image processing condition A as shown in FIG. Is set to display at 500 nm and B at 405 nm. Further, in addition to the above display setting, the image processing condition A is set with a strong brightness (brightness enhancement processing) correction amount and a contrast (color gradation processing) correction amount as normal (default value) as shown in FIG. And set the sharpness (structure enhancement processing) correction amount to normal (default value).

  Further, when the magnification value to be reduced is lower than a predetermined value (the magnification includes the same magnification), the image processing circuit 27 sets the RGB value as the image processing condition A, R is 525 nm, G is 495 nm, and B is 495 nm. Set the display so that Furthermore, in addition to the above display setting, the image processing condition B sets the brightness (lightness enhancement processing) correction amount to normal (default value) and sets the contrast (color gradation processing) correction amount as shown in FIG. The sharpness (structure enhancement processing) correction amount may be set strongly.

  On the other hand, when the magnification value to be magnified is higher than a predetermined value, the image processing circuit 27 sets the RGB value as the image processing condition B, as shown in FIG. 9, R is 525 nm, G is 495 nm, and B is 495 nm. Set the display as follows. Furthermore, in addition to the above display setting, the image processing condition B sets the brightness (brightness enhancement processing) correction amount to normal (default value), sets the contrast (color gradation processing) correction amount strongly, and sharpness (structure) Emphasis processing) Set the correction amount strongly.

  Further, when the magnification value to be reduced is higher than a predetermined value, the image processing circuit 27 sets the RGB value as the image processing condition B so that R is 520 nm, G is 500 nm, and B is 405 nm. Further, in addition to the above display setting, the image processing condition A is set with a strong brightness (brightness enhancement processing) correction amount, a contrast (color gradation processing) correction amount set to normal (default value), and sharpness (structure Emphasis processing) The correction amount may be set to normal (default value).

  The monitor 34 (display unit) displays the diagnostic image that has been subjected to the image processing based on either the image processing condition A or the image processing condition B (ST16).

  The image processing condition A and the image processing condition B may have display settings for changing from the current display mode to either the normal image display mode or the diagnostic image mode. When the magnification value to be enlarged is higher than the predetermined value, the normal image display mode can be automatically changed to the diagnostic display mode. Conversely, the display mode may be automatically changed.

  Further, when the magnification value to be reduced from the predetermined value is high, it is possible to automatically change from the normal image display mode to the diagnostic display mode. Conversely, the display mode may be automatically changed.

  Note that the input unit 41 may change the threshold value by an input of an image interpreter using a keyboard or the like. For example, when the threshold value is changed to 20 times, the image processing condition A is applied to the normal image or the diagnostic image when the magnification value is equal to 20 times. On the other hand, when the magnification value is 20 times or more, the image processing condition B is applied to the normal image or the diagnostic image.

  The database 46 may record a plurality of different image processing conditions in addition to the image processing conditions A and B.

  In addition to the image processing conditions A and B, different image processing conditions can be sequentially switched by pressing the switch 21, the foot switch, or the front panel switch having a touch panel function on the monitor 34. Thereby, the monitor 34 (display unit) displays the normal image or the diagnostic image subjected to the image processing based on either the image processing condition A or the image processing condition B (for example, ST6 and ST16), and then switches. It is possible to display a normal image or a diagnostic image that has been subjected to image processing according to the image processing conditions.

  Further, by applying a part recognition technique, the image processing conditions may be switched for each part, and a normal image or a diagnostic image subjected to image processing according to the switched image processing conditions may be displayed.

  As this part recognition technology, for example, a method using a discriminator obtained by machine learning using AdaBoost, support vector machine (SVM), adaptive vector machine (RVM), artificial neural network (ANN), etc. Template matching, comparison processing with a unique image, and the like.

  The database 46 may store a plurality of image processing conditions corresponding to each predetermined magnification value in advance. The image processing circuit 27 does not determine the image processing condition based on the threshold as described above, but receives a magnification value and determines a plurality of image processing conditions stored in the database 46 corresponding to the magnification value. Is possible.

  As is clear from the above description, in the present embodiment, in response to an instruction to perform zoom processing, the normal image or the diagnostic image is subjected to zoom processing at a predetermined magnification value, and the predetermined magnification value is obtained. The image processing conditions are changed in accordance with the image processing, the normal image or the diagnostic image is subjected to the image processing based on the changed image processing conditions, and the processed normal image or the image processing is performed. By displaying the diagnostic image thus obtained, a normal image or a diagnostic image that makes it easy to determine whether or not the target tissue on the zoomed image is a lesion or the like is displayed to the reader Can do.

  Further, in the present embodiment, a database 46 that previously stores a plurality of image processing conditions corresponding to each predetermined magnification value is further provided, and the image processing circuit 27 is provided for each predetermined magnification value stored in the database 46. By performing image processing based on a plurality of image processing conditions corresponding to the above, it is possible to determine whether or not the tissue subjected to zoom processing is a lesion or the like without performing complicated operations for the reader. It is possible to provide a normal image or a diagnostic image that can be accurately discriminated.

  Furthermore, in this embodiment, the interpreter can freely set the color balance of the diagnostic image, and the brightness of the diagnostic image can be automatically aligned with the brightness of the normal image. Even when the image and the diagnostic image are displayed on the monitor 34 at the same time, the sense of discomfort of the interpreter is reduced.

  In the above embodiment, the luminance adjustment is performed using the average value of the Y signal of the normal color image or diagnostic image for one field. However, the luminance value is not limited to such an average value. Further, the luminance adjustment may be performed based on the average value or maximum value of the Y signal related to a specific partial region in one field.

  In the present embodiment, the wavelength range from 400 nm to 700 nm can be divided into 61 wavelength ranges and can be selected. However, as the wavelength ranges λ1, λ2, and λ3, the wavelength range including the infrared range or the red range can be selected. By selecting a wavelength set for only the outer region, it is possible to obtain a diagnostic image that approximates an image obtained by irradiating infrared rays in the past without using a cut filter in the visible light region. Moreover, in the conventional endoscope, the fluorescence emitted from the cancer tissue or the like by the irradiation of the excitation light is photographed. By selecting the wavelength set of λ1, λ2, λ3 according to the fluorescence wavelength, In this case, a diagnostic image of a fluorescent part can be formed. In this case, there is an advantage that an excitation light cut filter is not required.

  Furthermore, in conventional endoscopes, pigments such as indigo and pioctanin are sprayed on the object to be observed, and images of tissues colored by pigment spraying are performed. As a wavelength set of the above λ1, λ2, and λ3 By selecting a wavelength range in which a tissue colored by pigment dispersion can be drawn, a diagnostic image equivalent to the image at the time of pigment dispersion can be obtained without performing pigment dispersion.

  In the present embodiment, when the pseudo color spectral image signals λ1t ′, λ2t ′, and λ3t ′ are input to the second color conversion circuit 32 as three-color image signals of Rs, Gs, and Bs, respectively, The image signals λ1t ′, λ2t ′, and λ3t ′ are assigned to the Rs, Gs, and Bs three-color image signals in that order. However, when the user desires a special color display, the order may be changed and assigned. Good. In such a case, for example, a diagnostic image in which a blood vessel portion is displayed in yellow or blue is displayed.

  In the present embodiment, by irradiating light with two narrow-band wavelengths (for example, 390 to 450 nm / 530 to 550 nm) that are easily absorbed by hemoglobin in blood, the capillaries and mucous membranes on the mucosal surface layer are irradiated. You may apply the diagnostic image obtained by the method (for example, Narrow Band Imaging-NBl) which emphasizes a fine pattern.

  In the present embodiment, the zoom process is described as being performed by optical zoom. However, in response to an instruction to perform electronic zoom received by the switch 21 (zoom instruction receiving unit), the zoom process is normally performed at a predetermined magnification value. By performing electronic zoom on the image or diagnostic image by the signal processing circuit (B) 26 (zooming unit) or the signal processing circuit (C) 33 (zooming unit), the image processing circuit 27 has a predetermined magnification value. A plurality of image processing conditions are changed according to the image processing, the normal image or the diagnostic image is subjected to image processing based on the changed plurality of image processing conditions, and the monitor 34 A diagnostic image that has undergone image processing may be displayed.

1 is a block diagram showing a configuration of an electronic endoscope apparatus according to an embodiment of the present invention. Figure showing an example of the spectral image wavelength range Schematic diagram of the image displayed on the monitor A graph showing an example of the wavelength range of a spectral image together with the spectral sensitivity characteristics of a primary color CCD A graph showing an example of the wavelength range of a spectral image along with the reflection spectrum of a living body The flowchart which shows the flow of the process using the normal image of this invention Image processing conditions for normal images The flowchart which shows the flow of the process using the diagnostic image of this invention Image processing condition 1 for diagnostic images Image processing condition 2 for diagnostic images

Explanation of symbols

DESCRIPTION OF SYMBOLS 5 Lens mechanism part 6 Zoom drive circuit 10 Scope part 12 Processor part 14 Light source device 14a Lamp 14b Lighting drive circuit 14c Aperture 14d Aperture drive part 15 CCD
17 CDS / AGC circuit 20, 35 Microcomputer 21 Switch 22 Lighting window 23 Light guide 24 DSP
25 Signal processing circuit (A)
26 Signal processing circuit (B)
27 Image processing circuit 28 First color conversion circuit 29 First color space conversion processing circuit 30 Second color space conversion processing circuit 31 Luminance adjustment circuit 32 Second color conversion circuit 33 Signal processing circuit (C)
34 Monitor 41 Input section 46 Database 47 Screen 51 Wavelength range display small screen 52 Gain display small screen

Claims (6)

An electronic endoscope apparatus that generates a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucous membrane,
A zoom instruction receiving unit capable of receiving an instruction to perform zoom processing at a predetermined magnification value on the normal image or the diagnostic image;
In response to an instruction to perform zoom processing received by the zoom instruction receiving unit, a zooming unit that performs zoom processing at the predetermined magnification value;
An image processing unit that changes a plurality of image processing conditions according to the predetermined magnification value, and performs image processing on the normal image or the diagnostic image based on the changed plurality of image processing conditions;
An electronic endoscope apparatus comprising: a display unit that displays a normal image subjected to the image processing or a diagnostic image subjected to the image processing. A database unit that stores in advance the plurality of image processing conditions corresponding to each predetermined magnification value;
2. The electronic endoscope according to claim 1, wherein the image processing unit performs image processing based on the plurality of image processing conditions corresponding to each of the predetermined magnification values stored in the database unit. Mirror device.   3. The electronic endoscope according to claim 1, wherein the plurality of image processing conditions relate to a correction amount of at least two of brightness enhancement processing, structure enhancement processing, and color gradation processing. apparatus.   The electronic endoscope apparatus according to claim 3, wherein the plurality of image processing conditions further include display setting of one of the normal image and the diagnostic image. An electronic endoscope display method for generating a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucous membrane,
Accepting an instruction to perform zoom processing at a predetermined magnification value for the normal image or the diagnostic image;
In response to the received instruction to perform zoom processing, perform zoom processing at the predetermined magnification value,
A plurality of image processing conditions are changed according to the predetermined magnification value, and image processing is performed on the normal image or the diagnostic image based on the changed plurality of image processing conditions,
An electronic endoscope display method comprising: displaying a normal image subjected to the image processing or a diagnostic image subjected to the image processing. An electronic endoscope display program that generates a normal image representing a biological mucous membrane and a diagnostic image that is a narrow-band spectral image representing the biological mucous membrane,
A function of receiving an instruction to perform zoom processing at a predetermined magnification value on the normal image or the diagnostic image;
A function of performing zoom processing at the predetermined magnification value in response to the received instruction to perform zoom processing;
A function of changing a plurality of image processing conditions according to the predetermined magnification value, and performing image processing on the normal image or the diagnostic image based on the changed plurality of image processing conditions;
An electronic endoscope display program for causing a computer to realize a function of displaying a normal image subjected to the image processing or a diagnostic image subjected to the image processing.

Images