Endoscopic system and its operating method

Abstract

Provided is an endoscope system (10) configured such that in a case in which a plurality of types of illumination light are to be emitted and a plurality of types of image signals are to be obtained, a highly reliable image signal can be obtained for each type of image signal to be acquired. Also provided is a method for operating the endoscope system.　A processor acquires a plurality of types of mutually different image signals in a first exposure period in which illumination light for observation is emitted and in a second exposure period in which illumination light for correction is emitted. The processor then calculates image signal brightness from some of the mutually different image signals among the image signals, outputs a control amount that varies according to the image signal brightness, generates mutually different exposure control signals on the basis of the image signals, and controls a light source and an imaging optical system according to the exposure control signals.

Description

Technical Field

[0005]

[0001] The present invention relates to an endoscope system that performs exposure control when obtaining a plurality of types of image signals, and an operating method thereof.

Background Art

[0002] In the medical field where endoscopes are used, oxygen saturation imaging is known as one of the techniques for facilitating the detection of lesions. In oxygen saturation imaging, narrow-band light is irradiated onto an observation target, and the oxygen saturation is calculated from an endoscope image by utilizing the difference in the absorption coefficients of oxidized hemoglobin and reduced hemoglobin in the tissue of the observation target, and lesions such as malignant tumors with relatively low oxygen saturation are detected. Although specific pigments such as yellow pigments different from oxidized hemoglobin and reduced hemoglobin exist in the tissue, there is a problem that the calculation accuracy of oxygen saturation decreases due to the influence of the light absorption of these specific pigments. To solve this problem, correction illumination light for calculating the influence of specific pigments is irradiated, and the algorithm for oxygen saturation calculation is corrected based on the signal obtained by correction imaging (see Patent Document 1).

[0003] Also, when sequentially emitting a plurality of types of illumination light onto an observation target, the target light amount is calculated using the brightness information of an endoscope image captured using the narrow-band light used for calculating oxygen saturation, and light amount control for correcting the brightness of a plurality of types of endoscope image signals acquired in subsequent frames in a time series is known (see Patent Document 2).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] The optimal brightness of an image signal differs depending on whether oxygen saturation is being calculated or correction calculations are being performed for oxygen saturation. To ensure the accuracy of oxygen saturation calculation and correction, the signal values, including brightness, of the image signal must be appropriate. Therefore, adjusting the brightness of one type of image signal using another type of image signal may result in problems such as the image being too dark or too bright, depending on the type of image signal being acquired. Consequently, when acquiring multiple types of image signals, there is a need for a technology that can obtain a new image signal with a more reliable signal value that is more suitable for the purpose of acquiring each image signal.

[0006] The present invention aims to provide an endoscope system and its operating method that enable the acquisition of highly reliable image signals for each type of image signal obtained when emitting multiple types of illumination light to acquire multiple types of image signals. [Means for solving the problem]

[0007] The endoscope system of the present invention comprises a light source device, an imaging optical system, and a processor, and illuminates an object to be observed and captures reflected light from the object to be observed. The light source device emits at least one light source and emits observation illumination light and correction illumination light having a different spectrum from the observation illumination light to the object to be observed. The imaging optical system captures the reflected light. The processor acquires multiple types of image signals that are different from each other during a first exposure period in which observation illumination light is emitted and a second exposure period in which correction illumination light is emitted, calculates the image signal brightness from some of the image signals that are different from each other, outputs a control amount that changes according to the image signal brightness, generates multiple types of exposure control signals that have different control amounts, and controls the light source and the imaging optical system according to the exposure control signals.

[0008] Preferably, the control variable is output based on the difference between the image signal brightness calculated from the image signal and a preset target brightness.

[0009] Preferably, the processor acquires an observation image signal as an image signal during the first exposure period, acquires a correction image signal as an image signal during the second exposure period, outputs a first control variable as a control variable based on the difference between a first image signal brightness, which is the image signal brightness calculated using the observation image signal, and a first target brightness, which is the target brightness, outputs a second control variable as a control variable different from the first control variable based on the difference between a second image signal brightness, which is the image signal brightness calculated using the correction image signal, and a second target brightness, which is the image signal brightness, generates a first exposure control signal as an exposure control signal based on the first control variable, and generates a second exposure control signal as an exposure control signal based on the second control variable. Preferably, the second control variable is larger than the first control variable.

[0010] The processor preferably outputs a control variable using a specific color signal from the image signal. The specific color signal is preferably the B signal.

[0011] Preferably, the processor generates an analysis image using the image signal, extracts a specific region from the analysis image, and outputs a control variable using the image signal in the specific region. Preferably, the specific region is extracted according to a specific shape.

[0012] Preferably, the processor extracts a first region as a specific region from a first analysis image, which is an analysis image generated using an observation image signal, outputs a first region control amount as a control amount based on the difference between the first region brightness, which is an image signal brightness calculated using the first region, and the first target brightness; extracts a second region as a specific region from a second analysis image, which is an analysis image generated using a correction image signal, outputs a second region control amount as a control amount based on the difference between the second region brightness, which is an image signal brightness calculated using the second region, and the second target brightness; generates a first region exposure control signal as an exposure control signal based on the first region control amount; and generates a second region exposure control signal as an exposure control signal based on the second region control amount.

[0013] Preferably, the processor switches between an observation mode for acquiring an observation image signal and a correction mode for acquiring both an observation image signal and a correction image signal, generates a first exposure control signal in observation mode, and generates a first exposure control signal and a second exposure control signal in correction mode.

[0014] The first exposure period includes a first illumination period in which white equivalent light included in the observation illumination light is emitted, and a second illumination period in which calculation illumination light included in the observation illumination light is emitted. The second exposure period includes a third illumination period in which correction illumination light is acquired. Preferably, the processor automatically switches between the first illumination period, the second illumination period, and the third illumination period, acquires a white light equivalent image signal as an image signal during the first illumination period, acquires a calculation image signal as an image signal during the second illumination period, acquires a correction image signal during the third illumination period, generates a first A exposure control signal as an exposure control signal based on the white light equivalent image signal, generates a first B exposure control signal as an exposure control signal based on the calculation image signal, generates a second exposure control signal based on the correction image signal, and controls the light source and imaging optical system according to the first A exposure control signal, the first B exposure control signal, and the second exposure control signal.

[0015] The control of the light source preferably includes the control of the light intensity, and the control of the imaging optical system preferably includes the control of the aperture value, the control of the exposure time, and the control of the gain.

[0016] A method of operating an endoscope system according to the present invention is a method of operating an endoscope system that illuminates an observation target and captures reflected light from the observation target, and includes a light source device, an imaging optical system, and a processor. The light source device emits at least one or more light sources, and sequentially emits observation illumination light and correction illumination light having a spectrum different from the observation illumination light to the observation target. The imaging optical system captures the reflected light. The processor obtains a plurality of different types of image signals during a first exposure period when the observation illumination light is emitted and a second exposure period when the correction illumination light is emitted. The processor calculates the image signal brightness from some of the image signals that are different from each other, outputs a control amount that varies according to the image signal brightness, generates a plurality of different types of exposure control signals according to the control amount, and controls the light source and the imaging optical system according to the exposure control signals.

Advantages of the Invention

[0017] According to the present invention, when emitting a plurality of types of illumination light to obtain a plurality of types of image signals, it is possible to obtain highly reliable image signals for each type of the obtained image signals.

Brief Description of the Drawings

[0018] [Figure 1] It is a schematic diagram of an endoscope system. [Figure 2] It is a block diagram showing the functions of an endoscope system. [Figure 3] It is a block diagram showing the configuration of a light source unit. [Figure 4] It is a graph showing the spectrum of normal light. [Figure 5] It is a graph showing the spectrum of white equivalent light. [Figure 6] It is a graph showing the spectrum of illumination light for calculation. [Figure 7] It is a graph showing the spectrum of correction illumination light. [Figure 8] It is an explanatory diagram showing a first light emission pattern. [Figure 9]It is an explanatory diagram showing the second light emission pattern. [Figure 10] It is an explanatory diagram showing the third light emission pattern. [Figure 11] It is a graph showing a blue transmission band, a green transmission band, and a red transmission band. [Figure 12] It is a block diagram showing the function of the tissue oxygen saturation image generation unit. [Figure 13] It is a graph showing the position of the oxygen saturation contour line in a two-dimensional space with the X-axis being ln(R1 / G1) and the Y-axis being ln(B2 / G1). [Figure 14] It is a graph showing the absorption coefficients of oxyhemoglobin and deoxyhemoglobin. [Figure 15] It is an explanatory diagram showing a method for calculating oxygen saturation. [Figure 16] It is a graph showing the absorption coefficient of yellow pigment. [Figure 17] It is a graph showing the position of the curved surface in a three-dimensional space with the X-axis being ln(R1 / G1), the Y-axis being ln(B2 / G1), and the Z-axis being ln(B3 / G3). [Figure 18] It is a block diagram showing the function of the exposure control unit. [Figure 19] It is an explanatory diagram showing the generation of an exposure control signal. [Figure 20] It is an explanatory diagram showing a method for calculating the image signal brightness and the brightness difference and outputting a control amount. [Figure 21] It is a block diagram showing the function of the brightness calculation unit. [Figure 22] It is an image diagram showing regions with high and low analysis accuracy in an endoscopic image. [Figure 23] It is an image diagram showing a case where a specific region is a region excluding only the outermost region of the image region for brightness calculation. [Figure 24] It is an image diagram showing a case where a specific region is a donut-shaped region of the image region for brightness calculation. [Figure 25] It is an image diagram showing a case where a specific region is a region close to the center of the image region for brightness calculation. [Figure 26]This is an explanatory diagram illustrating an example of extracting a first region from the first analysis image and performing exposure control. [Figure 27] This is an explanatory diagram illustrating an example of extracting a second region from a second analysis image and performing exposure control. [Figure 28] This is an explanatory diagram illustrating the case in which exposure control is performed by generating the 1A exposure control signal, the 1B exposure control signal, and the 2nd exposure control signal. [Figure 29] This is an explanatory diagram showing the specific details of exposure control. [Figure 30] This is a flowchart showing the exposure control flow. [Figure 31] This image diagram shows an example of an image used for correction. [Figure 32] This is a block diagram showing the functions of the correction image generation unit, the reliability calculation unit, and the correction determination unit. [Figure 33] This graph shows the table used to calculate the first confidence level. [Figure 34] This graph shows the table used to calculate the second confidence level. [Figure 35] This graph shows the table used to calculate the third confidence level. [Figure 36] This image diagram shows an example of a correction image when the saturation of the correction image is changed according to the confidence level. [Figure 37] This image shows an example of a correction image where a specific correction area is enclosed in a frame based on the level of confidence. [Figure 38] This image diagram shows an example of a correction image to demonstrate that the correction process can be performed properly. [Figure 39] This image shows an example of a correction image used when displaying a warning. [Figure 40] Block diagram showing the functions of the endoscopic system according to the second embodiment. [Figure 41] This is a plan view of the rotating filter. [Figure 42] This is an explanatory diagram showing the difference value ΔZ used in the calculation value correction process. [Figure 43] This is a schematic diagram of the endoscopic system according to the third embodiment. [Figure 44]This graph shows the spectrum of a mixed light containing primary blue light, secondary blue light, green light, and red light. [Figure 45] This is an explanatory diagram showing the function of the camera head of the third embodiment. [Figure 46] This graph shows the spectrum of light incident on the imaging sensor 511. [Figure 47] This graph shows the spectrum of light incident on the image sensor 512. [Figure 48] This graph shows the spectrum of light incident on the image sensor 513. [Figure 49] This graph shows the spectrum of light incident on the image sensor 514. [Figure 50] This is an explanatory diagram showing the function of the camera head of the fourth embodiment. [Figure 51] (A) is a graph showing the spectrum of the second mixed light. (B) is a graph showing the relationship between the reflectance and transmittance of light incident on the dichroic mirror of the fourth embodiment and the wavelength of light. (C) is a graph showing the relationship between the sensitivity of the image sensor 611 and the wavelength of light. [Figure 52] (A) is a graph showing the spectrum of the second mixed light. (B) is a graph showing the relationship between the reflectance and transmittance of light incident on the dichroic mirror of the fourth embodiment and the wavelength of light. (C) is a graph showing the relationship between the sensitivity of the image sensor 612 and the wavelength of light. [Figure 53] This is an explanatory diagram showing the light emission pattern of the oxygen saturation mode in the fourth embodiment. [Figure 54] This is an explanatory diagram showing the light emission pattern of the correction mode of the fourth embodiment. [Figure 55] (A) is a graph showing the spectrum of the corrective illumination light. (B) is a graph showing the relationship between the reflectance and transmittance of light incident on the dichroic mirror of the fourth embodiment and the wavelength of light. (C) is a graph showing the relationship between the sensitivity of the image sensor 611 and the wavelength of light. [Figure 56](A) is a graph showing the spectrum of the corrective illumination light. (B) is a graph showing the relationship between the reflectance and transmittance of light incident on the dichroic mirror of the fourth embodiment and the wavelength of light. (C) is a graph showing the relationship between the sensitivity of the image sensor 612 and the wavelength of light. [Figure 57] A block diagram showing the functions of the endoscopic system according to the fourth embodiment. [Figure 58] (A) is an image diagram showing the correction area. (B) is a magnified view of the correction area shown in (A). [Figure 59] This is a block diagram showing the functions of the reliability calculation unit, correction determination unit, and extended display control unit of the fourth embodiment. [Figure 60] This graph shows a table for calculating the first confidence score, with the horizontal axis representing the signal value of the G1 image signal. [Figure 61] This graph shows a table for calculating the third confidence level, with the vertical axis representing the signal ratio ln(B2 / G1). [Figure 62] This is an explanatory diagram showing the relationship between the light emission pattern in the correction mode of the fourth embodiment, the generated endoscopic image, and the image set. [Figure 63] This is an explanatory diagram showing the corresponding correction regions between the white light equivalent image, the second blue light image, and the corrective illumination light image. [Figure 64] This is an explanatory diagram showing how to calculate the correlation coefficient. [Figure 65] This image diagram shows an example of a display when a warning is displayed according to the fourth embodiment. [Modes for carrying out the invention]

[0019] [First Embodiment] As shown in Figure 1, the endoscope system 10 includes an endoscope 12, a light source device 13, a processor device 14, a display 15, and a user interface 16. The endoscope 12 is optically connected to the light source device 13 and electrically connected to the processor device 14.

[0020] The light source device 13 supplies illumination light to the endoscope 12. The display 15 displays normal light images, white light equivalent images, and / or tissue oxygen saturation images (oxygen saturation images and / or corrected oxygen saturation images). The user interface 16 has a keyboard, mouse, microphone, tablet, and stylus, and accepts input operations such as function settings. The processor device 14 controls the light source device 13 and performs image processing, analysis, and display control for image signals transmitted from the endoscope 12.

[0021] The endoscope 12 has an insertion section 12a, an operating section 12b, a bending section 12c, and a tip section 12d. The insertion section 12a is inserted into the body of the subject. The operating section 12b is provided at the base of the insertion section 12a. The bending section 12c and tip section 12d are provided on the tip side of the insertion section 12a. The bending section 12c is bent by operating the angle knob 12e of the operating section 12b. The tip section 12d is directed in a desired direction by the bending movement of the bending section 12c. The tip section 12d illuminates the object of observation with illumination light and receives reflected light from the object of observation to image the object. A forceps channel (not shown) for inserting treatment instruments, etc., may be provided from the insertion section 12a to the tip section 12d. Treatment instruments are inserted into the forceps channel through the forceps opening 12j.

[0022] The control unit 12b is equipped with an angle knob 12e, a mode selector switch 12f, a still image acquisition instruction switch 12h, and a zoom control unit 12i. The mode selector switch 12f is used to switch modes. The still image acquisition instruction switch 12h is used to instruct the acquisition of a still image of the subject. The zoom control unit 12i is used to operate the zoom lens 42.

[0023] The endoscope system 10 has three modes: normal observation mode (normal mode), tissue oxygen saturation observation mode (observation mode), and corrected tissue oxygen saturation calculation mode (correction mode). These three modes can be switched via the central control unit 50 by the user operating the mode switching switch 12f. The normal mode is a mode in which the observation target is illuminated with normal light, which is broadband white light, and a normal light image is generated and displayed based on the normal light image signal. Normal light is illumination light suitable for screening observation.

[0024] The observation mode involves irradiating the object of observation with white light equivalent light and illumination light for oxygen saturation calculation (calculation illumination light), acquiring a white light equivalent image signal and an oxygen saturation calculation image signal (calculation image signal), calculating the oxygen saturation of the object of observation based on the calculation image signal, and displaying an oxygen saturation image by superimposing the calculated oxygen saturation onto a base image generated from the white light equivalent image signal.

[0025] The white equivalent light and calculation illumination light used in observation mode are, as described below, illumination lights for observing oxygen saturation images, which are created by illuminating multiple light sources and combining multiple colors of light. Therefore, unless otherwise specified, the white equivalent light and calculation illumination light will be collectively referred to as observation illumination light. Also, when referring to the white light equivalent image signal and / or calculation image signal, they will be referred to as observation image signals.

[0026] The correction mode is a mode in which, in addition to the observation illumination light, a correction illumination light for calculating corrected oxygen saturation (correction illumination light) is irradiated onto the observation target, and when calculating oxygen saturation, the correction oxygen saturation is calculated by performing calculations that take into account the effect of specific dyes contained in the observation target, based on the correction image signal for calculating corrected oxygen saturation (correction image signal). In this specification, "tissue oxygen saturation" refers to either the oxygen saturation calculated in observation mode, or the correction oxygen saturation calculated in correction mode, or both.

[0027] In normal mode, the display 15 shows a normal light image with natural colors. In observation mode, the display 15 shows a white light equivalent image and an oxygen saturation image, which is an image of the calculated oxygen saturation using pseudo-color or similar methods. In correction mode, the display 15 shows a white light equivalent image and a corrected oxygen saturation image, which is an image of the corrected oxygen saturation.

[0028] In Figure 2, the light source device 13 comprises a light source unit 20, a light source control unit 21, and an optical path coupling unit 22. The light source unit 20 is composed of, for example, a semiconductor light source such as a multi-color LED (Light Emitting Diode), a laser light source, a combination of a laser diode and a phosphor, or a xenon lamp or halogen light source. The light source unit 20 has multiple light sources, and by turning each of these multiple light sources on or off, and controlling the amount of light emitted from each light source when they are on, it emits illumination light to illuminate the object of observation.

[0029] In the first embodiment, as shown in Figure 3, the light source unit 20 has five color LEDs (semiconductor light sources): V-LED (Violet Light Emitting Diode) 20a, BS-LED (Blue Short-wavelength Light Emitting Diode) 20b, BL-LED (Blue Long-wavelength Light Emitting Diode) 20c, G-LED (Green Light Emitting Diode) 20d, and R-LED (Red Light Emitting Diode) 20e. However, the combination of each color LED is not limited to this.

[0030] V-LED (Violet Light Emitting Diode) 20a emits violet light V with a central wavelength of 410 ± 10 nm. BS-LED 20b emits first blue light BS with a central wavelength of 450 nm ± 10 nm. BL-LED 20c emits second blue light BL with a central wavelength of 470 nm ± 10 nm. G-LED 20d preferably emits green light G in the green band with a central wavelength of approximately 540 nm. R-LED 20e preferably emits red light R in the red band with a central wavelength of approximately 630 nm. Note that the central wavelength and peak wavelength of each LED 20b to 20e may be the same or different.

[0031] The light source control unit 21 independently inputs control signals to each LED 20a to 20e. By independently controlling the on / off state and the amount of light emitted when each LED 20a to 20e is on, the light source control emits at least one type of illumination light, and ultimately emits normal light, white equivalent light, calculation illumination light, or correction illumination light according to the set mode. Exposure control by the light source control unit 21 will be described later.

[0032] When emitting normal light with a spectrum as shown in Figure 4, V-LED20a, BS-LED20b, G-LED20d, and R-LED20e are lit simultaneously to combine violet light V, first blue light BS, green light G, and red light R. When emitting white equivalent light with a spectrum as shown in Figure 5, BS-LED20b, G-LED20d, and R-LED20e are lit simultaneously to combine first blue light BS, green light G, and red light R. When emitting calculation illumination light with a spectrum as shown in Figure 6, BL-LED20c, G-LED20d, and R-LED20e are lit simultaneously to combine second blue light BL, green light G, and red light R. When emitting correction illumination light with a spectrum as shown in Figure 7, G-LED20d is lit to emit green light G. As shown in Figures 4 to 7, the normal light, observation illumination light (white equivalent light and calculation illumination light), and correction illumination light are illumination lights with different spectra.

[0033] Furthermore, the light source control unit 21 preferably changes the light intensity of V-LED20a, BS-LED20b, BL-LED20c, G-LED20d, and R-LED20e every frame F according to a specific pattern set in each mode, and emits normal light, white equivalent light, calculation illumination light, or correction illumination light in each frame. A frame is a unit of time that includes at least the period from the timing of illumination light emission to the completion of image signal readout by the imaging sensor 43. The light source control unit 21 automatically switches the illumination light emitted from the light source unit 20 for each illumination period of each frame.

[0034] In normal mode, a light emission pattern is repeated in which normal light is emitted during the normal light illumination period of each frame F. In observation mode, for example, as shown in Figure 8, a light emission pattern (first light emission pattern) is repeated in which one frame's worth of white equivalent light WL is emitted during the white equivalent light illumination period (first illumination period) P1 in which the object of observation is illuminated by white equivalent light WL, and one frame's worth of calculation illumination light OL is sequentially emitted during the calculation illumination period (second illumination period) P2 in which the object of observation is illuminated by calculation illumination light OL. The first exposure period EP1, which will be described later, includes the first illumination period P1 and the second illumination period P2. If the first illumination period P1 and the second illumination period P2 are consecutive, it is preferable to include the consecutive first illumination period P1 and second illumination period P2 in the first exposure period EP1.

[0035] In correction mode, for example, as shown in Figure 9, a light emission pattern (second light emission pattern) is repeated in which one frame of white equivalent light WL is emitted during the white equivalent light illumination period P1, one frame of calculation illumination light OL is emitted during the calculation illumination period P2, and one frame of calculation illumination light OL is sequentially emitted during the correction illumination period (third illumination period) P3, in which the object of observation is illuminated by the correction illumination light CL. The second exposure period EP2, which will be described later, includes the third illumination period P3. Each illumination period is automatically switched by the light source control unit 21. Note that the light emission pattern is not limited to this and can be set arbitrarily. For example, as shown in Figure 10, a light emission pattern (third light emission pattern) may be used in which white equivalent light WL, calculation illumination light OL, white equivalent light WL, and correction illumination light CL are emitted sequentially in each frame. In such a case, the first exposure period is defined as the period in which the first illumination period or the second illumination period is continuous, such as the first illumination period, the second illumination period, and the first illumination period.

[0036] The tip 12d of the endoscope 12 is provided with an illumination optical system 30 and an imaging optical system 40. The illumination optical system 30 is an optical system for irradiating the object of observation with illumination light and has an illumination lens 31. Illumination light such as normal light, white equivalent light, calculation illumination light, and correction illumination light is propagated by the light guide 23 and irradiated onto the object of observation via the illumination lens 31.

[0037] On the other hand, when the light source unit 20 is built into the tip 12d of the endoscope 12, illumination light such as normal light, white equivalent light, calculation illumination light, and correction illumination light is emitted by passing through the illumination lens 31 of the illumination optical system 30 without going through the light guide 23. The imaging optical system 40 is an optical system that captures reflected light from the observation target illuminated by the illumination light, and comprises an objective lens 41, a zoom lens 42, and an image sensor 43. The imaging optical system 40 also comprises an aperture 47 (not shown) and a shutter 48, which will be described later. The aperture 47 and shutter 48 are not components, but may be electronically controlled electronic apertures and electronic shutters. Reflected light from the observation target illuminated by the illumination light enters the image sensor 43 via the objective lens 41 and the zoom lens 42, and an image of the observation target is formed on the image sensor 43. The zoom lens 42 is a lens for magnifying the observation target, and moves between the telephoto end and the wide-angle end by operating the zoom operation unit 12i.

[0038] The imaging sensor 43 is a color imaging sensor that captures an object being observed while illuminated by illumination light. Each pixel of the imaging sensor 43 is provided with either a B pixel (blue pixel) having a B (blue) color filter, a G pixel (green pixel) having a G (green) color filter, or an R pixel (red pixel) having an R (red) color filter. Preferably, the imaging sensor 43 is a Bayer array color imaging sensor in which the ratio of the number of B pixels, G pixels, and R pixels is 1:2:1.

[0039] As shown in Figure 11, the B color filter BF primarily transmits light in the blue band, specifically light with a wavelength range of 380-560 nm (blue transmission band). The peak wavelength at which transmittance is maximum is around 460-470 nm. The G color filter GF primarily transmits light in the green band, specifically light with a wavelength range of 450-630 nm (green transmission band). The R color filter RF primarily transmits light in the red band, specifically light with a wavelength range of 580-760 nm (red transmission band).

[0040] As the image sensor 43, a CCD (Charge Coupled Device) image sensor or a CMOS (Complementary Metal-Oxide Semiconductor) image sensor can be used. Alternatively, instead of the primary color image sensor 43, a complementary color image sensor equipped with complementary color filters for C (cyan), M (magenta), Y (yellow), and G (green) may be used. When using a complementary color image sensor, a four-color image signal of CMYG is output, so by converting the four-color image signal of CMYG to a three-color image signal of RGB through complementary-to-primary color conversion, image signals of each RGB color similar to those of the image sensor 43 can be obtained. The image sensor 43 is driven and controlled by the image control unit 44.

[0041] The imaging sensor 43 senses reflected light from an object illuminated with normal light and outputs normal light image signals (Bc image signal, Gc image signal, Rc image signal). It also senses reflected light from an object illuminated with white light equivalent and outputs white light equivalent image signals (B1 image signal, G1 image signal, R1 image signal). Furthermore, it senses reflected light from an object illuminated with calculation illumination light and outputs calculation image signals (B2 image signal, G2 image signal, R2 image signal). In addition, it senses reflected light from an object illuminated with correction illumination light and outputs correction image signals (B3 image signal, G3 image signal, R3 image signal). The imaging control unit 44 controls the imaging sensor 43 to output normal light image signals in normal mode, white light equivalent image signals and calculation image signals in observation mode, and white light equivalent image signals, calculation image signals, and correction image signals in correction mode.

[0042] In other words, as shown in Table 1, when obtaining a normal light image signal in normal mode, the light source control unit 21 controls the V-LED20a, BS-LED20b, G-LED20d, and R-LED20e to light up and emit normal light during the normal light illumination period, and the imaging control unit 44 takes images for each frame and controls the imaging sensor 43 to output a Bc image signal from the B pixels, a Gc image signal from the G pixels, and an Rc image signal from the R pixels.

[0043] [Table 1]

[0044] Furthermore, as shown in Table 2, in observation mode, a white light equivalent image signal and a calculation image signal are acquired. When obtaining a white light equivalent image signal, the light source control unit 21 controls the BS-LED20b, G-LED20d, and R-LED20e to emit white light equivalent WL during the white light equivalent illumination period P1, and the imaging control unit 44 takes images every frame and controls the imaging sensor 43 to output a B1 image signal from the B pixels, a G1 image signal from the G pixels, and an R1 image signal from the R pixels. When obtaining a calculation image signal, the light source control unit 21 controls the BL-LED20c, G-LED20d, and R-LED20e to emit calculation illumination light OL during the calculation illumination period P2, and the imaging control unit 44 takes images every frame and controls the imaging sensor 43 to output a B2 image signal from the B pixels, a G2 image signal from the G pixels, and an R2 image signal from the R pixels.

[0045] [Table 2]

[0046] Furthermore, as shown in Table 3, in correction mode, a white light equivalent image signal, a calculation image signal, and a correction image signal are acquired. When obtaining a white light equivalent image signal or a calculation image signal in correction mode, the same control as in observation mode is performed. When obtaining a correction image signal, the light source control unit 21 controls the G-LED 20d to light up and emit correction illumination light, and the imaging control unit 44 takes images every frame and controls the imaging sensor 43 to output a B3 image signal from the B pixels, a G3 image signal from the G pixels, and an R3 image signal from the R pixels. Exposure control by the imaging control unit 44 will be described later. Note that the correction illumination light is not limited to monochromatic light as shown in Table 3, but may also be illumination light using multiple colors.

[0047] [Table 3]

[0048] The CDS / AGC (Correlated Double Sampling / Automatic Gain Control) circuit 45 performs correlated double sampling (CDS) and automatic gain control (AGC) on the analog image signal obtained from the imaging sensor 43. The image signal that has passed through the CDS / AGC circuit 45 is converted into a digital image signal by the A / D (Analog / Digital) converter 46. As a result, the digital image signal after A / D conversion is input to the processor device 14.

[0049] The processor unit 14 comprises a central control unit 50, an image signal acquisition unit 60, a DSP (Digital Signal Processor) 61, a noise reduction unit 62, an image processing switching unit 63, a normal light image generation unit 70, a tissue oxygen saturation image generation unit 80, a display control unit 100, and an exposure control unit 110 (see Figure 2). Programs for each process are stored in a program memory (not shown) of the processor unit 14. The central control unit 50, which is composed of the processor, executes the programs in the program memory, thereby realizing the functions of the image signal acquisition unit 60, DSP 61, noise reduction unit 62, image processing switching unit 63, normal light image generation unit 70, tissue oxygen saturation image generation unit 80, display control unit 100, and exposure control unit 110. Accordingly, the functions of the following components included in the tissue oxygen saturation image generation unit 80—the white light equivalent image generation unit 81, the base image generation unit 82, the signal ratio calculation unit 83, the oxygen saturation calculation unit 84, the oxygen saturation image generation unit 85, the corrected oxygen saturation calculation unit 86, and the corrected oxygen saturation image generation unit 87—as well as the exposure control signal generation unit 120, the control amount output unit 130, the brightness calculation unit 140, the analysis area setting unit 141, the area brightness calculation unit 142, and the brightness difference calculation unit 150—are realized.

[0050] The image signal acquisition unit 60 receives image signals (normal light image signal, white light equivalent image signal, calculation image signal, and correction image signal) input from the endoscope 12 and transmits the received image signals to the DSP 61. The DSP 61 performs various signal processing on the received image signals, such as defect correction processing, offset processing, gain correction processing, linear matrix processing, gamma conversion processing, demosaicing processing, and YC conversion processing. In defect correction processing, the signals of defective pixels in the imaging sensor 43 are corrected. In offset processing, the dark current component is removed from the image signal after defect correction processing, and an accurate zero level is set. Gain correction processing adjusts the signal level of each image signal by multiplying the image signal of each color after offset processing by a specific gain. After gain correction processing, the image signal of each color is subjected to linear matrix processing to improve color reproduction.

[0051] Subsequently, gamma conversion processing adjusts the brightness and saturation of each image signal. After linear matrix processing, the image signal is subjected to demosaicing (also called isotropization or simulcasting), and interpolation generates signals for the missing colors of each pixel. Demosaicing ensures that all pixels have signals for each RGB color. DSP61 applies YC conversion processing to each image signal after demosaicing and outputs the luminance signal Y, chrominance signal Cb, and chrominance signal Cr to the noise reduction unit 62.

[0052] The noise reduction unit 62 applies noise reduction processing, such as the moving average method or the median filter method, to the image signal that has undergone demosaicing or other processing by the DSP 61. The noise-reduced image signal is input to the image processing switching unit 63 and the exposure control unit 110, which will be described later.

[0053] The image processing switching unit 63 transmits the image signal from the noise reduction unit 62 to the normal light image generation unit 70 in normal mode, and to the tissue oxygen saturation image generation unit 80 in observation mode or correction mode. The normal light image generation unit 70 further applies color conversion processing such as 3x3 matrix processing, grayscale conversion processing, and 3D LUT (Look Up Table) processing to the Bc image signal, Gc image signal, and Rc image signal for one frame that has been input. Next, various color enhancement processing is applied to the RGB image data after color conversion processing. Structural enhancement processing such as spatial frequency enhancement is then applied to this RGB image data after color enhancement processing. The RGB image data that has undergone structural enhancement processing is transmitted to the display control unit 100 as a normal light image.

[0054] As shown in Figure 12, the tissue oxygen saturation image generation unit 80 includes a white light equivalent image generation unit 81, a base image generation unit 82, a signal ratio calculation unit 83, an oxygen saturation calculation unit 84, an oxygen saturation image generation unit 85, a corrected oxygen saturation calculation unit 86, and a corrected oxygen saturation image generation unit 87.

[0055] The white light equivalent image generation unit 81 further applies color conversion processing such as 3x3 matrix processing, gradation conversion processing, and 3D LUT processing to the B1 image signal, G1 image signal, and R1 image signal for one frame that has been input. Next, various color enhancement processing is applied to the RGB image data that has undergone color conversion processing. Structural enhancement processing such as spatial frequency enhancement is then applied to this RGB image data that has undergone color enhancement processing. The RGB image data that has undergone structural enhancement processing is transmitted to the display control unit 100 as a white light equivalent image. The white light equivalent image is an image that corresponds to the normal light image obtained in normal mode, but because the spectrum of the illumination light irradiated on the object of observation is different, the color tone is different from that of the normal light image.

[0056] The base image generation unit 82 performs various signal processing on the input image signal for one frame, for example, the B1 image signal, the G1 image signal, and the R1 image signal, to generate a base image for generating an oxygen saturation image or a corrected oxygen saturation image. The image signal input to the base image generation unit 82 is not limited to the B1 image signal, the G1 image signal, and the R1 image signal, but may be any image signal for generating an image that the user wants to use as the base image. In observation mode, the base image is transmitted to the oxygen saturation image generation unit 85, and in correction mode, it is transmitted to the corrected oxygen saturation image generation unit 87.

[0057] The oxygen saturation calculation unit 84 calculates oxygen saturation in observation mode based on the white light equivalent image signal and the calculation image signal. The method for calculating oxygen saturation will be described later. The calculated oxygen saturation information is transmitted to the oxygen saturation image generation unit 85. The oxygen saturation image generation unit 85 generates an oxygen saturation image in which the oxygen saturation information is represented as pseudo-color on a base image. The base image is, for example, a white light equivalent image or an image with adjusted saturation of a white light equivalent image. Using a white light equivalent image with reduced saturation as the base image has the advantage of improving visibility when the area for displaying oxygen saturation is small. Alternatively, an endoscopic image based on an image signal acquired using other illumination light may be used as the base image. Alternatively, a normal light image may be used as the base image. The oxygen saturation image is transmitted to the display control unit 100.

[0058] The corrected oxygen saturation calculation unit 86 calculates the corrected oxygen saturation in the correction mode based on the white light equivalent image signal, the calculation image signal, and the correction image signal. The method for calculating the corrected oxygen saturation will be described later. The calculated corrected oxygen saturation information is transmitted to the corrected oxygen saturation image generation unit 87. The corrected oxygen saturation image generation unit 87 generates an oxygen saturation image by superimposing the corrected oxygen saturation information as pseudo-color onto the base image. The corrected oxygen saturation image is transmitted to the display control unit 100.

[0059] The display control unit 100 converts the normal light image from the normal light image generation unit 70, and the white light equivalent image, oxygen saturation image, and corrected oxygen saturation image from the tissue oxygen saturation image generation unit 80 into video signals that enable full-color display on the display 15. The converted video signals are input to the display 15. As a result, the normal light image, white light equivalent image, oxygen saturation image, and / or corrected oxygen saturation image are displayed on the display 15. Alternatively, multiple displays 15 may be connected to the processor device 14, and the display control unit 100 may cause the normal light image, white light equivalent image, oxygen saturation image, or corrected oxygen saturation image to be displayed on different displays.

[0060] The method for calculating oxygen saturation is described below. The signal ratio calculation unit 83 of the tissue oxygen saturation image generation unit 80 calculates calculated values ​​for oxygen saturation by signal ratio processing based on the B2 image signal, G1 image signal, and R1 image signal. Specifically, the signal ratio calculation unit 83 calculates the signal ratio B2 / G1 of the B2 image signal and the G1 image signal, and the signal ratio R1 / G1 of the R1 image signal and the G1 image signal, as calculated values ​​for oxygen saturation used to calculate oxygen saturation. It is preferable to logarithmically transform (ln) the signal ratios B2 / G1 and R1 / G1. In addition, color difference signals Cr, Cb, saturation S, hue H, etc., calculated from the B2 image signal, G1 image signal, and R1 image signal may be used as calculated values ​​for oxygen saturation.

[0061] The oxygen saturation calculation unit 84 calculates oxygen saturation by referring to the oxygen saturation calculation table 84a and using calculated values ​​for oxygen saturation calculation, such as the signal ratio B2 / G1 and the signal ratio R1 / G1. The oxygen saturation calculation table 84a stores the correlation between the logarithmic values ​​of the signal ratios B2 / G1 and R1 / G1, obtained from simulations and endoscopic images of multiple phantoms that mimic living organisms with multiple levels of oxygen saturation, and the oxygen saturation. When the correlation is represented in a two-dimensional space formed by X (X=ln(R1 / G1)), which is the value in the X-axis direction obtained by logarithmically calculating the signal ratio R1 / G1, and Y (Y=ln(B2 / G1)), which is the value in the Y-axis direction obtained by logarithmically calculating the signal ratio B2 / G1, contour lines connecting parts with the same oxygen saturation are formed almost along the X-axis (horizontal axis), as shown in Figure 13. Furthermore, as oxygen saturation increases, the contour lines are located further down the Y-axis (vertical axis). For example, contour line 84b, which represents 100% oxygen saturation, is located below contour line 84c, which represents 0% oxygen saturation.

[0062] The correlation between the signal ratios B2 / G1 and R1 / G1 and oxygen saturation is closely related to the absorbance and light scattering characteristics of oxyhemoglobin (Graph 84d) and deoxyhemoglobin (Graph 84e), as shown in Figure 14. As shown in Figure 14, in wavelength bands such as 470±10nm for the second blue light BL, where the difference in extinction coefficients between oxyhemoglobin and deoxyhemoglobin is large, the amount of absorbance changes depending on the oxygen saturation of hemoglobin, making it easy to handle oxygen saturation information. Also, in the wavelength band of 540±20nm for green light G, the extinction coefficient of hemoglobin is relatively high, so the amount of absorbance tends to change depending on the blood concentration.

[0063] The B2 image signal, obtained by irradiating the observation target with calculation illumination light including the second blue light BL, changes in signal value particularly strongly depending on oxygen saturation. Similarly, the G1 image signal, obtained by irradiating the observation target with white equivalent light, changes in signal value particularly strongly depending on blood concentration. Furthermore, the R1 image signal, obtained by irradiating the observation target with white equivalent light, changes in signal value with moderate dependence on oxygen saturation. Note that the signal values ​​of the B2, G1, and R1 image signals all fluctuate depending on oxygen saturation, blood concentration, and brightness. Also, "blood concentration" refers to the hemoglobin concentration (amount of blood pigment) in the blood.

[0064] Here, in order to eliminate the influence of brightness from the B2, G1, and R1 image signals, a signal ratio B2 / G1, which is mainly dependent on oxygen saturation, and a signal ratio R1 / G1, which is mainly dependent on blood concentration, are used. This makes it possible to calculate oxygen saturation while considering the influence of blood concentration. The signal ratio B2 / G1 is a value that depends moderately on blood concentration. The signal ratio R1 / G1 is a value that depends moderately on oxygen saturation.

[0065] The oxygen saturation calculation unit 84 refers to the oxygen saturation calculation table 84a and calculates the oxygen saturation corresponding to the signal ratio B2 / G1 and R1 / G1 for each pixel. For example, as shown in Figure 15, the oxygen saturation corresponding to the signal ratio B2* / G1* and R1* / G1* of a particular pixel is "40%". Therefore, the oxygen saturation calculation unit 84 calculates the oxygen saturation of the particular pixel as "40%".

[0066] Furthermore, the signal ratios B2 / G1 and R1 / G1 rarely become extremely large or extremely small. In other words, the combinations of values ​​for the signal ratios B2 / G1 and R1 / G1 rarely fall below the upper limit contour line 84c (see Figure 13) representing 100% oxygen saturation, or conversely, above the lower limit contour line 84b (see Figure 13) representing 0% oxygen saturation. However, if the distribution falls above the upper limit contour line 84c, the oxygen saturation is set to 100%, and if it falls below the lower limit contour line 84b, the oxygen saturation calculation unit 84 sets the oxygen saturation to 0%. In addition, if the points corresponding to the signal ratios B2 / G1 and R1 / G1 do not fall between the upper limit contour line 84c and the lower limit contour line 84b, it may be indicated that the reliability of the oxygen saturation at that pixel is low, and the oxygen saturation may not be calculated.

[0067] The oxygen saturation image generation unit 85 generates an oxygen saturation image by using the oxygen saturation calculated by the oxygen saturation calculation unit 84. Specifically, the oxygen saturation image generation unit 85 acquires the B2 image signal, G1 image signal, and R1 image signal, and applies an oxygen saturation image generation gain to each pixel of these image signals according to the oxygen saturation. Then, it generates RGB image data using the B2 image signal, G1 image signal, and R1 image signal to which the oxygen saturation image generation gain has been applied, and superimposes it onto the base image to form an oxygen saturation image.

[0068] For example, the oxygen saturation image generation gain is multiplied so that the gains of the B and G channels are relatively increased and the gain of the R channel is relatively decreased. Specifically, the oxygen saturation image generation unit 83 multiplies the B2 image signal, G1 image signal, and R1 image signal with the same oxygen saturation image generation gain of "1" for pixels with an oxygen saturation of 60% or more. In contrast, for pixels with an oxygen saturation of less than 60%, the oxygen saturation image generation gain of "1" or more is multiplied for the B2 image signal, and the oxygen saturation image generation gain of less than "1" is multiplied for the G1 and R1 image signals. The RGB image data generated using the B2, G1, and R1 image signals after such oxygen saturation image generation gain processing, and superimposed on the base image as pseudocolor, is the oxygen saturation image. Note that the image signals to which the oxygen saturation image generation gain is applied to generate the oxygen saturation image are not limited to the image signals given in the example above.

[0069] In the oxygen saturation image generated by the oxygen saturation image generation unit 83, for example, it is preferable to represent the high-oxygen region (the region where oxygen saturation is 60-100%) with the same color as the base image, while representing the low-oxygen region (the region where oxygen saturation is 0-60%), where oxygen saturation falls below a specific value, with a different color (pseudo-color) than the base image. In this case, the oxygen saturation image generation unit 83 represents the low-oxygen region by multiplying the low-oxygen region by a gain for generating an oxygen saturation image that converts it to pseudo-color.

[0070] Alternatively, the high-oxygen region may be represented by multiplying it by a gain for generating an oxygen saturation image corresponding to the oxygen saturation level. Furthermore, a color map representing oxygen saturation with a pseudo-color gradient may be generated and used as the oxygen saturation image. In addition, the low-oxygen and high-oxygen regions are separated by 60% oxygen saturation, but this boundary is arbitrary. For example, a color map divided into low-oxygen, medium-oxygen, and high-oxygen regions may be generated by setting multiple specific values. Furthermore, an oxygen saturation image may be generated that displays oxygen saturation as text information such as "A%".

[0071] The method for calculating corrected oxygen saturation is described below. The signal ratio calculation unit 83 of the tissue oxygen saturation image generation unit 87 calculates calculated values ​​for calculating corrected oxygen saturation by signal ratio processing based on the B2 image signal, G1 image signal, R1 image signal, B3 image signal, and G3 image signal. Specifically, the signal ratio calculation unit 83 calculates the signal ratio B2 / G1 of the B2 image signal and the G1 image signal, the signal ratio R1 / G1 of the R1 image signal and the G1 image signal, and the signal ratio B3 / G3 of the B3 image signal and the G3 image signal as calculated values ​​for calculating oxygen saturation used to calculate oxygen saturation. It is preferable to logarithmically scale (ln) the signal ratios B2 / G1, R1 / G1, and B3 / G3. Furthermore, the calculated values ​​for oxygen saturation may include chrominance signals Cr and Cb, or saturation S and hue H, calculated from the B2, G1, R1, B3, and G3 image signals. Alternatively, the signal ratio B1 / G3 of the B1 and G3 image signals may be used instead of the signal ratio B3 / G3. In this case, the pigment value calculation table 86a using the signal ratio B1 / G3 will be referenced as the pigment value calculation table 86a described later.

[0072] The observed substances may include specific dyes other than oxidized or reduced hemoglobin that affect the calculation of oxygen saturation. Examples of such specific dyes include yellow pigments. As shown in Figure 16, the extinction coefficient of yellow pigments has a peak that is highest around a wavelength of 450 ± 10 nm. Therefore, the signal value of the B2 image signal, acquired based on light transmitted through the B color filter BF from the observation illumination light including the second blue light BL with a central wavelength of 470 nm ± 10 nm, is highly dependent on the amount of yellow pigment. Because its value fluctuates significantly depending on the amount of yellow pigment, it affects the calculation of oxygen saturation.

[0073] Specifically, the greater the amount of yellow pigment, the lower the signal value of the B2 image signal, which in turn lowers the signal ratio B2 / G1, resulting in an apparently higher oxygen saturation. Therefore, when the observed object contains yellow pigment, it is possible to calculate corrected oxygen saturation, which is tissue oxygen saturation that takes the effect of yellow pigment into account, by acquiring image signals with relatively low dependence on the amount of yellow pigment (e.g., B3 and G3 image signals). Note that the dependence of the G1 image signal on yellow pigment is low to moderate, and the dependence of the R1 image signal on yellow pigment is low.

[0074] The signal value of the B3 image signal obtained by irradiating the observation target with corrective illumination light containing green light (G) is brightness-dependent, has low oxygen saturation dependence, high blood concentration dependence, and moderate yellow pigment amount dependence. Similarly, the signal value of the G3 image signal obtained by irradiating the observation target with corrective illumination light containing green light (G) is brightness-dependent, has low oxygen saturation dependence, has high blood concentration dependence, and has low to moderate yellow pigment amount dependence.

[0075] The corrected oxygen saturation calculation unit 86 uses signal ratios B2 / G1, R1 / G1, and B3 / G3, which eliminate the effect of brightness from the B2 image signal, G1 image signal, R1 image signal, B3 image signal, and G3 image signal, and obtains corrected oxygen saturation according to the amount of yellow pigment by referring to the pigment value calculation table 86a and the corrected oxygen saturation calculation table 86b. The pigment value calculation table 86a stores the correlation between the signal ratios B2 / G1, R1 / G1, and B3 / G3, as well as the pigment value according to the amount of a specific pigment. The pigment value is a value obtained in advance using simulations or endoscopic images taken of multiple phantoms that mimic living organisms with multiple levels of oxygen saturation.

[0076] When the correlation is represented in a three-dimensional space formed by the values ​​in the X-axis direction (X=ln(R1 / G1)), the values ​​in the Y-axis direction (Y=ln(B2 / G1)), and the values ​​in the Z-axis direction (Z=ln(B3 / G3)), surfaces 86c to 86g are formed for each of the multiple pigment values, as shown in Figure 17. Surfaces 86c to 86g are contour lines of tissue oxygen saturation for a certain pigment value. In Figure 17, surface 86c for pigment value "0", surface 86d for pigment value "1", surface 86e for pigment value "2", surface 86f for pigment value "3", and surface 86g for pigment value "4" are formed according to the signal ratios ln(R1 / G1), ln(B2 / G1), and ln(B3 / G3). In other words, the values ​​of the signal ratios ln(R1 / G1), ln(B2 / G1), and ln(B3 / G3) determine the surfaces used to calculate the corrected oxygen saturation (corrected oxygen saturation calculation table 86b). In addition, in the pigment value calculation table 86a shown in Figure 17, as the amount of yellow pigment increases, Y=ln(B2 / G1) decreases, X=ln(R1 / G1) increases, and Z=ln(B3 / G3) decreases.

[0077] The corrected oxygen saturation calculation unit 86 calculates the corrected oxygen saturation by referring to the corrected oxygen saturation calculation table 86b corresponding to the dye value. The corrected oxygen saturation calculation table 86b, like the oxygen saturation calculation table 84a, stores the correlation between the signal ratio B2 / G1 and the signal ratio R1 / G1 and the corrected oxygen saturation (a two-dimensional space formed by the value in the X-axis direction, X=ln(R1 / G1), and the value in the Y-axis direction, Y=ln(B2 / G1), as shown in Figure 13).

[0078] Note that the corrected oxygen saturation calculation table 86b does not have to be pre-stored. In this case, the corrected oxygen saturation calculation unit 86 creates the corrected oxygen saturation calculation table 86b each time, according to the dye value, which is the correlation between the signal ratio B2 / G1, the signal ratio R1 / G1, and the corrected oxygen saturation. The created corrected oxygen saturation calculation table 86b may be used as the oxygen saturation calculation table 84a in observation mode.

[0079] Alternatively, instead of using the pigment value calculation table 86a, a corrected oxygen saturation that takes into account the effect of yellow pigment may be determined using a corrected oxygen saturation calculation table 86b, which has pre-stored correlations with signal ratios R1 / G1, B2 / G1, and B3 / G3, as well as tissue oxygen saturation.

[0080] The corrected oxygen saturation image generation unit 87 generates a corrected oxygen saturation image by using the corrected oxygen saturation calculated by the corrected oxygen saturation calculation unit 86. Specifically, the corrected oxygen saturation image generation unit 87 acquires the B2 image signal, G1 image signal, and R1 image signal, and applies a corrected oxygen saturation image generation gain corresponding to the oxygen saturation to each pixel of these image signals. Then, it generates RGB image data using the B2 image signal, G1 image signal, and R1 image signal to which the corrected oxygen saturation image generation gain has been applied, and superimposes it onto the base image as a pseudo-color image to obtain the corrected oxygen saturation image. Note that the image signals to which the corrected oxygen saturation image generation gain is applied in order to generate the corrected oxygen saturation image are not limited to the image signals listed in the example above.

[0081] In the corrected oxygen saturation image, similar to the oxygen saturation image, the hypoxic region (oxygen saturation of 0-60%) may be shown with pseudocolor, or the hyperxic region (oxygen saturation of 60-100%) and the hypoxic region (oxygen saturation of 0-60%) may be shown with different colors (pseudocolor). Alternatively, a color map corresponding to the corrected oxygen saturation may be generated and used as the corrected oxygen saturation image, or a corrected oxygen saturation image may be generated that shows the corrected oxygen saturation as text information such as "B%".

[0082] Furthermore, the oxygen saturation image generation unit 85 or the corrected oxygen saturation image generation unit 87 may generate a tissue oxygen saturation image in black if there are regions where the image signal brightness of the image signal to which the tissue oxygen saturation image generation gain (oxygen saturation image generation gain or corrected oxygen saturation image generation gain) is applied is extremely low, such as when the object of observation is far away from the tip 12d of the endoscope 12, or when, for example, the illumination light is not sufficiently irradiated onto the object of observation due to insufficient light intensity. In addition, if there are regions where the image signal brightness of the image signal to which the tissue oxygen saturation image generation gain is applied is extremely high, such as when the reflected light from the object of observation contains a lot of specular reflection, the oxygen saturation image generation unit 85 or the corrected oxygen saturation image generation unit 87 may generate a tissue oxygen saturation image in white if there are regions where the image signal brightness of the image signal to which the tissue oxygen saturation image generation gain is applied is extremely high.

[0083] The exposure control in this embodiment will now be described. The exposure control unit 110 generates different exposure control signals based on the image signals acquired during the first exposure period and the second exposure period, and performs control (exposure control) according to each exposure control signal. As shown in Figure 18, the exposure control unit 110 includes an exposure control signal generation unit 120 and a control amount output unit 130. The control amount output unit 130 further includes a brightness calculation unit 140 and a brightness difference calculation unit 150. The exposure period refers to the period from the timing when various illumination lights are emitted until the timing when the imaging sensor 43 senses the reflected light from the object of observation and outputs an image signal. Exposure control refers to controlling units that operate during the exposure period, such as the light source unit 20 and the imaging optical system 40.

[0084] A specific example of generating different exposure control signals for each exposure period will be explained using Figure 19. Figure 19 shows exposure control when various illumination lights are emitted in the second light emission pattern, but the light emission pattern is not limited to this. First, the exposure control unit 110 receives the observation image signal 111, which is either a white light equivalent image signal IS1 or a calculation image signal IS2 acquired in the first exposure period EP1, from the noise reduction unit 62. It also receives the correction image signal IS3 acquired in the second exposure period EP2.

[0085] Next, the brightness calculation unit 140 of the control amount output unit 130 calculates the observation image signal brightness Y1 from the observation image signal 111 and the correction image signal brightness Y2 from the correction image signal IS3. Here, the brightness calculation unit 140 outputs the image signal brightness from different parts of the image signal depending on the type of image signal received. This is because calculating the image signal brightness for outputting the control amount from different parts of the image signal allows for more appropriate exposure control for the observation image signal and the correction image signal, respectively. The specific method for calculating the image signal brightness from a part of the acquired image signal will be described later. Image signal brightness refers to the brightness information contained in each type of image signal.

[0086] Next, the control quantity output unit 130 outputs a first control quantity CQ1 based on the observation image signal 111, and also outputs a second control quantity CQ2 based on the correction image signal IS3. The control quantity is a value that varies according to the brightness of the image signal. This is because the appropriate exposure amount differs depending on the type of image signal to be acquired, and furthermore, even if the same type of illumination light is emitted, it will differ depending on the image signal obtained for each frame.

[0087] This explains why the control amount differs depending on the type of image signal acquired. Both white light and calculation illumination light are illumination lights that combine multiple colors of light and have similar spectra (see Figures 3 and 4), so the control amounts output for exposure control will be close to each other. On the other hand, correction illumination light has a significantly different spectrum from observation illumination light (see Figures 3 to 5), so the control amount based on the correction image signal output for exposure control will be a different value from the control amount based on the observation image signal.

[0088] The exposure control signal generation unit 120 generates a first exposure control signal ECS1 based on a first control variable CQ1, and a second exposure control signal ECS2 based on a second control variable CQ2. Each exposure control signal is transmitted from the exposure control unit 110 to the central control unit 50, and a first exposure control EC1 corresponding to the first exposure control signal ECS1 and a second exposure control EC2 corresponding to the second exposure control signal ECS2 are performed on the light source unit 20 and the imaging optical system 40. The first exposure control EC1 and the second exposure control EC2 control the emission of illumination light and the acquisition of image signals, which are performed in an exposure period that is chronologically later than the exposure period in which the image signal was acquired. The specific details of the exposure control will be described later. Since the control variables that determine the content of each exposure control signal are different, the first exposure control EC1 and the second exposure control EC2 are controls with different content from each other.

[0089] With the above configuration, exposure control is performed based on the brightness of each image signal acquired from multiple different types of frames, preventing under- or over-brightness between different types of image signals, and allowing image signals captured in later frames in the time series to be obtained under more appropriate conditions based on the control amount.

[0090] Preferably, the control amount is output based on the difference between the image signal brightness calculated from the image signal and the target brightness. The brightness calculation unit 140 calculates the image signal brightness using the image signal. The brightness difference calculation unit 150 calculates the difference (brightness difference) between the preset target brightness and the image signal brightness. The target brightness refers to the target value of the brightness information that each type of image signal possesses, which is preset for each type of image signal. Next, the control amount output unit 130 outputs the control amount based on the brightness difference. However, the method of outputting the control amount is not limited to this. For example, the amount of light emitted by the light source when acquiring the image signal, the aperture value, or the exposure time may be stored, and the difference between these values ​​and their respective target values ​​may be calculated and the control amount output. With the above configuration, appropriate exposure control can be sequentially performed based on the respective image signal brightness obtained according to the observation purpose.

[0091] It is preferable to calculate the image signal brightness from the image signals acquired according to each exposure period, then calculate the brightness difference, output a control amount from the brightness difference, and sequentially generate exposure control signals based on the control amount for later frames in the time series. Specifically, as shown in Figure 20, first, the brightness calculation unit 140 calculates the observation image signal brightness (first image signal brightness) Y1 from the observation image signal 111 and the correction image signal brightness (second image signal brightness) Y2 from the correction image signal IS3. The image signal brightness is calculated according to the image signal obtained in each mode.

[0092] Next, the brightness difference calculation unit 150 calculates the difference between a preset target brightness and the brightness of each image signal. Two target brightnesses are set: a target brightness for the observation image signal (first target brightness) and a target brightness for the correction image signal (second target brightness). The first and second target brightnesses may be the same or different. As shown in Figure 20, the observation image signal brightness difference ΔY1, which is the brightness difference between the observation image signal brightness Y1 and the observation image signal target brightness, is calculated. Additionally, the correction image signal brightness difference ΔY2, which is the brightness difference between the correction image signal brightness Y2 and the correction image signal target brightness, is calculated. The brightness difference is calculated separately for each mode.

[0093] Next, the control variable output unit 130 outputs a control variable based on the brightness difference. As shown in Figure 20, the first control variable CQ1 is calculated based on the brightness difference ΔY1 of the observation image signal. The second control variable CQ1 is calculated based on the brightness difference ΔY2 of the correction image signal. Specifically, for example, the first control variable CQ1 is calculated by substituting the observation image signal brightness Y1 and the observation image signal brightness difference ΔY1 into the first control variable calculation function f(Y1,ΔY1), and the second control variable CQ1 is calculated by substituting the correction image signal brightness Y2 and the correction image signal brightness difference ΔY2 into the second control variable calculation function f(Y2,ΔY2). Note that the method of calculating the control variable is not limited to this.

[0094] Next, the exposure control signal generation unit 120 generates exposure control signals based on the control quantities. As shown in Figure 20, a first exposure control signal ECS1 is generated based on a first control quantity CQ1, and a second exposure control signal ECS2 is generated based on a second control quantity CQ2. The generated exposure control signals are transmitted to the central control unit 50, where a first exposure control EC1 corresponding to the first exposure control signal ECS1 and a second exposure control EC2 corresponding to the second exposure control signal ECS2 are performed. With the above configuration, exposure control appropriate for the exposure period in which observation illumination light is emitted and the exposure period in which correction illumination light is emitted can be performed, respectively.

[0095] It is preferable that the second control variable CQ2 be larger than the first control variable CQ1. This is because the observation illumination light is a bright illumination light that combines multiple colors of light, while the correction illumination light is a dim illumination light that has only one color. The amount of exposure to the image sensor 43 during the second exposure period, when the emitted illumination light is dim, is smaller than the amount of exposure during the first exposure period. Therefore, by increasing the control variable, the brightness of the image signal obtained during the first exposure period can be increased. Another reason is that in order to accurately calculate the corrected oxygen saturation, it is preferable to continuously acquire an image signal with a certain level of image signal brightness. The brightness of the image signal acquired fluctuates depending on the imaging magnification, etc., but the correction image signal, which requires a more precise signal value than the observation image signal, needs to be precisely controlled to the image signal brightness that is particularly necessary for each imaging scene. Therefore, by making the second control variable larger than the first control variable, optimal exposure control for the imaging scene can be continuously performed, and a precise signal value can be obtained.

[0096] The brightness calculation unit 140 of the control amount output unit 130 preferably outputs image signal brightness using an image signal (specific color signal) obtained from a pixel of a specific color. For example, it is preferable to use the B image signal (B signal) as the specific color signal. This is because, among the acquired specific color signals, the image signal brightness of the B signal (B1 image signal, B2 image signal, and B3 image signal) is relatively small compared to the G image signal (G signal) and R image signal (R signal), so by performing exposure control according to the B signal, an image signal with a large image signal brightness can be acquired. In this case, the B1 image signal and / or B2 image signal are used as the observation image signal, and the B3 image signal is used as the correction image signal, and the image signal brightness and brightness difference are calculated and the first control amount and the second control amount are output.

[0097] Specifically, if the target brightness for the B signal is set to "B=100", the brightness difference is calculated according to the brightness of the B signal (B1 image signal, B2 image signal, or B3 image signal). Furthermore, the target brightness for specific color signals may be changed according to each image signal. For example, the target brightness for the B1 and B2 image signals may be set to "B=50", and the target brightness for the B3 image signal may be set to "B=100". Note that the specific color signals may be G, R, or the four CMYG image signals, or combinations thereof, and the target brightness can be arbitrarily set according to each specific color signal. Also, the specific color signal used to calculate the image signal brightness may be changed depending on the type of image signal. For example, the G1 image signal may be used for observation images, and the B3 image signal for correction images.

[0098] Furthermore, when dividing the G pixels of the imaging sensor 43 into Gb pixels and Gr pixels, obtaining a Gb image signal from the Gb pixels and a Gr image signal from the Gr pixels, and calculating the brightness of the image signal by applying these to the formula Y=(K1*Gb+K2*Gr+K3*B+K4*R+K5) / 256 (where Gb, Gr, B, and R are pixel values, and K1, K2, K3, K4, and K5 are arbitrarily set coefficients), the brightness difference may be calculated after changing the coefficients according to the type of image signal, and a control amount may be output.

[0099] The control variable output unit 130 preferably calculates the image signal brightness using the image signal in a portion of the analyzed image generated from the image signal. In this case, as shown in Figure 21, it is preferable to provide the brightness calculation unit 140 with an analysis region setting unit 141 and a region brightness calculation unit 142. The analysis region setting unit 141 generates an analyzed image from each image signal and extracts a specific region from the analyzed image. The region brightness calculation unit 142 calculates the brightness from the image signal in the specific image region.

[0100] The specific region is preferably a region where the reliability and analytical accuracy of the signal values ​​of the image signal are high. As shown in Figure 22, when observing the lumen of the tubular organs of the esophagus, stomach, small intestine, and large intestine, the endoscopic image 141a is divided into a region 141b on the near side, i.e., the side closer to the tip 12d of the endoscope 12, and a region farther away, i.e., the side further from the tip 12d of the endoscope 12, when observing so that the back of the lumen is visible. The far region 141b does not receive sufficient illumination and therefore does not have sufficient brightness, resulting in low analytical accuracy. On the other hand, the near region 141b is bright because it is illuminated with sufficient light, resulting in high analytical accuracy and making it suitable as a specific region. In addition, generally, the closer to the edge of the endoscopic image 141a, the greater the image distortion, resulting in lower analytical accuracy. On the other hand, the closer to the center of the endoscopic image 141a, the less the image distortion, resulting in higher analytical accuracy and making it suitable as a specific region.

[0101] Preferably, the specific region is extracted according to a specific shape that is pre-set depending on the type of image signal acquired. The specific shape is a shape set considering the accuracy of the analysis. For example, there is a shape 141d (shaded area) as shown in Figure 23, which excludes only the outermost region; a donut-shaped shape 141e (shaded area) as shown in Figure 24, which excludes the region near the edge and the region near the center corresponding to the back of the lumen; and a shape 141f (shaded area) as shown in Figure 25, which extracts the region near the center. Alternatively, the specific region may be extracted based on the brightness calculated from each image signal. This method can be used to address cases where the method of extracting to a predetermined shape is unsuitable, such as when the back of the lumen appears at the edge of the endoscopic image, or when the back of the lumen is not included in the endoscopic image as a result of high-magnification imaging of the mucosa.

[0102] The analysis region setting unit 141 extracts a specific region using an observation image signal, which is either a white light equivalent image signal or a calculation image signal. It also extracts a specific region using a correction image signal. In this case, an observation analysis image (first analysis image) is generated using the observation image signal, and the observation analysis region (first region) is extracted as the specific region. Furthermore, a correction analysis image (second analysis image) is generated using the correction image signal, and the correction analysis region (second region) is extracted as the specific region. Next, the region brightness calculation unit 142 calculates the observation region brightness (first region brightness) as the region brightness from the first region. It also calculates the correction region brightness (second region brightness) from the second region.

[0103] Next, the brightness of the first region and the brightness of the second region are transmitted to the brightness difference calculation unit 150. The brightness difference calculation unit 150 calculates the brightness difference of the observation image signal region as the brightness difference between the brightness of the first region and the first target brightness. It also calculates the brightness difference of the correction image signal region as the brightness difference between the brightness of the second region and the second target brightness. The control amount output unit 130 outputs a first region control amount based on the brightness difference of the observation image signal region, and outputs a second region control amount based on the brightness difference of the correction image signal region.

[0104] The exposure control signal generation unit 120 generates a first region exposure control signal based on the first region control amount, and also generates a second exposure control signal based on the second region control amount. Finally, the first exposure control EC1 corresponding to the first region exposure control signal and the second exposure control EC2 corresponding to the second region exposure control signal are performed sequentially, resulting in appropriate exposure control according to each exposure period.

[0105] Diagnosis by physicians and other users using white light equivalent images (or normal light images) and oxygen saturation images should preferably be performed over a wide area of ​​the endoscopic image (for example, the area included in the shapes shown in Figures 23 and 24), excluding areas near the edges where distortion is significant and areas deep within the lumen that are difficult to observe. Furthermore, when calculating corrected oxygen saturation, it is desirable to use the image signal from an area close to the center of the image with high analytical accuracy (for example, the area with the shape included in Figure 25) in order to calculate corrected oxygen saturation with high accuracy. For this reason, it is preferable to use different image signals for exposure control between the observation image signal (white light equivalent image signal and / or calculation image signal) and the correction image signal by changing the analysis area.

[0106] Specifically, when an observation image signal 111 is acquired, as shown in Figure 26, a specific region 151a (first region) is extracted from the first analysis image 151 (the observation analysis image of the previous frame in Figure 26), and an exposure control signal for the first region is generated using the observation image signal of the first region. As a result, the brightness of the observation image signal 152 (the endoscopic image (observation illumination light) of the later frame in Figure 26) acquired during the subsequent first exposure period is adjusted. On the other hand, when a correction image signal is acquired, for example, as shown in Figure 27, a specific region 153a (second region) is extracted from the second analysis image 153 (the correction analysis image of the previous frame in Figure 27), and an exposure control signal for the second region is generated using the correction image signal of the second region. As a result, the brightness of the correction image signal 154 (the endoscopic image (correction illumination light) of the later frame in Figure 27) acquired during the second exposure period is adjusted.

[0107] In Figures 26 and 27, the intensity of the shaded areas indicates the brightness of the image signal. Specifically, when exposure control is performed using the first region 151a, which is a relatively bright region suitable for diagnosis within the already relatively bright observation image signal, the control amount is small, and the change in brightness compared to the subsequently acquired image signal is also small. On the other hand, when exposure control is performed using the second region 153a, which is an even darker region within the already dark correction image signal but offers high analysis accuracy, the control amount is large, and the change in brightness compared to the subsequently acquired image signal is also large. In this way, the image signal used for exposure control can be changed according to the purpose. With the above configuration, an image signal with a more appropriate brightness can be obtained depending on the type of image signal.

[0108] Furthermore, when generating an exposure control signal by outputting a control amount using a portion of the acquired image signal, a method using a specific color signal and a method using an image signal in a specific region may be combined. For example, to calculate the brightness of the observation image signal, the G1 image signal included in the specific region shown in Figures 23 and 24 may be used, and to calculate the brightness of the correction image signal, the B3 image signal included in the specific region shown in Figure 25 may be used.

[0109] In the observation mode, which acquires an observation image signal, and the correction mode, which acquires a correction image signal in addition to the observation image signal, it is preferable to generate different exposure control signals and perform different exposure control in each mode. Since the image signals acquired are different in the observation mode and the correction mode, the resulting exposure control signals will also be different, as shown in Figure 19, for example. Therefore, different exposure control can be performed depending on the mode.

[0110] It is preferable to generate different exposure control signals based on three types of image signals: a white light equivalent image signal IS1, a calculation image signal IS2, and a correction image signal IS3, thereby performing different exposure controls. A specific example will be explained using Figure 28. In this case, the exposure control signal generation unit 120 generates a first A exposure control signal ECS1 based on the white light equivalent image signal IS1 acquired during the first illumination period. The exposure control signal generation unit 120 also generates a first B exposure control signal ECS1B based on the calculation light image signal IS2 acquired during the second illumination period. Furthermore, the exposure control signal generation unit 120 generates a second exposure control signal ECS2 based on the correction image signal IS3 acquired during the third illumination period.

[0111] The exposure control unit 110 transmits each exposure control signal to the central control unit 50, and finally, the first exposure control EC1A corresponding to the first exposure control signal ECS1A, the first exposure control EC1B corresponding to the first exposure control signal ECS1B, and the second exposure control EC2 corresponding to the second exposure control signal ECS2 are performed. With the above configuration, a white light equivalent image signal, a calculation image signal, and a correction image signal with appropriate image signal brightness can be obtained by observing the target of observation and calculating tissue oxygen saturation.

[0112] The exposure control performed by the exposure control unit 110 preferably includes controlling the light intensity that controls the light source unit 20, controlling the aperture value of the aperture 47 in the imaging optical system 40, controlling the exposure time length that controls the shutter 48, or controlling the gain that controls the imaging sensor 43. As shown in Figure 29, the exposure control unit 110 controls the light source control unit 21 or the imaging control unit 44 via the central control unit 50 by transmitting each exposure control signal. When an exposure control signal is transmitted to the light source control unit 21, the light source control unit 21 controls the light source unit 20 and adjusts the light intensity and light intensity ratio.

[0113] When controlling the amount of light, a target amount of light may be set. When a target amount of light is set, the exposure control unit 110 transmits a light amount control signal, which is an exposure control signal, to the light source control unit 21 via the central control unit 50, according to the image signal brightness and brightness difference based on the image signal. The light source control unit 21 controls the amount of light from each light source of the light source unit 20 based on the light amount control signal.

[0114] Specifically, the exposure control unit 110 generates a light intensity control signal based on a first control variable CQ1 output using the observation image signal 111, and the light source control unit 21 adjusts the light intensity and light intensity ratio of BS-LED20b, BL-LED20c, G-LED20d, and R-LED20e. In addition, the light source control unit 21 dims the G-LED20d based on a second control variable CQ2 calculated using the correction image signal IS3, thereby generating a light intensity control signal. With the above configuration, the illumination light for the next frame is sequentially emitted with a light intensity appropriate for each frame, and an appropriate exposure amount can be obtained. The control of light intensity is not limited to this.

[0115] When an exposure control signal is transmitted to the image control unit 44, the image control unit 44 controls the aperture 47, the shutter 48, and / or the image sensor 43. When exposure control is performed by controlling the aperture 47, for example, the exposure control unit 110 sets a target aperture value and generates an exposure control signal, and adjusts the aperture value by controlling the aperture 47. When exposure control is performed by controlling the shutter 48, for example, the exposure control unit 110 sets a target frame rate and a target shutter speed and generates an exposure control signal, and adjusts the exposure time by controlling the shutter 48. When exposure control is performed by controlling the image sensor 43, for example, the exposure control unit 110 sets a target gain and controls the image sensor 43 to adjust the analog gain. The control of the imaging optical system is not limited to these.

[0116] The method of exposure control is not limited to the above. Furthermore, exposure control may be configured to combine the above methods. With the above configuration, exposure control can be performed by selecting an appropriate method based on the image signal acquired during each exposure period.

[0117] In this embodiment, the endoscope 12 is a flexible endoscope, but the present invention is also suitable when a rigid endoscope (laparoscope) used for surgical procedures is used. When a flexible endoscope is used, the observation target is the superficial mucosa viewed from the luminal side of a tubular organ, and endoscopic images such as tissue oxygen saturation images calculated from the tissue oxygen saturation of the superficial mucosa are displayed. When a laparoscope is used, the observation target is the organ viewed from the serosal side, and endoscopic images such as tissue oxygen saturation images calculated from the tissue oxygen saturation of the organ surface are displayed.

[0118] The following describes the sequence of steps for exposure control for each frame in this embodiment. As shown in Figure 30, the image signal acquisition unit 60 acquires an image signal (S101), the brightness calculation unit 140 calculates the brightness of the image signal (S102), the brightness difference calculation unit 150 calculates the brightness difference between the calculated brightness and the target brightness (S103), the control amount output unit outputs a control amount (S104), the exposure control signal generation unit 120 generates an exposure control signal based on the control amount (S105), and the exposure control unit 110 performs exposure control for the next frame (S106).

[0119] The above embodiment makes it possible to obtain image signals of later frames in a time series under more appropriate conditions when acquiring multiple types of image signals. "Multiple types of image signals" refers to the normal light image signal, white light equivalent image signal, calculation image signal, or correction image signal as shown in Tables 1 to 3.

[0120] The exposure control in this embodiment is based on multiple types of image signals obtained by irradiating the observation target with illumination light that has different acquisition purposes and significantly different spectra. By sequentially performing different exposure controls depending on the exposure period for acquiring these image signals, it is possible to prevent insufficient or excessive brightness in each image signal that can occur with exposure control based on a single type of image signal, and to obtain image signals captured under more appropriate conditions.

[0121] In the correction mode, when calculating the corrected oxygen saturation by referring to the corrected oxygen saturation calculation table 86b (surfaces 86c to 86g shown in Figure 17) according to the dye value, some of the image signals obtained in the correction mode may be used. Hereinafter, selecting the corrected oxygen saturation calculation table 86b according to the dye value will be referred to as the "correction process". Some of the image signals are the image signals in a specific correction region of the correction image described later (see Figure 31 described later). In this case, it is preferable that the specific correction region is a region in which disturbances affecting the accuracy of oxygen saturation calculation are less likely to occur. In order to determine the degree of disturbance in the specific correction region, the reliability of the image signals in the specific correction region is calculated.

[0122] Disturbances are those observed in the endoscopic image captured by the endoscope 12 that, other than specific dyes, can cause a decrease in the accuracy of oxygen saturation calculation, such as halation, dark areas, bleeding, fat, and deposits on the mucosal surface. Halation and dark areas are related to the brightness of the endoscopic image. Halation refers to areas where the image is overexposed due to strong light entering the imaging sensor 43. Dark areas are areas where the image is dark due to shadows from internal structures such as folds and colonic flexures, instruments, or because illumination light does not easily reach the deepest parts of the lumen.

[0123] Bleeding includes extraserous (intraperitoneal) or intraperitoneal bleeding, and intramucosal bleeding. Fat includes fat observed extraserous (intraperitoneal) in areas such as the greater omentum, lesser omentum, and mesentery, and fat observed on the mucosal surface of the intraperitoneal lumen. Adhesion on the mucosal surface includes biologically derived adhesion such as mucus, blood, and exudate; extracorporeal adhesion such as staining solutions and water supplied from water supply devices; and adhesion that is a mixture of biologically derived and extracorporeal adhesion, or residual liquid or residue.

[0124] In correction mode, when correction processing is performed using the image signal in a specific correction area, a correction image 200, as shown in Figure 31, is first displayed on the display 15 at the moment the system switches to correction mode. The display of the correction image 200 is controlled by the display control unit 100. The correction image 200 displays the specific correction area 201 in a manner that is visible to the user.

[0125] The shape of the correction-specific region 201 is not limited to a circular shape as shown in Figure 31. Furthermore, the position of the correction-specific region 201 is not limited to the center of the image as shown in Figure 31. For example, the correction-specific region may be a donut-shaped area excluding the peripheral part of the correction image 200 where distortion is more pronounced due to the curvature of the lens, and the central part of the correction image 200 which is dark because it corresponds to the back of the lumen. The correction image is preferably a color image (e.g., a white light equivalent image) generated using the B1, G1, and R1 image signals. The correction image may also be an image generated using other image signals.

[0126] Furthermore, when displaying a correction image and calculating the reliability using the image signal in a specific correction region, the correction image is generated by the correction image generation unit 210 of the tissue oxygen saturation image generation unit 80, as shown in Figure 32. In this case, the tissue oxygen saturation image generation unit 80 is further provided with a reliability calculation unit 220 and a correction determination unit 230 in addition to the correction image generation unit 210, as shown in Figure 32.

[0127] When a correction image 200, as shown in Figure 31, is displayed on the display 15, and a confidence calculation instruction is input, the confidence calculation unit 220 calculates the confidence level for each pixel included in the correction specific region 201 based on the image signal in the correction specific region 201. The input of the confidence calculation instruction may be performed according to an input instruction via the user interface, or it may be performed automatically at the same time as the control for displaying the correction image 200.

[0128] The following explains how confidence is calculated. Confidence levels include (1) confidence in the brightness of the endoscopic image, (2) confidence in the degree of bleeding contained in the endoscopic image, and (3) confidence in the degree of fat contained in the endoscopic image.

[0129] The calculation of the reliability of brightness will now be explained. In this case, the reliability calculation unit 220 uses the G2 image signal and calculates the reliability by referring to the first reliability calculation table 221, as shown in Figure 33. The first reliability calculation table 221 is a pre-generated table that shows the relationship between the signal value of the G2 image signal and the reliability. The signal value of the G2 image signal is, for example, the luminance value obtained by performing a conversion process using the G2 image signal. In this case, the reliability is calculated as a value from 0 to 1. In the first reliability calculation table 221, as shown in Figure 33, the reliability of the signal value of the G1 image signal outside a certain range Rx is lower than the reliability of the luminance value of the G1 image signal within a certain range Rx. Specifically, the case in which the signal value of the G1 image signal is outside a certain range Rx is when the pixel contains halation and has a high luminance value, or when the correction specific area contains a dark area and has a very low luminance value, etc. Note that the G1 image signal may be used instead of the G2 image signal to calculate the reliability of brightness (see Figure 60, which will be described later).

[0130] The calculation of confidence based on the degree of bleeding will now be explained. In this case, the confidence calculation unit 220 uses the signal ratio ln(R1 / G1) and the signal ratio ln(B1 / G1) and calculates the confidence by referring to the second confidence calculation table 222, as shown in Figure 34. In the second confidence calculation table 222, the definition line DFX is plotted on a two-dimensional coordinate system where the X axis is ln(R1 / G1) and the Y axis is the signal ratio ln(B1 / G1). In this case, the confidence is calculated such that the lower the coordinate (X1,Y1)=(ln(R1 / G1),ln(B1 / G1)) calculated using the X component value as the signal value ln(R1 / G1) and the Y component value as the signal ratio ln(B1 / G1), the lower the position in the second confidence calculation table 222. Furthermore, if the coordinate (X1,Y1) is located in the region to the upper left of the definition line DFX, the confidence level based on the degree of bleeding is set to a fixed value that indicates high confidence. Note that the signal ratio ln(R1 / G1) is the value obtained by normalizing the R1 image signal by the G1 image signal and taking the logarithmic scale. Similarly, the signal ratio ln(B1 / G1) is the value obtained by normalizing the B1 image signal by the G1 image signal and taking the logarithmic scale.

[0131] The calculation of confidence based on the degree of fat will now be explained. In this case, the confidence calculation unit 220 uses the signal value ln(R2 / G2) and the signal ratio ln(B2 / G2) to calculate the confidence by referring to the third confidence calculation table 223, as shown in Figure 35. In the third confidence calculation table 223, the definition line DFY is plotted on a two-dimensional coordinate system where the X axis is ln(R2 / G2) and the Y axis is the signal ratio ln(B2 / G2). In this case, the confidence is calculated such that the lower left the coordinate (X2,Y2)=(ln(R2 / G2),ln(B2 / G2)) calculated using the X component value as the signal value ln(R2 / G2) and the Y component value as the signal ratio ln(B2 / G2) is located in the third confidence calculation table 223, the lower the confidence. Furthermore, if the coordinate (X2,Y2) is located in the region to the upper right of the definition line DFY, the confidence level based on the degree of fat is set to a fixed value that indicates high confidence. The signal ratio ln(R2 / G2) is the value obtained by normalizing the R2 image signal by the G2 image signal and taking the logarithmic scale. Similarly, the signal ratio ln(B2 / G2) is the value obtained by normalizing the B2 image signal by the G2 image signal and taking the logarithmic scale. Alternatively, by using the G1 image signal instead of the G2 image signal, and the R1 image signal instead of the R2 image signal, the confidence level based on the degree of fat may be calculated based on the position in the confidence level calculation table of the coordinate (X2,Y2)=(ln(R1 / G1),ln(B2 / G1)), where the X component value is the signal value ln(R1 / G1) and the Y component value is the signal ratio ln(B2 / G1) (see Figure 61 described later).

[0132] The confidence calculation unit 220 calculates at least one confidence level from among the confidence levels for brightness (first confidence level), the confidence level for the degree of bleeding (second confidence level), and the confidence level for the degree of fat (third confidence level). The calculated confidence levels are used for notification to prevent areas with low confidence levels from entering the specific correction area, or for weighting the signal values ​​of the image signal used in the correction process (signal ratios ln(R1 / G1), ln(B2 / G1), and ln(B3 / G3)).

[0133] When notification is given to prevent low-reliability areas from entering the specific correction area, the calculated reliability is transmitted to the correction determination unit 230. The correction determination unit 230 uses a pre-set reliability determination threshold to determine the reliability of each pixel in the specific correction area and outputs a determination result indicating whether each pixel is a high-reliability pixel or a low-reliability pixel.

[0134] The correction determination unit 230, for example, designates pixels whose reliability is equal to or greater than the reliability determination threshold as high-reliability pixels, and pixels whose reliability is less than the reliability determination threshold as low-reliability pixels. The correction determination unit 230 transmits the determination result, which is the determination made regarding the reliability of each pixel, to the display control unit 100. The display control unit 100 controls the display mode of the correction image 200 displayed on the display 15 according to the determination result.

[0135] The display control unit 100, for example, as shown in Figure 36, sets the saturation of the low-reliability region 201a in the correction-specific region 201 to be higher than the saturation of the high-reliability region 201b. The low-reliability region is a set of pixels that have low-reliability pixels. The high-reliability region is a set of pixels that have high-reliability pixels. By displaying a correction image 200 in which the display patterns of the low-reliability region 201a and the high-reliability region 201b are different, as shown in Figure 36, the user can be notified that the correction-specific region includes a low-reliability region containing a relatively large amount of disturbance. Note that the method of differentiating the display patterns of the low-reliability region 201a and the high-reliability region 201b is not limited to changing the saturation of each region. For example, the brightness, hue, etc., may be different.

[0136] Furthermore, when calculating multiple confidence levels from the first, second, and third confidence levels, the confidence level used for determining the confidence level may be the lowest of the first, second, or third confidence levels. Alternatively, a confidence level determination threshold may be set for each confidence level. For example, a first confidence level determination threshold for the first confidence level, a second confidence level determination threshold for the second confidence level, and a third confidence level determination threshold for the third confidence level may be set in advance, and if any of the confidence levels falls below the confidence level determination threshold, the pixel for which that confidence level was calculated may be determined to be a low-confidence pixel.

[0137] Furthermore, the correction determination unit 230 may also perform a determination on the number of high-reliability pixels based on the reliability calculated for each pixel. In this case, the display control unit 100 changes the display mode of the correction specific area depending on whether the number of high-reliability pixels in the correction specific area is equal to or greater than the threshold for determining the number of high-reliability pixels, or whether it is less than the threshold for determining the number of high-reliability pixels. For example, if the number of high-reliability pixels in the correction specific area is equal to or greater than the threshold for determining the number of high-reliability pixels, the correction image 200 is displayed in a highlighted state by surrounding the correction specific area with a frame 202 of the first determination result color, as shown in Figure 37. By highlighting the correction specific area with a frame of the first determination result color, the user can be informed that the correction process can be performed with minimal influence from disturbances.

[0138] Furthermore, if the number of high-reliability pixels in the specific correction area is less than the threshold for determining the number of high-reliability pixels, the correction image 200 may be displayed in which the specific correction area is highlighted by surrounding it with a frame of a second judgment result color, which is different from the first judgment result color. By highlighting the specific correction area by surrounding it with a frame of the second judgment result color, the user can be notified that the number of pixels with little influence from disturbances is less than a certain value.

[0139] Furthermore, the correction determination unit 230 may determine the number of low-reliability pixels, and the display control unit 100 may change the display mode of the specific correction area depending on whether the number of low-reliability pixels in the specific correction area is equal to or greater than the low-reliability pixel count determination threshold, or less than the low-reliability pixel count determination threshold. In this way, by using a reliability pixel count determination threshold (high reliability pixel count determination threshold or low reliability pixel count determination threshold) and changing the display mode of the correction image depending on the number of high or low reliability pixels, it is possible to inform the user of the degree to which disturbances are present in the specific correction area and prompt the user to operate the endoscope to perform the correction process appropriately.

[0140] Furthermore, the correction determination unit 230 may determine the reliability of each pixel in the specific correction region using a reliability determination threshold and / or a reliability pixel count determination threshold, and if it determines that the influence of disturbances in the specific correction region is small, it may display a message on the correction image 200 indicating that the correction process can be performed appropriately. For example, as shown in Figure 38, a message MS1 such as "The correction process will be performed appropriately" may be superimposed on the correction image 200.

[0141] Furthermore, the correction determination unit 230 may determine the reliability of each pixel in the specific correction area using a reliability determination threshold and / or a reliability pixel count determination threshold, and display a warning if the specific correction area includes a low reliability area, or if the number of low reliability pixels is equal to or greater than the reliability pixel count determination threshold. For example, as shown in Figure 39, a message MS2 such as "Operate the endoscope for correction processing" is superimposed on the correction image 200. Also, if it is determined that the reliability of brightness has a particularly large impact, a message MS3 such as "Avoid dark areas" may be superimposed on the correction image 200, as shown in Figure 39.

[0142] As described above, by changing the display mode of the correction image 200, the user can be notified that the specific correction area includes a low-confidence area containing a relatively large number of disturbances, or that the correction process can be performed appropriately. In addition to the image displayed on the display 15, audio notification may also be provided.

[0143] By providing such notification, the user can be encouraged to operate the endoscope 12 so that areas with minimal disturbance influence fall within the correction-specific area 201. In other words, the user can be encouraged to operate the endoscope 12 so that low-reliability areas are excluded as much as possible, while high-reliability areas are excluded as much as possible, within the correction-specific area.

[0144] Furthermore, if the system notifies that the correction process can be performed appropriately and the user inputs an operation to instruct the execution of the correction process, the correction process in correction mode will be performed. Alternatively, the system may use a confidence threshold and / or a confidence pixel count threshold to determine the confidence level of each pixel in the specific correction region. If it is determined that the influence of disturbances in the specific correction region 201 is minimal, the system may automatically execute the correction process without any user input to instruct the execution of the correction process.

[0145] Alternatively, without displaying the correction image 200 on the display 15, the processor device 14 may calculate the reliability in the specific correction region as an internal process, make a determination on the reliability of each pixel, and then perform the correction process using the image signal in the specific correction region.

[0146] Furthermore, when performing correction processing, the confidence level calculated for each pixel in a specific region may be used to weight the signal values ​​of the B1, G1, R1, B2, B3, and / or G3 image signals, thereby reflecting the confidence level in the correction processing. Additionally, when calculating the signal ratio ln(R2 / G2), signal ratio ln(B1 / G2), and signal ratio ln(B3 / G3) using the average value (average signal value) of the signal values ​​of the B1, G1, R1, B2, B3, and / or G3 image signals in a specific correction region during the correction processing, the weighted average value obtained by weighting the average signal value may be used to calculate these signal ratios.

[0147] [Second Embodiment] In the second embodiment, the light source unit 20 is replaced with a broadband light source 400 that emits broadband light, such as a white LED, xenon lamp, or halogen light source, instead of the color LEDs 20a to 20e shown in the first embodiment. The light emitted from the light source device 13 is used as illumination light to illuminate the subject by combining the broadband light source 400 with the rotating filter 410. The following describes the differences between the endoscope system 10 and the first embodiment, and the descriptions of the common parts are omitted.

[0148] In the second embodiment, as shown in Figure 40, the light source device 13 of the endoscope system 10 is provided with a broadband light source 400, a rotating filter 410, and a filter switching unit 420. The filter switching unit 420 is controlled by the light source control unit 21. The other configurations are the same as those of the endoscope system 10 in the first embodiment. In the second embodiment, it is preferable that the imaging sensor 43 be a monochrome imaging sensor.

[0149] The broadband light source 400 emits broadband light having a wavelength range from blue to red. Broadband light is, for example, white light. The rotating filter 410, as shown in Figure 41, comprises an inner filter 411 located on the inside and an outer filter 412 located on the outside. The filter switching unit 420 moves the rotating filter 410 radially. In normal mode, the filter switching unit 420 inserts the inner filter 411 of the rotating filter 410 into the optical path of white light. In observation mode or correction mode, the filter switching unit 420 inserts the outer filter 412 of the rotating filter 410 into the optical path of white light.

[0150] As shown in Figure 41, the inner filter 411 is provided with a B1 filter 411a that transmits light in the wavelength bands of violet light V and first blue light BS from white light, a G filter 411b that transmits light in the wavelength bands of green light G from white light, and an R filter 411c that transmits light in the wavelength bands of red light R from white light, along the circumferential direction. Therefore, in normal mode, illumination light having the wavelength bands of violet light V and first blue light BS, illumination light having the wavelength band of green light G, and illumination light having the wavelength band of red light R are emitted from the light source device 13 in accordance with the rotation of the rotating filter 410.

[0151] Furthermore, as shown in Figure 41, the outer filter 412 is provided with a B1 filter 412a that transmits light in the wavelength band of the second blue light BL among the white light, a B2 filter 412b that transmits light in the wavelength band of the first blue light BS among the white light, a G filter 412c that transmits light in the wavelength band of the green light G among the white light, an R filter 412d that transmits light in the wavelength band of the red light R among the white light, and a B3 filter 412e that transmits blue-green light BG among the white light. The blue-green light BG is light in the wavelength band that is transmitted when the corrective illumination light passes through the B color filter BF of the image sensor 43, which is a color image sensor (see Figure 11). Therefore, in observation mode or correction mode, illumination light having the wavelength bands of the first blue light BS, second blue light BL, green light G, red light R, and blue-green light BG is emitted from the light source device 13 in accordance with the rotation of the rotating filter 410.

[0152] In the endoscope system 10, in normal mode, a Bc image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength bands of violet light V and first blue light BS using a monochrome image sensor. A Gc image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of green light G using a monochrome image sensor. An Rc image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of red light R using a monochrome image sensor. Then, based on the Bc image signal, Gc image signal, and Rc image signal, a normal light image is generated in the same manner as in the first embodiment.

[0153] On the other hand, in observation mode or correction mode, a B1 image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of the first blue light BS using a monochrome image sensor. A G1 image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of green light G using a monochrome image sensor. A R1 image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of red light R using a monochrome image sensor. A B2 image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of the second blue light BL using a monochrome image sensor. A B3 image signal is output by capturing the reflected light obtained by illuminating the subject with illumination light having the wavelength band of blue-green light BG using a monochrome image sensor. Next, based on the B1, B2, G2, R2, and B3 image signals, an oxygen saturation image is generated in the same manner as in the first embodiment, and a correction process is performed to generate a corrected oxygen saturation image. However, in the second embodiment, instead of the signal ratio ln(B3 / G3), the signal ratio ln(B3 / G1), which is obtained by normalizing the B3 image signal with the G1 image signal, is used.

[0154] As a correction process for calculating corrected oxygen saturation in correction mode, it is also possible to select a table 86b for calculating corrected oxygen saturation according to the dye value, or to calculate corrected oxygen saturation by performing a calculation value correction process that adds or subtracts a correction value obtained from a specific calculation value to the oxygen saturation calculated by referring to the oxygen saturation calculation table 84a.

[0155] When performing the calculation value correction process, the correction value used for correcting oxygen saturation is calculated by referring to the two-dimensional coordinate system 430 shown in Figure 42. The vertical axis of the two-dimensional coordinate system 430 is a specific calculated value obtained based on the B1 image signal, G1 image signal, R1 image signal, B2 image signal, and B3 image signal, and the horizontal axis is ln(R1 / G1). The specific calculated value is determined by the following formula A). Formula A) B2 / G1×cosφ-B3 / G1×sinφ

[0156] In the two-dimensional coordinate system 430, a reference line 431a showing the distribution of predetermined reference baseline information and a measured line 431b showing the distribution of measured baseline information obtained by actually photographing the observed object are shown. The difference value ΔZ between the reference line 431a and the measured line 431b is calculated as a correction value. The reference baseline information is defined as information obtained without specific dyes and independent of oxygen saturation. Specifically, the reference baseline information is a value obtained by adjusting φ so that the above equation A) remains constant even when oxygen saturation changes.

[0157] [Third Embodiment] In the third embodiment, the endoscope 12 is a rigid endoscope equipped with a camera head 500 at the proximal end of the insertion section 12a, as shown in Figure 43. When the endoscope 12 is a laparoscope, the insertion section 12a is the part inserted into the abdominal cavity of the patient. The camera head 500 is equipped with an imaging optical system 40. In the first and second embodiments, the imaging optical system 40, which has an objective lens 41 and an imaging sensor 43, is provided at the tip of the endoscope 12, but in the third embodiment, the imaging sensor of the imaging optical system 40 is provided on the camera head 500 instead of the tip. The camera head 500 captures the reflected light guided from the tip of the endoscope 12. The image signal captured by the camera head 500 is transmitted to the processor device 14. Note that in Figure 43, the angle knob 12e, mode switching switch 12f, still image acquisition instruction switch 12h, and zoom operation unit 12i are omitted. The following describes the differences between the endoscope system 10 and the first and second embodiments, while the common parts will not be described.

[0158] In normal mode, the light source device 13 emits normal light including violet light V, first blue light BS, green light G, and red light R. In observation mode and correction mode, the light source device 13 emits illumination light (first mixed light), which is a mixed light including first blue light BS, second blue light BL, green light G, and red light R, as shown in Figure 44. Note that in Figure 44 and Figures 46, 47, 48, and 49 described later, the spectra of each illumination light are simplified by assuming that the light intensity of each illumination light is a constant value, but the light intensity of each illumination light is not limited to a constant value.

[0159] As shown in Figure 45, the camera head 500 comprises dichroic mirrors 501, 502, and 503, and image sensors 511, 512, 513, and 514, which are monochrome image sensors. The dichroic mirror 501 reflects light in the wavelength bands of violet light V and first blue light BS from the reflected light from the subject, and transmits light in the wavelength bands of second blue light BL, green light G, and red light R. The light reflected by the dichroic mirror 501 and incident on the image sensor 511 has the wavelength band of violet light V or first blue light BS, as shown in Figure 46. The image sensor 511 outputs a Bc image signal in normal mode and a B1 image signal in oxygen saturation or correction mode.

[0160] The dichroic mirror 502 reflects light in the wavelength band of the second blue light BL and transmits light in the wavelength bands of the green light G and red light R from the light that has passed through the dichroic mirror 501. The light reflected by the dichroic mirror 502 and incident on the imaging sensor 512 has the wavelength band of the second blue light BL, as shown in Figure 47. The imaging sensor 512 stops outputting the image signal in normal mode and outputs the B2 image signal in observation mode or correction mode.

[0161] The dichroic mirror 503 reflects light in the wavelength range of green light G and transmits light in the wavelength range of red light R from the light that has passed through the dichroic mirror 502. The light reflected by the dichroic mirror 503 and incident on the image sensor 513 has the wavelength range of green light G, as shown in Figure 48. The image sensor 513 outputs a Gc image signal in normal mode and a G1 image signal in oxygen saturation or correction mode.

[0162] The light transmitted through the dichroic mirror 503 and incident on the imaging sensor 514 has a wavelength band of red light R, as shown in Figure 49. The imaging sensor 514 outputs an Rc image signal in normal mode and an R1 image signal in observation mode or correction mode.

[0163] In other words, in the third embodiment, in normal mode, the Bc image signal, Gc image signal, and Rc image signal are output from the camera head, and in observation mode or correction mode, the B1 image signal, G1 image signal, R1 image signal, and B2 image signal are output from the camera head. The image signals output from the imaging sensors 511, 512, 513, and 514 are transmitted to the processor device 14.

[0164] [Fourth Embodiment] In the fourth embodiment, as in the third embodiment, the endoscope 12 is a rigid endoscope equipped with a camera head at the base end of the insertion portion 12a. The following describes the parts that differ from the first, second, and third embodiments, and the parts that are common will be omitted. In the fourth embodiment, the camera head 600 shown in Figure 50 is provided instead of the camera head 500 of the third embodiment.

[0165] As shown in Figure 50, the camera head 600 includes a dichroic mirror 601 and image sensors 611 and 612. The dichroic mirror 601 reflects light in the wavelength ranges of violet light V, first blue light BS, green light G, and red light R from the reflected light from the subject, and transmits light in the wavelength range of second blue light BL (see Figures 51(B) and 52(B) described later).

[0166] The image sensor 611, which receives light reflected by the dichroic mirror 601, is a color image sensor equipped with a B color filter BF for the B pixels, a G color filter GF for the G pixels, and an R color filter RF for the R pixels. The image sensor 612, which receives light transmitted through the dichroic mirror 601, is a monochrome image sensor.

[0167] In normal mode, white light is emitted from the light source device 13, and the reflected light from the subject reflected by the dichroic mirror 601 is received by the image sensor 611, which is a color image sensor. As a result, Bc image signals, Gc image signals, and Rc image signals are output from the image sensor 611, respectively. In normal mode, the image sensor 612, which is a monochrome image sensor, stops outputting image signals.

[0168] In observation mode, the light source device 13 emits illumination light (hereinafter referred to as the second mixed light) containing a first blue light BS, a second blue light BL, a green light G, and a red light R, as shown in Figure 51(A). The dichroic mirror 601 spectrally separates the reflected light from the subject illuminated by the second mixed light by reflecting and transmitting it. Figure 51(B) shows the relationship between the reflectance (dashed line 601a) and transmittance (solid line 601b) of light incident on the dichroic mirror 601 and the wavelength of the light. Note that in Figure 51(A) and Figures 52(A), 55(A), and 55(A) described later, for simplicity, the spectra of each illumination light are shown with the light intensity of each color of illumination light assumed to be constant, but the light intensity of each color of illumination light is not limited to constant.

[0169] Of the light reflected from a subject illuminated by the second mixed light, the light reflected by the dichroic mirror 601 is received by the image sensor 611, which is a color image sensor. The sensitivity of the B pixel B, G pixel G, and R pixel R of the image sensor 611 and the wavelength of light have the relationship shown in Figure 51(C). Therefore, the B pixel B of the image sensor 611 outputs a B1 image signal by sensing the light in the wavelength band B1 of the first blue light BS. The G pixel G of the image sensor 611 outputs a G1 image signal by sensing the light in the wavelength band G1 of the green light G. Furthermore, the R pixel R of the image sensor 611 outputs an R1 image signal by sensing the light in the wavelength band R1 of the red light R.

[0170] Meanwhile, of the reflected light from the subject illuminated by the second mixed light, the light transmitted through the dichroic mirror 601 is received by the image sensor 612, which is a monochrome image sensor. The sensitivity of the image sensor 612 and the wavelength of light have the relationship shown in Figure 52(C). Therefore, the image sensor 612 outputs a B2 image signal by sensing the light in wavelength band B2 of the second blue light BL that is transmitted through the dichroic mirror 601. Figure 52(A), like Figure 51(A), shows the wavelength band of light contained in the second mixed light. Also, Figure 52(B), like Figure 51(B), shows the relationship between the reflectance (dashed line 601a) and transmittance (solid line 601b) of the light incident on the dichroic mirror 601 and the wavelength of light.

[0171] In the observation mode of the fourth embodiment, as shown in Figure 53, the emission pattern in which the second mixed light ML is emitted is repeated once per frame F. Therefore, in the observation mode of the fourth embodiment, for each frame, the B1 image signal, G1 image signal, and R1 image signal are output from the image sensor 611, which is a color image sensor, and the B2 image signal is output from the image sensor 612, which is a monochrome image sensor. The B1 image signal, G1 image signal, R1 image signal, and B2 image signal output from the image sensor 611 or the image sensor 612 in observation mode are transmitted to the processor device 14. The method for calculating oxygen saturation in observation mode is the same as in the first embodiment.

[0172] In the correction mode of the fourth embodiment, as shown in Figure 54, when the second mixed light ML is being emitted, if a light emission switching instruction is input by user operation, the emission pattern is such that there is a period of no light emission NL for 2 frames F, followed by the emission of the correction illumination light CL for 2 frames F, and then, after several more frames of no light emission NL, the second mixed light ML is emitted for 2 frames F. The frames of the no-light emission NL are a period for switching between the second mixed light ML and the correction illumination light, and neither illumination light is emitted during this period. The light emission switching instruction may be input by operating an illumination light switching switch (not shown) provided on the endoscope 12 or the user interface, or by toggling the mode switching switch 12f.

[0173] In correction mode, in frames where the second mixed light ML is emitted, the B1 image signal, G1 image signal, and R1 image signal are output from the imaging sensor 611, and the B2 image signal is output from the imaging sensor 612, just as in observation mode.

[0174] In correction mode, in frames where the correction illumination light CL is emitted, the light source device 13 emits correction illumination light (correction illumination light) including green light G, as shown in Figure 55(A). In addition, the image sensor 611, which is a color image sensor, receives reflected light from the subject reflected by the dichroic mirror 601 (see Figure 55(B)). As a result, as shown in Figure 55(C), the B pixel B of the image sensor 611 outputs a B3 image signal by sensing the light in the wavelength band B3 of the green light G. Furthermore, as shown in Figure 56(C), the G pixel G of the image sensor 611 outputs a G3 image signal by sensing the light in the wavelength band G3 of the green light G. Also, the R pixel R of the image sensor 611 outputs an R3 image signal (not shown) by sensing the light in the wavelength band of the green light G.

[0175] Figure 56(A), like Figure 55(A), shows the wavelength range of light contained in the corrective illumination light. Also, Figures 55(B) and 56(B), like Figure 51(B), show the relationship between the reflectance (dashed line 601a) and transmittance (solid line 601b) of light incident on the dichroic mirror 601 and the wavelength of light. Furthermore, Figures 55(C) and 56(C) show Figure 5 Similar to 1(C), this shows the relationship between the sensitivity of the B pixels (B), G pixels (G), and R pixels (R) of the image sensor 611 and the wavelength of light.

[0176] In correction mode, the B1 image signal, G1 image signal, R1 image signal, B2 image signal, B3 image signal, and G3 image signal output from the imaging sensor 611 or imaging sensor 612 are transmitted to the processor device 14. The calculation of reliability in the fourth embodiment will be described below.

[0177] In the first embodiment, an example was shown in which a correction image 200 is displayed and the confidence level is calculated for each pixel included in a specific correction region 201 contained in the correction image 200. In the fourth embodiment, unlike the first embodiment, the confidence level is calculated for each correction region using the image obtained in the frame in which the second mixed light ML is emitted (white light equivalent image and second blue light image) and the image obtained in the frame in which the correction illumination light CL is emitted (correction illumination light image). The correction region corresponds to the specific correction region in the first embodiment. Furthermore, the term "correction region" is used to refer to "a set of sub-regions divided into multiple parts" or "the sub-region itself (the Nth correction region, where N is a natural number of 1 or more)," as will be described later.

[0178] The white light equivalent image is an endoscopic image generated using the B1, G1, and R1 image signals output in the frame in which the second mixed light ML is emitted. The second blue light image is an endoscopic image generated using the B2 image signal output in the frame in which the second mixed light ML is emitted. The corrective illumination light image is an endoscopic image generated using the B3, G3, and R3 image signals output in the frame in which the corrective illumination light CL is emitted. The white light equivalent image, the second blue light image, and the corrective illumination light image are generated by the DSP61.

[0179] The white light equivalent image, the second blue light image, and the correction illumination light image are transmitted to the feature calculation unit 620 of the processor device 14, as shown in Figure 57. The feature calculation unit 620 is preferably configured by a different processor than the central control unit 50 in the processor device 14. For example, the feature calculation unit 620 is preferably configured by an FPGA (Field Programmable Gate Array).

[0180] The feature calculation unit 620 calculates region features for each of the multiple correction regions shown in Figure 58 from the white light equivalent image, the second blue light image, and the correction illumination light image. Region features will be described later. The correction region 622 is a small region divided into multiple regions in the white light equivalent image 621, as shown in Figure 58. In the example shown in Figure 58, the correction region 622 is a region obtained by dividing the horizontal length a of the white light equivalent image 621 into a1, a2, and a3, and the vertical length b into b1, b2, and b3, and then further dividing the region where the column of a2 and the row of b2 intersect into 16 equal parts (see Figure 58(A)). The position of the region to be used as the correction region 622 and the number of divisions of the region to be used as the correction region 622 are not limited to these. For example, the correction region 622 may be divided into 9 equal parts or into 25 equal parts.

[0181] Figure 58(B) is an enlarged view of the correction region 622 shown in Figure 58(A). As shown in Figure 58(B), the correction region 622 is divided into 16 regions, from the first correction region 622a to the 16th correction region 622p. In Figure 58(B), the division of the correction region 622 into 16 regions, from the first correction region 622a to the 16th correction region 622p, is indicated by the numbers 1 through 16.

[0182] The feature calculation unit 620 determines whether each pixel constituting the correction region of the Nth correction region (in the example shown in Figure 58, from the 1st correction region to the 16th correction region) is a valid pixel. The determination of whether a pixel is a valid pixel is made by setting lower and upper channel threshold values ​​for each channel (B channel, G channel, R channel) of each pixel.

[0183] For B channels, a lower threshold and an upper threshold are set. For G channels, a lower threshold and an upper threshold are set. For R channels, a lower threshold and an upper threshold are set.

[0184] The feature calculation unit 620 determines a pixel to be an effective pixel if, for the white light equivalent image, the second blue light image, and the corrective illumination light image, the pixel values ​​of all color channels for each pixel in each correction region are within the range of greater than or equal to the lower channel threshold and less than the upper channel threshold for each color.

[0185] For white light equivalent images and corrective illumination light images, a pixel is determined to be an effective pixel if the pixel value of the B channel of each pixel constituting the white light equivalent image and corrective illumination light image is within the range of B channel lower threshold to B channel upper threshold, G channel pixel value is within the range of G channel lower threshold to G channel upper threshold, and R channel pixel value is within the range of R channel lower threshold to R channel upper threshold.

[0186] For the second blue light image, a pixel is determined to be an effective pixel if its pixel value is within the range of the lower limit of the monochrome image channel and less than the upper limit of the monochrome image channel.

[0187] Next, the feature calculation unit 620 calculates region features for each correction region from the first correction region to the sixteenth correction region. Region features include the number of effective pixels, the sum of the pixel values ​​of the effective pixels, the sum of the squares of the pixel values ​​of the effective pixels, and the variance of the pixel values ​​of the effective pixels.

[0188] Specifically, the feature calculation unit 620 calculates regional features for each correction region of each channel in the white light equivalent image. It also calculates regional features for each correction region of each channel in the correction illumination light image. Furthermore, it calculates regional features for each correction region of the second blue light image. The regional features of each correction region of each channel in each endoscopic image calculated by the feature calculation unit 620 are transmitted to the reliability calculation unit 220 of the tissue oxygen saturation image generation unit 80.

[0189] In the fourth embodiment, the reliability calculation unit 220 calculates a reliability score for determining the degree of influence of disturbances in the correction region. The reliability calculation unit 220 also calculates a dye value for determining the degree of movement of the endoscope 12. The degree of movement of the endoscope 12 is the degree to which the endoscope 12 was moved during the switching of illumination light in the correction mode of the fourth embodiment (i.e., the non-illuminating state NL). If the endoscope 12 moves in the non-illuminating state NL, the object of observation shown in the endoscope image also moves, which may prevent the correction process from being performed properly. Therefore, by calculating the degree of movement of the endoscope 12 in the correction region, as described later, a determination regarding the movement of the endoscope 12 can be made based on the degree of movement of the endoscope 12, and if the degree of movement of the endoscope 12 is large, the user can be notified not to move the endoscope 12.

[0190] Furthermore, in the fourth embodiment, the correction determination unit 230 of the tissue oxygen saturation image generation unit 80 determines the degree of influence of disturbances using the reliability and / or determines the degree of movement of the endoscope 12 using the dye value.

[0191] In the fourth embodiment, a method for determining the degree of influence of disturbances and a method for determining the degree of movement of the endoscope 12 will be described below. In the fourth embodiment, as shown in Figure 59, the reliability calculation unit 220 has a region reliability calculation unit 630 and a dye value calculation unit 650. The correction determination unit 230 has a region reliability determination unit 640 and a dye value determination unit 660.

[0192] The region reliability calculation unit 630 of the reliability calculation unit 220 calculates region reliability using region features of each correction region of each channel of the white light equivalent image, second blue light image, and correction illumination light image generated from the image signal output in each frame. Region reliability includes the average value of pixel values ​​in the correction region, the standard deviation of pixel values ​​in the correction region, the effective pixel ratio in the correction region, the reliability of brightness in the correction region, the reliability of the degree of bleeding included in the correction region, the reliability of the degree of fat included in the correction region, etc. Region reliability is one form of "reliability" in the first embodiment.

[0193] The average value of pixel values ​​within the correction area is calculated using the number of pixels within the correction area and the pixel values ​​of the effective pixels within the correction area. The standard deviation of pixel values ​​within the correction area is calculated using the number of pixels within the correction area and the variance of the pixel values ​​of the effective pixels. The effective pixel ratio within the correction area is calculated using the number of pixels within the correction area and the number of effective pixels. The confidence level for brightness within the correction area is calculated by applying the average value of the G1 image signal within the correction area (i.e., the pixel values ​​within the correction area of ​​the G channel of the white light equivalent image) to the first confidence calculation table 221 (see Figure 33), as shown in Figure 60, where the horizontal axis is the signal value of the G1 image signal. It is preferable that the average value of the G1 image signal be the average value of the luminance values ​​obtained by performing a conversion process using the G1 image signal.

[0194] The reliability based on the degree of bleeding within the correction region is calculated by using the average values ​​of the B1, G1, and R1 image signals within each correction region of the white light equivalent image (i.e., the average of the pixel values ​​within each correction region for each color channel of the white light equivalent image) to determine the region-average signal ratio ln(R1 / G1) and region-average signal ratio ln(B1 / G1). These signal ratios are then applied to the second reliability calculation table 222 (see Figure 34).

[0195] The reliability based on the degree of fat contained within the correction region is calculated by using the average values ​​of the G1 image signal and R1 image signal within each correction region of the white light equivalent image (i.e., the average of the pixel values ​​in each correction region of the G channel and R channel of the white light equivalent image) and the average value of the B2 image signal within each correction region of the second blue light image to calculate the region average signal ratio ln(R1 / G1) and region average signal ratio ln(B2 / G1). These signal ratios are then applied to the third reliability calculation table 223 (see Figure 35), as shown in Figure 61, with the vertical axis set to signal ratio ln(B2 / G1). The region reliability calculated by the region reliability calculation unit 630 is transmitted to the region reliability determination unit 640 of the correction determination unit 230 (see Figure 59).

[0196] The following describes a method for determining and reporting the degree of influence of disturbances in the fourth embodiment. The region reliability determination unit 640 uses a preset region reliability determination threshold to output a determination result for each correction region in the white light equivalent image, the second blue light image, and the correction illumination light image, indicating whether it is a high-reliability correction region or a low-reliability correction region.

[0197] The threshold for determining regional confidence may be set according to the type of regional confidence. For example, a first regional confidence determination threshold can be set for the "average value of pixel values ​​within the correction region," and if the "average value of pixel values ​​within the correction region" is equal to or greater than the first regional confidence determination threshold, the correction region is determined to be a "high-confidence correction region." On the other hand, if the "average value of pixel values ​​within the correction region" is less than the first regional confidence determination threshold, the correction region is determined to be a "low-confidence correction region."

[0198] Similarly, a second-domain confidence threshold is set for the "standard deviation of pixel values ​​within the correction region," a third-domain confidence threshold for the "effective pixel ratio within the correction region," a fourth-domain confidence threshold for the "confidence regarding brightness within the correction region," and a fifth-domain confidence threshold for the "confidence regarding the degree of fat contained within the correction region," and the judgment results are output.

[0199] The region reliability determination unit 640 uses the determination result for each correction region—whether it is a high-reliability correction region or a low-reliability correction region—to further determine the reliability of the white light equivalent image, the second blue light image, and the correction illumination light image. In this case, the image determination result is output according to the number of correction regions determined to be "low-reliability correction regions" among all correction regions in the white light equivalent image, the second blue light image, and the correction illumination light image. The image determination result is output, for example, by pre-setting a threshold for outputting the first image determination result.

[0200] For example, if the threshold for outputting the first image judgment result is set to "10" and the number of correction regions for the white light equivalent image is set to 16, then if there are 10 or more low-reliability correction regions in the white light equivalent image, the system outputs an image judgment result indicating that the overall reliability of the correction regions is high, meaning that there is little influence from disturbances and the correction process can be performed appropriately. On the other hand, if there are fewer than 10 low-reliability correction regions in the white light equivalent image, the system outputs an image judgment result indicating that the overall reliability of the correction regions is low, meaning that there is some influence from disturbances and the correction process cannot be performed appropriately.

[0201] Furthermore, the calculation of regional confidence and the output of image judgment results may be performed for all white light equivalent images, second blue light images, and corrective illumination images, or for some of the white light equivalent images, second blue light images, and corrective illumination images. This is to speed up the calculation process.

[0202] The image determination result output by the region reliability determination unit 640 is transmitted to the display control unit 100. The display control unit 100 preferably changes the display mode on the display 15 according to the image determination result. For example, if the image determination result is "low reliability for the entire correction region", the display control unit 100 displays a message on the display 15 indicating that the correction process can be performed appropriately (see Figure 38). On the other hand, if the image determination result is "high reliability for the entire correction region", a message such as "Please operate the endoscope for correction processing" is displayed on the display 15 as a warning (see Figure 39). Such a message may be superimposed on the white light equivalent image displayed on the display.

[0203] Furthermore, the region reliability determination unit 640 may calculate the average reliability for image determination using each correction region in the white light equivalent image, the second blue light image, and the correction illumination light image. The average reliability for image determination is calculated, for example, by dividing the sum of the reliability values ​​of all correction regions in the white light equivalent image by the number of correction regions. In this case, the region reliability determination unit 640 sets a threshold for outputting a second image determination result for the average reliability for image determination, and outputs an image determination result of "high reliability for the correction region as a whole" if the average reliability for image determination is equal to or greater than the threshold for outputting a second image determination result. On the other hand, if the average reliability for image determination is less than the threshold for outputting a second image determination result, outputs an image determination result of "low reliability for the correction region as a whole". In this case as well, it is preferable that the display control unit 100 changes the display mode on the display 15 according to the image determination result.

[0204] The following describes a method for determining and reporting the degree of movement of the endoscope 12 in the fourth embodiment. In this case, the determination result of whether each correction region in the white light equivalent image, the second blue light image, and the correction illumination light image output by the region reliability determination unit 640 is a high-reliability correction region or a low-reliability correction region is transmitted to the dye value calculation unit 650 (see Figure 59).

[0205] It is preferable that the pigment value calculation unit 650 performs an exclusion process to exclude from the calculation of pigment values ​​correction regions that have been determined to be "low-reliability correction regions" among the correction regions in the white light equivalent image, the second blue light image, and the correction illumination light image.

[0206] Furthermore, it is preferable that the pigment value calculation unit 650 excludes a portion of the white light equivalent image, the second blue light image, and the corrective illumination light image from the light emission pattern in the correction mode of the fourth embodiment (see Figure 54). Specifically, for example, as shown in Figure 62, the excluded images are the white light equivalent image 652a and the second blue light image 652b generated based on the image signal output in frame 651b, the corrective illumination light image 652c generated based on the image signal output in frame 651c, the corrective illumination light image 653c generated based on the image signal output in frame 651d, and the white light equivalent image 653a and the second blue light image 653b generated based on the image signal output in frame 651e.

[0207] Of these, the white light equivalent image 652a, the second blue light image 652b, and the corrective illumination light image 652c are referred to as the first image set 652d. The white light equivalent image 653a, the second blue light image 653b, and the corrective illumination light image 653c are referred to as the second image set 653d. The dye value calculation unit 650 preferably performs exclusion processing on the images included in the first image set 652d and the images included in the second image set 653d. Hereafter, the correction area determined to be a "high-reliability correction area" that is not subject to exclusion processing will be referred to as the effective area. On the other hand, the correction area determined to be a "low-reliability correction area" that is subject to exclusion processing will be referred to as the excluded area.

[0208] The pigment value calculation unit 650 performs exclusion processing so that the positions of each effective region included in the first image set 652d and the positions of each excluded region correspond between the white light equivalent image 652a, the second blue light image 652b, and the corrective illumination light image 652c.

[0209] Specifically, for example, as shown in Figure 63, the excluded areas of the white light equivalent image 652a are designated as correction areas 654d and 654h, and the correction areas of the entire correction area 654 other than correction areas 654d and 654h are designated as effective areas. Similarly, the excluded areas of the second blue light image 652b are designated as correction areas 655d and 655h, and the correction areas of the entire correction area 655 other than correction areas 655d and 655h are designated as effective areas. In addition, the excluded areas of the corrective illumination light image 652c are designated as correction areas 656d and 656h, and the correction areas of the entire correction area 656 other than correction areas 656d and 656h are designated as effective areas. By performing such exclusion processing, the positions of the effective areas can be made to correspond between the images included in the first image set 652d.

[0210] The exclusion process method will now be explained. The pigment value calculation unit 650 performs the exclusion process for each image set using a pre-set exclusion threshold. The exclusion threshold is set as multiple values ​​so that the region confidence of each correction region can be evaluated and calculated on a 5-point scale from "1" to "5". It is preferable to set the exclusion threshold according to the type of region confidence.

[0211] The pigment value calculation unit 650 first calculates a five-level confidence level for region determination for each correction region in the white light equivalent image, the second blue light image, and the correction illumination image included in the image set. Next, it selects the correction region with the lowest region determination confidence level from among the corresponding correction regions in the white light equivalent image, the second blue light image, and the correction illumination image included in the image set.

[0212] Next, a threshold for determining whether a correction region is a "high-reliability correction region" or a "low-reliability correction region" is applied to the correction region with the lowest region determination reliability. Correction regions determined to be "low-reliability correction regions" are excluded. In this case, all correction regions in the white light equivalent image, the second blue light image, and the correction illumination light image corresponding to the correction region determined to be a "low-reliability correction region" are excluded.

[0213] The pigment value calculation unit 650 calculates pigment values ​​from the first image set 652d and the second image set 653d, respectively. The calculation of pigment values ​​will be explained in detail below. When calculating the pigment values ​​of the first image set 652d, the region average signal ratio ln(R1 / G1), region average signal ratio ln(B2 / G1), and region average signal ratio ln(B3 / G3) are calculated for each effective region based on the signal values ​​in each effective region at corresponding positions in the white light equivalent image 652a, the second blue light image 652b, and the corrective illumination light image 652c.

[0214] The region-averaged signal ratio ln(R1 / G1) is calculated using the average value of the R1 image signal within each effective region of the white light equivalent image 652a and the average value of the G1 image signal within each effective region (i.e., the average of the pixel values ​​within each effective region of the R channel and the average of the pixel values ​​within each effective region of the G channel of the white light equivalent image).

[0215] The region-averaged signal ratio ln(B2 / G1) is calculated using the average value of the B1 image signal within each effective region of the second blue light image 652b (i.e., the average of the pixel values ​​within each effective region of the second blue light image) and the average value of the G1 image signal within each effective region of the white light equivalent image 652a.

[0216] The region-averaged signal ratio ln(B3 / G3) is calculated using the average value of the B3 image signal within each effective region of the corrective illumination light image 652c and the average value of the G3 image signal within each effective region (i.e., the average of the pixel values ​​within each effective region of the B channel and the average of the pixel values ​​within each effective region of the G channel of the corrective illumination light image).

[0217] The pigment value calculation unit 650 calculates the region-average signal ratio ln(R1 / G1), region-average signal ratio ln(B2 / G1), and region-average signal ratio ln(B3 / G3), which are calculated for each corresponding effective region of the first image set 652d, by referring to the pigment value calculation table 86a (see Figure 17). In the pigment value calculation table 86a, surfaces CV0 to CV4 are distributed according to the concentration of yellow pigment in a three-dimensional coordinate system with the signal ratio ln(R1 / G1) on the X axis, the signal ratio ln(B2 / G1) on the Y axis, and the signal ratio ln(B3 / G3) on the Z axis.

[0218] The pigment value calculation unit 650 refers to the pigment value calculation table 86a, which is a three-dimensional coordinate system, and calculates the pigment value for the surface CV0 to CV4 where the coordinates (X3, Y3, Z3) = (region average signal ratio ln(R1 / G1), region average signal ratio ln(B2 / G1), region average signal ratio ln(B3 / G3)) overlap or are closest in distance. Surfaces CV0 to CV4 each indicate a pigment value from "0" to "4". For example, if the coordinates (X3, Y3, Z3) overlap with surface CV2, the pigment value is calculated to be "2". The pigment value calculation unit 650 calculates the pigment value for each effective region of the first image set 652d. Similarly, the pigment value for the second image set 653d is also calculated for each effective region.

[0219] The pigment values ​​calculated for each effective region of the first image set 652d and the pigment values ​​calculated for each effective region of the second image set 653d are transmitted to the pigment value determination unit 660 of the correction determination unit 230. It is also preferable to transmit to the pigment value determination unit 660 the X component values, Y component values, and Z component values ​​of the coordinates (X3, Y3, Z3) calculated for each effective region of the first image set 652d and the X component values, Y component values, and Z component values ​​of the coordinates (X3, Y3, Z3) calculated for each effective region of the second image set 653d.

[0220] As explained in Figure 64, the pigment value determination unit 660 calculates a correlation coefficient 663 between the pigment value 661 calculated for each effective region of the first image set 652d and the pigment value 662 calculated for each effective region of the second image set 653d. In Figure 64, for illustrative purposes, the pigment values ​​661 and 662 calculated for each effective region are shown, with the vertical axis representing the pigment value and the horizontal axis representing the number of the correction region assigned to the effective region (i.e., the "N" in the Nth correction region).

[0221] The dye value determination unit 660 determines that "the degree of endoscope movement is large" if the correlation coefficient is smaller than a preset threshold for motion determination. On the other hand, if the correlation coefficient is larger than the threshold for motion determination, it determines that "the degree of endoscope movement is small". In this case, the dye value determination unit 660 outputs the determination result of "the degree of endoscope movement is large" or "the degree of endoscope movement is small" as the motion determination result and transmits it to the display control unit 100.

[0222] The display control unit 100 preferably changes the display mode on the display 15 according to the motion detection result. For example, if the motion detection result is "the degree of endoscope movement is small," the display control unit 100 displays a message on the display 15 indicating that the correction process can be performed appropriately (see Figure 38). On the other hand, if the motion detection result is "the degree of endoscope movement is large," a message MS4 such as "Please keep the endoscope still for correction processing," as shown in Figure 65, is displayed on the display 15 as a warning.

[0223] When switching illumination light by manually inputting a light emission switching instruction, the movement of the endoscope 12 may become excessive during the switching process. In such cases, it may not be possible to properly perform the correction processing according to the influence of the concentration of the specific dye. Therefore, by determining the degree of movement of the endoscope 12 and notifying the user if the degree of movement is excessive, the user can be encouraged to refrain from moving the endoscope 12. As a result, if the degree of movement of the endoscope 12 is reduced, the correction processing can be performed appropriately.

[0224] As described above, by notifying the image judgment result that determines the degree of the disturbance's influence and / or the motion judgment result that determines the degree of movement of the endoscope 12, the user can be prompted to perform operations to appropriately carry out the correction process. In the correction process, in order to perform table correction processing according to the dye value, it is preferable to perform correction processing by determining the dye value by a robust estimation method based on the dye value for each correction area calculated using the first image set and the second image set, and selecting the correction oxygen saturation calculation table 86b corresponding to the dye value. Furthermore, if the motion judgment result is "the degree of movement of the endoscope is small" and the image judgment result is "high reliability for the entire correction area", the correction process may be performed by determining the dye value using the area-average signal ratio in the correction area determined to be a "high-reliability correction area".

[0225] In the above embodiment, the hardware structure of the processing unit that performs various processes such as the central control unit 50, image signal acquisition unit 60, DSP 61, noise reduction unit 62, image processing switching unit 63, normal light image generation unit 70, tissue oxygen saturation image generation unit 80, display control unit 100, and exposure control unit 110 is the following type of processor. The types of processors include general-purpose processors that execute software (programs) and function as various processing units, such as CPUs (Central Processing Units), GPUs (Graphical Processing Units), programmable logic devices (PLDs) whose circuit configuration can be changed after manufacturing, such as FPGAs, and dedicated electrical circuits which are processors with circuit configurations specifically designed to perform various processes.

[0226] A single processing unit may be composed of one of these various processors, or it may be composed of a combination of two or more processors of the same or different types (for example, multiple FPGAs, a combination of a CPU and an FPGA, or a combination of a CPU and a GPU). Alternatively, multiple processing units may be composed of a single processor. Examples of composing multiple processing units with a single processor include, firstly, a configuration where one or more CPUs and software are combined to form a single processor, and this processor functions as multiple processing units, as is typical of computers such as clients and servers. Secondly, a configuration where a processor is used that realizes the functions of the entire system, including multiple processing units, on a single IC (Integrated Circuit) chip, as is typical of a System on a Chip (SoC). Thus, various processing units are configured, in terms of hardware structure, using one or more of the above-mentioned various processors.

[0227] Furthermore, the hardware structure of these various processors is, more specifically, an electrical circuit formed by combining circuit elements such as semiconductor devices. The hardware structure of the memory unit is a storage device such as an HDD (hard disk drive) or SSD (solid state drive). [Explanation of Symbols]

[0228] 10 Endoscopy Systems 12 Endoscopes 12a Insertion section 12b Operation section 12c curved section 12d Tip 12e Angle Knob 12f Mode Selector Switch 12h Still image acquisition instruction switch 12i Zoom Control Section 12j forceps mouth 13 Light source device 14 Processor Unit 15 displays 16 User Interface 20 Light source section 20a V-LED 20b BS-LED 20c BL-ELD 20d G-LED 20e R-LED 21 Light source control unit 22 Optical path coupling section 23 Light Guide 30 Illumination optical system 31 Illumination Lens 40 Imaging Systems 41 Objective lens 42 Zoom Lens 43, 511, 512, 513, 514, 611, 612 Image sensors 44 Imaging Control Unit 45 CDS / AGC circuit 46 A / D converters 47 aperture 48 shutters 50 Central control unit 60 Image signal acquisition unit 61 DSP 62 Noise reduction unit 63 Image processing switching unit 70 Normal light image generation unit 80 Tissue oxygen saturation image generation unit 81 White light equivalent image generation unit 82 Base image generation unit 83 Signal ratio calculation unit 84 Oxygen saturation image calculation unit 84a Oxygen saturation calculation table 86b Corrected oxygen saturation calculation table 84b, 84c Contour lines 84d, 84e Absorption coefficient graphs 85 Oxygen saturation image generation unit 86 Corrected oxygen saturation calculation unit 86a Pigment value calculation table 86c, 86d, 86e, 86f, 86g Surfaces 87 Corrected oxygen saturation image generation unit 100 Display control unit 111 Observation image signal 110 Exposure control unit 120 Exposure control signal generation unit 130 Control amount output unit 140 Brightness calculation unit 141 Analysis area setting unit 141a Endoscope image 141b, 141c Areas of the endoscope image 141d, 141e, 141f Areas of the analysis image 142 Area brightness calculation unit 150 Brightness difference calculation unit 151 First analysis image 151a First area 152 Observation image signal 153 Second analysis image 153a Second area 154 Correction image signal 200 Correction image 201 Specific correction area 201a Low confidence region 201b High confidence region 202 slots 210 Correction Image Generation Unit 220 Confidence Calculation Unit 221 Table for calculating the first confidence level 222 Second Confidence Level Calculation Table 223 Table for calculating the third confidence level 230 Correction judgment section 400 broadband light sources 410 Rotation Filter 411 Internal filter 411a, 412a B1 filter 411b, 412c G filters 411c, 412d R filters 412 External filter 412b B2 filter 412e B3 filter 420 Filter switching section 430 2D coordinate system 431a Reference line 431b Measured line 500, 600 camera heads 501, 502, 503, 601 Dichroic Mirror 620 Feature Calculation Unit 621, 652a, 653a White light equivalent image 622, 654, 654d, 654h, 655, 655d, 655h, 656, 656d, 656h Correction area 622a 1st correction area 622p 16th correction area 630 Area Reliability Calculation Unit 640 Area Confidence Determination Unit 650 Pigment Value Calculation Unit 601a Reflectance 601b Transmittance 651a, 651b, 651c, 651d, 651e, 651f frames 652b, 653b Second blue light image 652c, 653c Correction illumination light image 652d First image set 653d Second image set 660 Pigment value determination unit 661, 662 Pigment values 663 Correlation coefficient

Claims

1. In an endoscope system that illuminates an object to be observed and captures reflected light from the object to be observed, A light source device that emits light from at least one light source and emits observation illumination light and correction illumination light having a different spectrum from the observation illumination light to the object being observed, An imaging optical system for imaging the reflected light, Equipped with a processor, The aforementioned processor, During the first exposure period in which the observation illumination light is emitted and the second exposure period in which the correction illumination light is emitted, multiple different types of image signals are acquired. The image signal brightness is calculated from a subset of the aforementioned image signals that are different from each other. The control quantity that varies according to the brightness of the aforementioned image signal is output, Multiple types of exposure control signals with different control amounts are generated, An endoscope system that controls the light source and the imaging optical system in accordance with the exposure control signal.

2. The endoscope system according to claim 1, wherein the control amount is output based on the difference between the brightness of the image signal and a preset target brightness.

3. The aforementioned processor, During the first exposure period, an observation image signal is acquired as the image signal. During the second exposure period, a correction image signal is acquired as the image signal. Based on the difference between the first image signal brightness, which is the image signal brightness calculated using the aforementioned observation image signal, and the first target brightness, which is the target brightness, a first control quantity is output as the control quantity. Based on the difference between the second image signal brightness, which is the image signal brightness calculated using the correction image signal, and the second target brightness, which is the image signal brightness, a second control amount different from the first control amount is output. A first exposure control signal is generated as the exposure control signal based on the first control amount. The endoscope system according to claim 2, which generates a second exposure control signal as the exposure control signal based on the second control amount.

4. The endoscopic system according to claim 3, wherein the second control amount is greater than the first control amount.

5. The aforementioned processor, The endoscope system according to claim 3 or 4, wherein the brightness of the image signal is calculated using a specific color signal from the aforementioned image signal.

6. The endoscope system according to claim 5, wherein the specified color signal is a B signal.

7. The aforementioned processor, An analysis image is generated using the aforementioned image signal. A specific region is extracted from the aforementioned analyzed image, The endoscope system according to claim 3 or 4, which calculates the brightness of the image signal using the image signal in the specified region.

8. The endoscope system according to claim 7, wherein the aforementioned specific region is extracted according to a specific shape.

9. The aforementioned processor, From the first analysis image, which is the analysis image generated using the aforementioned observation image signal, the first region, which is the specific region, is extracted. Based on the difference between the first region brightness, which is the image signal brightness calculated using the first region, and the first target brightness, the first region control amount is output as the control amount. From the second analysis image, which is the analysis image generated using the correction image signal, the second region, which is the specific region, is extracted. The second region control amount is output as the control amount based on the difference between the second region brightness, which is the image signal brightness calculated using the second region, and the second target brightness. A first region exposure control signal is generated as the exposure control signal based on the first region control amount. The endoscope system according to claim 8, which generates a second region exposure control signal as the exposure control signal based on the second region control amount.

10. The aforementioned processor, The system switches between an observation mode for acquiring the aforementioned observation image signal and a correction mode for acquiring the aforementioned observation image signal and the aforementioned correction image signal. The first exposure control signal is generated in the observation mode. The endoscope system according to claim 3, wherein the first exposure control signal and the second exposure control signal are generated in the correction mode.

11. The first exposure period includes a first illumination period in which white equivalent light included in the observation illumination light is emitted, and a second illumination period in which calculation illumination light included in the observation illumination light is emitted. The second exposure period includes a third illumination period for acquiring the corrective illumination light, The aforementioned processor, The first illumination period, the second illumination period, and the third illumination period are automatically switched. During the first illumination period, an image signal equivalent to white light is acquired as the image signal. During the second illumination period, a calculation image signal is acquired as the image signal. During the third illumination period, the correction image signal is acquired. Based on the white light equivalent image signal, a first A exposure control signal is generated as the exposure control signal. Based on the calculation image signal, a first B exposure control signal is generated as the exposure control signal. Based on the correction image signal, the second exposure control signal is generated. The endoscope system according to claim 3, wherein the light source and the imaging optical system are controlled in accordance with the first A exposure control signal, the first B exposure control signal, and the second exposure control signal.

12. The control of the light source includes the control of the light intensity. The endoscope system according to claim 1, wherein the control of the imaging optical system includes control of the aperture value, control of the exposure time, and control of the gain.

13. An endoscope system operating method for illuminating an object to be observed and imaging reflected light from the object to be observed comprises a light source device, an imaging optical system, and a processor. The light source device emits at least one or more light sources and emits observation illumination light and correction illumination light having a different spectrum from the observation illumination light onto the object to be observed. The imaging optical system includes the step of imaging the reflected light, The aforementioned processor, The steps include acquiring multiple types of image signals that are different from each other during a first exposure period in which the observation illumination light is emitted and a second exposure period in which the correction illumination light is emitted, A step of calculating the brightness of an image signal from a subset of the image signals that are different from each other, The steps include outputting a control amount that varies according to the brightness of the image signal, The steps include generating multiple types of exposure control signals with mutually different control amounts, A method for operating an endoscope system, comprising the step of controlling the light source and the imaging optical system in accordance with the exposure control signal.

Images