JP6929662B2 - Imaging equipment, imaging methods and programs - Google Patents

Description

The present invention relates to an imaging device, an imaging method and a program for obtaining an image of the subject by scanning light on the subject.

Currently, as a method for non-invasively observing the internal structure of the eye, an optical coherence tomography imager (hereinafter referred to as OCT device) using an optical coherence tomography method (hereinafter referred to as OCT) that applies an optical interference phenomenon. ) Is used. In the OCT apparatus, the light emitted from the low-infrared coherent light source is divided into the measurement light and the reference light. Then, by interfering the scattered reflected light of the measurement light irradiating the eye with the reference light, high-resolution and high-sensitivity tomographic information is obtained in the depth direction (measurement optical axis direction).

Conventionally, as a scanning mode of the measurement light of the OCT device in the eye, particularly the fundus, a raster scan in which the scanning of the measurement light in the X direction is repeated while shifting in the Y direction perpendicular to the X direction has been used. On the other hand, there is a technique for obtaining tomographic information of the object to be inspected three-dimensionally by scanning the object to be inspected in a plane so that the measurement light draws a Lissajous figure (hereinafter referred to as Lissajous scan). (See Patent Document 1 and Non-Patent Document 1).

In the case of raster scan, there is a difference in the acquisition time of tomographic information between, for example, the first scanning position and the last scanning position in the Y direction of the measurement light. That is, there is a difference in the acquisition time of tomographic information depending on the part on the fundus. When the object to be inspected is the eye, the eye constantly makes a movement called fixation tremor, so in the case of raster scan, it is necessary to deal with the movement of the part on the fundus that may expand due to this time difference. Become. On the other hand, in the case of the resage scan, since the measurement light is scanned so as to draw a loop with a certain width, the difference in the acquisition time of the tomographic information in this loop becomes small. In the resage scan, the tomographic information from the plurality of loops obtained in this way is combined to generate three-dimensional tomographic information. Therefore, it is not necessary to consider the influence of the time difference due to the difference in the site where the tomographic information is acquired, and it becomes easier to deal with the fixation tremor in the risage scan as compared with the case of the raster scan.

JP-A-2016-017915

Yiwei Chen, Young-Joo Hong, Shuichi Makita and Yoshiaki Yasuno "Correction of eye motion artifacts in three-dimensional optical coherence tomography imaging based on the Lissajous scanning pattern", Optical Society of America (2016)

In recent years, attempts have been made to put into practical use OCTA (Optical coherence tomography angiografy), which acquires a plurality of tomographic images of the same location using OCT and draws a blood vessel image from changes between these tomographic images. In practical use of OCTA, it is necessary to repeat scanning on the same scanning line with the measurement light at relatively short time intervals in order to extract changes between tomographic images at the same location. However, in the case of the above-mentioned resage scan, the scanning lines of the measurement light are sequentially and continuously switched, and after scanning on a certain scanning line by the measurement light, all of the series of scanning lines are drawn by the measurement light, or not again. It is not possible to scan on that scan line. Therefore, it is not suitable for repeating the measurement optical scanning on the same scanning line.

Further, if it is attempted to repeatedly scan the same scanning line with the measurement light in a short time in the process of performing the resage scan, it is necessary to forcibly interrupt the operation of the scanner or the like that scans the measurement light. Further, after the operation is interrupted, it is necessary to perform the operation of returning the irradiation position so that the measurement light is irradiated to the drawing start point of each loop by the above-mentioned scanner or the like. Such a forced operation generally puts a heavy load on a scanner for measuring optical scanning such as a galvano scanner. In addition, there may be a problem in terms of the accuracy of the measurement light irradiation position at the time of returning or the stability of operation.

The present invention has been made in view of the above circumstances, and when scanning light so as to draw a figure similar to a Lissajous figure to obtain information on an object to be inspected, scanning the light on the same scanning line. Provided are an imaging device, an imaging method, and a program capable of repeating the above.

In order to solve the above problems, the imaging device according to one embodiment of the present invention is
A first scanning means for reciprocating light in a first direction in a first cycle on an object to be inspected, and a first scanning means.
A second scanning means that reciprocates the light in a second direction different from the first direction in the second cycle on the object to be inspected.
A control means for delaying the scanning of the light of the second scanning means by a predetermined time with respect to the scanning of the light of the first scanning means.
An image generating means for generating an image of the inspected object by using the information obtained from the return light of the light from the inspected object is provided .
The control means delays the predetermined time when the drive waveform applied to the second scanning means has the maximum amplitude after the light is scanned by the second scanning means for a predetermined cycle. characterized Rukoto applied to the second scanning means.

According to the present invention, when light is scanned so as to draw a figure similar to a Lissajous figure to obtain information on an object to be inspected, it is possible to repeat scanning of light on the same scanning line.

It is a figure which shows the schematic structure of the OCT apparatus used in the embodiment of this invention. It is a figure which shows the drive voltage waveform applied to the scanner when scanning the measurement light so as to draw a Lissajous figure. It is a figure explaining the scanning style of the measurement light when drawing a Lissajous figure by a scanner. It is a figure which shows the drive voltage waveform applied to the scanner in 1st Embodiment of this invention. It is a figure explaining the scanning mode of the measurement light at the time of drawing a figure similar to a Lissajous figure by the measurement light in 1st Embodiment. It is a flowchart which shows the image generation processing in 1st Embodiment. It is a flowchart which shows the group index creation process in 1st Embodiment. It is a flowchart which shows the data rearrangement processing in 1st Embodiment. It is a figure explaining the method of data interpolation in the data rearrangement processing. It is a flowchart which shows the alignment / combination process in 1st Embodiment. It is a figure which shows the drive voltage waveform applied to the scanner in the 2nd Embodiment of this invention. It is a figure explaining the scanning mode of the measurement light at the time of drawing a figure similar to a Lissajous figure in the 2nd Embodiment. It is a figure explaining the relationship between the scanning locus of the measurement light on the display screen and a sampling point in 2nd Embodiment. It is a figure explaining the scanning mode of the measurement light at the time of drawing a figure similar to a Lissajous figure in the modification of the 2nd Embodiment.

Embodiments of the present invention will be described in detail below with reference to the drawings. The shape described in the following embodiments, the relative positions of the components, and the like are arbitrary and can be changed according to the configuration of the device to which the present invention is applied or various conditions. Also, in the drawings, the same reference numerals are used between the drawings to indicate elements that are the same or functionally similar.

[First Embodiment]
Hereinafter, the configuration of the OCT apparatus according to the first embodiment will be described with reference to FIG. 1, the scanning mode of the measurement light will be described with reference to FIGS. 2 to 5, and the image generation process will be described with reference to FIGS. 6 to 10. In the following description, the case where the fundus of the eye to be inspected 120 is measured as an object to be inspected will be taken as an example.

<Device configuration>
FIG. 1 shows the overall configuration of the optical tomography imaging apparatus of this embodiment. The OCT device 100 according to the present embodiment includes an OCT optical system, a control device 109, a control PC 111, and a display device 112. The OCT optical system mainly includes a low coherence light source 101, a beam splitter 103, a scanning optical system 104, an eyepiece lens system 105, a reference mirror 107, and a detection unit 110. The control PC 111 can be configured by using an arbitrary general-purpose computer, but may be a computer dedicated to the OCT device. The display device 112 can be configured by any display. Further, although the control device 109, the control PC 111, and the display device 112 are shown individually, they may be integrated as appropriate. The control PC 111 is accompanied by an input device for inputting measurement parameters of the OCT device, selecting various modes for data acquisition, reading and executing a measurement program stored in advance, and the like. The input device may be arranged on the display device 112 side. Further, the control PC 111 may display various images described later on the display device 112 as display control means, or may assign various tomographic information described later to each pixel on the display screen of the display device 112.

The light emitted from the low coherence light source 101 passes through the optical fiber and becomes parallel light by the fiber collimator 102. This parallel light is split into measurement light and reference light by the beam splitter 103.

The measurement light is applied to the eye 120 to be inspected through the scanning optical system 104 and the eyepiece lens system 105, which are composed of two galvano scanners controlled by the control device 109. The two galvano scanners include an X galvano scanner that changes the irradiation position of the measurement light in the X direction and a Y galvano scanner that changes the irradiation position of the measurement light in the Y direction orthogonal to the X direction, and constitutes a scanning means. By operating these two galvano scanners, the scanning optical system 104 can change the irradiation position of the measurement light on the fundus in two dimensions. Further, the eyepiece lens system 105 is on the electric stage and can move in the optical axis direction according to the control signal of the control device 109. By moving the eyepiece system 105 in the optical axis direction, the focal position of the measurement light can be changed. The return light reflected or scattered from the fundus is returned to the beam splitter 103 via the eyepiece system 105 and the scanning optical system 104 through the above path in the reverse direction. The path through these optical systems to which the measurement light is guided is called a measurement optical path.

On the other hand, the reference light is reflected from the reference mirror 107 through the dispersion compensating glass 106, and is returned to the beam splitter 103 through the same path in the opposite direction. The reference mirror 107 is installed on an electric stage that moves in the optical axis direction, and can move its position in the optical axis direction according to a control signal of the control device 109. By adjusting the position of the reference mirror 107, it is possible to adjust the reference optical path length which is the optical path length of the reference light returning from the beam splitter 103 to the beam splitter 103 via the reference mirror 107. The dispersion compensating glass 106 has a configuration in which two dispersion prisms having the shape of a right triangle are arranged so that their hypotenuses face each other. The reference optical path length can also be adjusted by shifting the positions of these dispersion prisms. Normally, since the optical systems constituting the measurement optical path and the reference optical path are different, the amount of wavelength dispersion of each is different, which is not the optimum interference condition. Optimal interference conditions are obtained by inserting the dispersion compensating glass 106 into the reference optical path in order to adjust the wavelength dispersion amount.

The return light reflected or scattered from the fundus of the eye 120 to be inspected and the reference light reflected by the reference mirror 107 are combined by the beam splitter 103. When the optical path length of the measurement light and the optical path length of the reference light are the same, this combined wave light becomes interference light showing interference fringes. Since each of these interference fringes corresponds to a layer or the like at the back of the fundus, it is possible to obtain information (tomographic information) in the depth direction of the fundus by analyzing the interference fringes. The interference light is input to the optical fiber by the fiber collimator 108 and input to the detection unit 110. The detection unit 110 includes a diffraction grating that disperses the input interference light and a line sensor unit that detects the dispersed light, and the line sensor unit converts the dispersed light into a digital detection signal and the detection signal. Is sent to the control device 109.

The control device 109 sends a detection signal to the control PC 111. The control PC 111 has a processing unit (not shown) that executes image generation processing software that generates an image of the fundus of the eye. The processing unit operates as a module corresponding to various steps in the image generation process described later, and executes these steps using the input detection signal. By executing these steps, the OCT apparatus generates tomographic information used for generating a tomographic image at a position on the fundus where the measurement light is irradiated.

As the OCT device used in the present embodiment, a Fourier domain type OCT (FD-OCT) device that acquires the depth information of the measurement target included in the detected light by replacing it with the frequency information is illustrated. Further, as the FD-OCT (Fourier Domain Optical Coherence Tomography) device, a spectral domain type OCT (SD-OCT) device is particularly exemplified. However, the OCT apparatus to be used is not limited to this, and for example, a wavelength sweep type OCT (SS-OCT) apparatus may be used as the FD-OCT apparatus. It is also possible to use other known OCT devices.

<Imaging method>
The imaging method according to the present embodiment for imaging a three-dimensional tomographic image from the fundus using the above-mentioned OCT apparatus will be described below. The imaging method of the present embodiment is based on a method of scanning the fundus in a plane with the measurement light so as to draw a Lissajous figure on the fundus with the measurement light. A scanning mode in which a Lissajous figure is drawn with measurement light in this way is called a Lissajous scan. In the following description, first, an imaging method in the case of a Lissajous scan in which a plane scan is performed with a general Lissajous figure will be briefly described, and then an imaging method according to the present embodiment will be described.

The drive waveform of the galvano scanner when drawing a Lissajous figure with the measurement light will be described with reference to FIGS. 2 and 1. For convenience of explanation, the galvano scanner that scans the measurement light in the X-axis direction, which is an arbitrary direction on the bottom surface of the eye, is referred to as the above-mentioned X galvano scanner, and similarly, the galvano scanner that scans the measurement light in the Y-axis direction on the bottom surface of the eye. Is the Y galvano scanner described above. The X-axis and the Y-axis are generally orthogonal to each other, but here, they do not necessarily have to be orthogonal to each other and may intersect at a specific angle.

When drawing a Lissajous figure with the measurement light, a cosine wave having _{different periods T x} and T _y is given to the X galvano scanner and the Y galvano scanner, respectively. In this case, since each period is different, when one of the sine waves reaches one period, the other sine wave is less than one period or has passed one period and transitions to the next period. It will be in the state of being. That is, a drive waveform such that the phases of the two cosine waves are shifted for each cycle is applied to the corresponding galvano scanners. It should be noted that one cycle and the other cycle may be different from each other. For example, when one cycle reaches one cycle, the other cycle may be deviated from an integral multiple of the cycle.

ただし、式１において、０≦ｔ_ｘ＜Ｔ_ｘ、０≦ｔ_ｙ＜Ｔ_ｙであり、ｔ＝（Ｌ_ｘ-１）Ｔ_ｘ+ｔ_ｘ＝（Ｌ_ｙ-１）Ｔ_ｙ+ｔ_ｙである。 That is, as two cosine waves represented by the following equation 1, a drive waveform whose phase shifts with each period is applied to each galvano scanner to drive them.

However, in Equation 1, 0 ≦ t _x <T _x , 0 ≦ t _y <T _y , and t = (L _x -1) T _x + t _x = (L _y -1) T _y + _ty . be.

Here, f (t) indicates the position of irradiation of the measurement light on the resage locus when t seconds have passed from the start of driving the galvano scanner, and A _x and A _y are the X galvano scanner and the Y galvano scanner, respectively. The amplitude of the drive waveform of is shown. L _x and _Ly are indexes indicating the number of cycles at which the point at time t is in the cosine wave, which is each drive waveform. In the example of the drive waveform shown in FIG. 2, the case where the conditions of A _x = A _y = 5 [V], T _x = 13 [ms], and T _y = 12 [ms] is illustrated is illustrated. ..

FIG. 3 shows the trajectory of the measurement light when the galvano scanner is driven using the drive waveform shown in FIG. 2 and the fundus is scanned with the measurement light. When the galvano scanner is driven by the drive waveform, when the locus of the measurement light drawn on the fundus is plotted for each cycle of the X galvano scanner, 12 patterns of annular loci 301 are obtained. The Lissajous figure 302 is obtained by assembling (synthesizing) these circular loci 301. Hereinafter, for convenience of explanation, this annular locus 301 obtained for each cycle of the X galvano scanner will be referred to as a loop. Here, the loop may be a locus for each cycle of the Y galvano scanner. Further, as a figure collectively called a Lissajous figure, a mode in which the drive cycle ratio of the X galvano scanner and the Y galvano scanner is set to, for example, 2: 3 and the drive speed difference is relatively large can be considered. However, in reality, it is preferable that the drive speeds of both galvano scanners do not differ so much due to the device configuration. Therefore, the ratio of both scanner drive cycles in practical use such as this embodiment is assumed to be close to 1: 1.

In the annular locus 301, loops L = 1 to L = 6 are drawn clockwise, and loops L = 7 to L = 12 are drawn counterclockwise. As can be seen from the circular locus 301, the end point of each loop and the start point of the next loop are continuous points. This is clear from the drive waveform of the galvano scanner shown in FIG.

However, as described above, due to the characteristics of the scanner and the like, the resage scan is not suitable for repeatedly acquiring a plurality of tomographic information from the same position by repeatedly drawing the same loop with the measurement light. Specifically, when repeatedly acquiring tomographic information, one galvano scanner is repeatedly driven for each cycle, but the other galvano scanner is less than one cycle or has passed one cycle and the next cycle. It is in a state of transition to. Therefore, at the moment when the irradiation position of the measurement light is changed from the end point of a certain loop to the start point of the loop, one galvano scanner is forced to drive with a large load connecting the discontinuities. Such driving imposes an excessive load on the driving system of the galvano scanner and causes deterioration of the position accuracy when scanning the measurement light. Deterioration of the position accuracy leads to a decrease in the correlation value in the generation of the superimposed image described later, which causes a decrease in the number of superimposed images and a blur of the image. Further, in OCTA, the decorrelation value increases when calculating the correlation between tomographic images, and as a result, there is a risk of erroneously detecting a tissue that is not a blood vessel as a blood vessel.

When drawing a Lissajous figure with measurement light, basically, a galvano scanner that reciprocates the measurement light in an arbitrary X direction and a galvano scanner that reciprocally scans the measurement light in the Y direction perpendicular to the X direction are used in the X direction. And the measurement light in the Y direction are scanned at the same time. A Lissajous figure is obtained by drawing a loop whose shape changes with time because the drive cycle in the X direction and the drive cycle in the Y direction are different.

Here, instead of driving each galvano scanner in a different cycle, a figure similar to the Lissajous figure can be obtained by repeatedly driving each galvano scanner in the same cycle and changing the phase at a predetermined timing. It can be drawn with measurement light. When both galvano scanners are driven in the same cycle, the same galvano scanner can be used because the scanning speed of the measurement light is the same, and the drive system and the like can be standardized. When two galvanos scanners are driven in such a driving mode, the measurement light is repeatedly drawn in the same loop until the phase is changed, and the position accuracy can be expected to be maintained during repeated scanning.

FIG. 4 shows an example of a drive waveform used when driving such a galvano scanner. In this case, the X galvanometer scanner and Y galvanometer scanners, have decided to use a drive waveform having the same period T _y. Then, when the drive waveform given to one galvano scanner (X direction) has passed an arbitrary cycle (one cycle in the illustrated example), the phase is shifted with respect to the drive waveform of the other galvano scanner (Y direction). , The phase is delayed by a predetermined time. The timing of changing the phase does not necessarily have to be every one cycle, and may be changed at a predetermined cycle interval or a predetermined time interval. In the example shown in FIG. 4, in the above equation 1, A _x = A _y = 5 [V] and T _x = T _y = 12 [ms]. However, the X galvano scanner is driven so as to delay the phase by 1 [ms] for each cycle.

As described above, in the example shown in FIG. 4, the phase is delayed for each cycle. However, the timing of delaying the transfer is not limited to this example. For example, the phase may be delayed at a timing that has passed a plurality of cycles such as every 10 cycles instead of every 1 cycle. That is, after the measurement light is scanned by any of the galvano scanners for a predetermined cycle, a delay of a predetermined time may be given. By setting in this way, the measurement light draws the same loop a plurality of times (10 times in this case), then transitions to the next loop, and draws the next loop a plurality of times. Therefore, a plurality of tomographic information can be continuously and repeatedly obtained from the same position (cross section). The plurality of tomographic information acquired in this way is effective when performing addition processing to generate an image in which so-called speckle noise is reduced, or when executing the above-mentioned OCTA.

It is desirable that the timing for delaying the phase between the drive waveforms is the timing at which the delay drive waveform has the maximum amplitude. At the timing of the maximum amplitude, the displacement speed of the galvano scanner is minimized, and the load on the galvano scanner generated when the phase delay is applied is minimized.

FIG. 5 shows a trajectory of the measurement light drawn when the fundus is scanned by the measurement light using the drive waveform of FIG. When the locus of the measurement light is plotted for each loop in the same manner as in FIG. 3, a locus 501 such as a ring of 12 patterns is obtained as shown in the figure. By assembling (synthesizing) these circular loci 501, a Lissajous figure 502 similar to the Lissajous figure can be obtained. In the example shown here, a series of scanning of the measurement light is completed by drawing 12 patterns of loops for the sake of facilitation of explanation and illustration. However, in reality, more patterns such as 1024 patterns are drawn. The figure 502 drawn by the measurement light according to the driving mode of the two galvano scanners described above is similar to the Lissajous figure 302, but is not a Lissajous figure. Therefore, in the present specification, the figure obtained by delaying the scanning of one of the galvano scanners in this way is referred to as a Lisage-like figure.

As described above, in the present embodiment, both galvano scanners are driven by using cosine waves of the same period, and the galvano scanners are driven so that the phase is delayed with respect to one reciprocating scan at predetermined intervals. This makes it possible to draw a scanning line on the fundus with the measurement light, which is similar to a Lissajous figure and has high repetitive position accuracy. Therefore, it is possible to acquire a highly accurate addition image for reducing so-called speckle noise and an OCTA image for extracting blood flow. Further, according to the present embodiment, plane scanning is performed in a style of drawing a figure similar to a Lissajous figure that does not require a tracking function, and the scanning position accuracy is improved when the measurement light is scanned so as to draw the same loop. It becomes possible to plan. Furthermore, by driving two galvano scanners that move around different axes at a constant speed, it is possible to arrange the imaging points in a grid pattern in the imaging region, and it is possible to shorten the processing time in image generation. ..

In the above-mentioned example, the X-galvano scanner and the Y-galvano scanner reciprocally scan the measurement light at the same cycle. However, for example, when it is desired to scan the measurement light over the entire scanning range more quickly, the cycle of one reciprocating scan of both galvano scanners may be set to be an integral multiple of the cycle of the other reciprocating scanning. Alternatively, when the galvano scanner is driven at a certain scanning speed, the cycle of one reciprocating scan may be 1/2 times the cycle of the other reciprocating scan, or 1 times an integer. In this setting, as described above, when one galvano scanner finishes the reciprocating scan, a delay is caused with respect to the other galvano scanner to provide a phase difference, so that the same effect as the above-mentioned example can be obtained. However, in this case, it is necessary to adjust so that the phases of the two drive waveforms do not shift until the two drive waveforms given to both galvano scanners are out of phase. Further, it is necessary to adjust the phase difference so that it occurs in the intended amount and at the intended timing.

As described above, when the reciprocating scan cycle of the other galvano scanner is an integral multiple of the reciprocating scanning cycle of one galvano scanner, the phase of the reciprocating scanning of one of the galvano scanners should be shifted at an appropriate timing. Then, the above-mentioned Lisage-like figure can be drawn by the measurement light. However, the period ratio is set to 1.5 times, and the scanning start position of both scanners returns to the original scanning start position by scanning two round trips by one scanner and scanning three round trips by the other scanner. It may be. In this case, if the reciprocating scanning of one of the galvano scanners is delayed when the original scanning start position is returned, a risage-like figure can be drawn as in the above example. However, if the ratio of the reciprocating scanning cycle of one galvano scanner to the reciprocating scanning cycle of the other galvano scanner is too large, the scanning speed of the other galvano scanner becomes extremely high and the load on the galvano scanner becomes large. In addition, when it is necessary to pass a plurality of cycles until the irradiation positions of the light measured by both galvano scanners return to the scanning start position and match, from the acquisition of one tomographic information to the acquisition of the next tomographic information in the same cross section. It takes time. Therefore, in order to prevent the two galvano scanners from having an extreme load difference and to acquire tomographic information from the same cross section in a short time, the ratio of each period should be 1.5 times or 2 times. It is preferable to set the degree.

<Image generation method>
The image generation process performed based on the tomographic information obtained by applying the drive waveform described above to each of the two galvano scanners will be described below. In the image generation process of the present embodiment, the measurement data is rearranged based on the drive waveform of the galvano scanner, and then the position is corrected by utilizing the correlation of the brightness values.

Hereinafter, the image generation process executed in the present embodiment will be described with reference to FIG. FIG. 6 shows a flowchart of a process of generating an image using the interference signal acquired by using the above-mentioned OCT apparatus. Each process in the following flowcharts is executed by a module corresponding to each process arranged in the control PC 111 functioning as a control means.

When the image generation process is started, in step S101, the control PC 111 acquires a detection signal transmitted from the control device 109, that is, an interference signal. The interference signal is a one-dimensional data string arranged in the depth direction of the fundus of the eye, and at each measurement point (sampling point) arranged on the scanning line at regular time intervals when the imaging target is scanned two-dimensionally with the measurement light. To be acquired.

In step S102, the control PC 111 converts the acquired interference signal into a wavenumber function, then executes a Fourier transform process, extracts the amplitude value of the obtained complex number data, and obtains the brightness value. By Fourier transforming the data string on the wave number axis, it is possible to obtain the data string (fault information string) of the brightness value (fault information) at each measurement point on the fundus. These data strings are stored as three-dimensional tomographic information of the fundus in a storage unit (not shown) arranged in the control PC 111.

In step S103, the control PC 111 calculates the average value of the data strings (fault information strings) of the luminance values arranged in the depth direction. At this time, the average value may be calculated in the entire depth range, or may be calculated only in a desired depth range. As will be described later, when an image is generated from the acquired three-dimensional tomographic information, it is necessary to align and combine the tomographic information sequences acquired from each loop. In the present embodiment, an average value or the like is obtained as a representative value from a one-dimensional data string of each measurement point, and various types described later are used on the XY plane on which the Lissajous figure shown in FIG. 5 is drawn. Perform processing. In the following step S104, the average value for all the measurement points calculated by the control PC 111 in step S103 is stored in the storage unit described above.

In step S105, the control PC 111 groups the average value groups included in the highly correlated loop in which the fundus (the eye to be inspected 120) is presumed not to move significantly in the measurement data, and groups Group Index into each group. assign. The procedure for creating and allocating the Group Index will be described with reference to the flowchart of FIG. FIG. 7 is a flowchart showing a series of processes for obtaining the correlation between each loop and assigning Group Index to each loop.

When the group index creation process is started, in step S201 in FIG. 7, since the control PC 111 executes the process sequentially, indexes i and j are assigned to the loop and the group index assigned to the loop, respectively. The initial value of i is 0, and the initial value of j is -1.

Next, in step S202, the control PC 111 relates to the average value of the luminance values calculated in step S103, and has a correlation coefficient with the average value of the corresponding measurement points in the immediately preceding loop (= Loop (i-1)). Ask for. In the case of the example shown in FIG. 5, the correlation between the adjacent loops (L = 1) and the loops (L = 2) may be very low. However, the figure is shown by simplifying the number of loops for the sake of explanation, and the number of loops actually drawn is 500 to 600. Therefore, when the adjacent loops are considered at substantially the same position or the pixels corresponding to the scanning positions on the display screen, the tomographic information is obtained from the same pixel. Therefore, if the eye to be inspected does not move due to fixation tremor or the like, the correlation coefficient always shows a value of a certain value or more.

In step S203, the control PC 111 determines whether or not the correlation coefficient calculated in step S202 is equal to or greater than the threshold value. If the correlation coefficient is equal to or greater than the threshold value, the control PC 111 determines that the correlation with the adjacent loop is high, and advances the flow to step S205. If the correlation coefficient is less than the threshold value, it is determined that the loop is not drawn at a predetermined position due to the movement of the eye to be inspected or the like, and the control PC 111 advances the flow to step S204.

In step S204, the control PC 111 increments the value of j by one. For example, in the first loop (= Loop (0)), since there is no target for calculating the correlation, it is determined in step S203 that the correlation coefficient is less than the threshold value, the flow is advanced to step S204, and j = by the control PC 111. (-1) +1 = 0 is calculated.

In step S205, the control PC 111 assigns the Group Index (j) to the Loop (i). At that time, if there is no eye movement between the time of drawing the immediately preceding loop (= Loop (i-1)), the same Group Index (j-1) as the immediately preceding loop is assigned and there is movement. If so, a different Group Index (j) is assigned. In step S206, the control PC 111 ends the Loop (i) process, returns the flow to step S201, and shifts to the Loop (i + 1) allocation process. By repeating the processes of steps S201 to S206, each loop for rearrangement, which will be described later, is grouped according to the movement of the eyes.

After the loop processing is completed, the control PC 111 associates Loop (i) with Group Index (j) in step S207 and stores the loop processing in the storage unit described above. With the above processing, the process of creating Group Index in step S105 in the main flow is completed, and the flow shifts to step S106 in the flowchart of FIG.

The above processing is applied, for example, as follows. For example, in FIG. 5, if the correlation coefficients of the loops of L = 1 and L = 2 and L = 2 and L = 3 are each equal to or more than the threshold value, these three loops are set to group 0. On the other hand, when the correlation coefficient of the loop of L = 3 and L = 4 does not reach the threshold value, the loop of L = 4 is assigned to group 1. If the correlation coefficient of the loop of L = 4 and L = 5 also does not reach the threshold value, the loop of L = 5 is set in group 2. Next, if the loops of L = 5 and L = 6 are equal to or larger than the threshold value, the loop of L = 6 is also set to group 2. The correlation coefficients of these loops illustrated here are not actually high, but here, for the sake of explanation, each correlation coefficient is assumed and described.

In the main flow of FIG. 6, in step S106, the control PC 111 reads out the average value (luminance data) of the luminance stored in step S104. In step S107, the control PC 111 performs Log conversion on the read luminance data and converts it into an OCT image (a plane image forming one loop which is a scanning locus of the measurement light). Assuming that the luminance data is I, the conversion formula used for image conversion is as follows.

In step S108, the control PC 111 arranges the OCT image data acquired by the Log conversion on each pixel arranged on the XY plane of the display screen of the display device 112. At the time of arrangement, the control PC 111 assigns a sampling point to each pixel with reference to the Group Index (j) created in step S105. The rearrangement of data will be described with reference to FIGS. 8 and 9. FIG. 8 shows a flowchart of the OCT image data rearrangement process, and FIG. 9 is a diagram showing a pixel arrangement for explaining the content of the data interpolation process performed in the data rearrangement process.

When the data rearrangement process is started, in step S301, the OCT image data is rearranged into the XY coordinate system based on Equation 1 for each pixel arranged as the XY coordinate system on the display screen of the display device 112. Deploy. When a plurality of OCT image data are arranged in the same pixel by rearrangement, the average value of the OCT image data is assigned to the pixel.

Even when the OCT image data is rearranged in step S301, the OCT image data may not be arranged in the pixels depending on the arrangement in which the scanning lines are drawn. In the case of the resage scan, the scan line spacing in the vicinity of the central portion of the imaging region is relatively wide, and the OCT image data may not be arranged in the pixels. In step S302, by the rearrangement process as described above, it is determined whether or not there is a pixel in the XY coordinate system included in the area for generating an image in which the OCT image data is not arranged. When it is determined by the PC 111 that there are no pixels in which no data is arranged, the OCT image data is arranged in all the pixels. In such a case, the control PC 111 ends step S108 for performing data rearrangement processing, and advances the flow to the next step. However, when it is determined that there are pixels for which data is not arranged, the control PC 111 advances the flow to step S303 and performs data complement processing.

In step S303, the control PC 111 complements the pixels for which no data is arranged. There are various methods of data complementation. For example, the average value of the OCT image data arranged in a large block of 9 pixels consisting of 3 rows and 3 columns including the surrounding pixels may be applied to the pixels in which the data is not arranged, or may be adaptively applied from the surrounding pixels. It may be interpolated. Hereinafter, an adaptive data interpolation method will be described with reference to FIG.

FIG. 9 shows a matrix of pixels arranged in the XY coordinate system. Adaptive interpolation detects the correlation between the upper, lower, left, and right data of the pixel to be interpolated, performs interpolation from the upper and lower data when the vertical correlation is high, and complements from the left and right data when the horizontal correlation is high. By performing interpolation on pixels in which data is not arranged in this way, it is possible to form a continuous image on the display screen.

２．求めた差の絶対値に基づいて、補間の方法（補間にＯＣＴ画像データを用いる画素）を選択する。例えば、ＹＤｉｆｆ＞ＸＤｉｆｆならば、水平方向の相関が高いと判断し、式４を用いて補間する。

また、ＸＤｉｆｆ＞ＹＤｉｆｆならば、垂直方向の相関が高いと判断し、式５を用いて補間する。

For example, the process performed by the control PC 111 when interpolating the central pixel P11 in FIG. 9 is as follows.
1. 1. The absolute value (XDiff, YDiff) of the difference between the upper, lower, left and right pixels of the complement target is obtained by the equation 3.

2. The interpolation method (pixels that use OCT image data for interpolation) is selected based on the absolute value of the obtained difference. For example, if YDiff> XDiff, it is determined that the correlation in the horizontal direction is high, and interpolation is performed using Equation 4.

Further, if XDiff> YDiff, it is determined that the correlation in the vertical direction is high, and interpolation is performed using Equation 5.

As an example of the interpolation method, the correlation values are compared with the arrangement of the upper, lower, left, and right pixels of the central pixel P11, but the pixel data used for the interpolation is not limited to that of the pixels of this arrangement. For example, adaptive interpolation may be performed using four pixels P00, P02, P20, and P22 diagonally adjacent to the pixel to be interpolated. In the present embodiment, since the locus of the measurement light is drawn so as to travel diagonally on the XY plane, it is considered that such adaptive interpolation using four diagonally adjacent pixels is also effective. As described above, in the present embodiment, the data obtained by adaptively interpolating the OCT image data is arranged in the pixels in which the data is not arranged.

After the data rearrangement is completed, the control PC 111 advances the flow to step S109 shown in FIG. In step S109, the control PC 111 aligns and combines each of the loop images generated by rearranging the OCT image data in step S107. Alignment and coupling will be described with reference to FIG. FIG. 10 shows a flowchart of the processing of the alignment and the combination of the loop images.

In step S109, the control PC 111 starts the process of aligning and combining the loop images. When the process is started, in step S401, the control PC 111 uses only the pixels in which the OCT image data of the same loop of the Group Index (j) created in Step S105 is rearranged, and each Group Index (j) is used. Generate a circular image for each.

Next, in step S402, the control PC 111 rearranges the generated circular image in the order of the circular image having the largest number of pixels. In the following step S403, the control PC 111 selects the circular image having the largest number of pixels as the reference image.

The loop process is performed after step S404, and the control PC 111 combines the reference image with other circular images to enlarge the image. First, in step S404, the control PC 111 assigns an index (j') to the rearranged circular image in order to rotate the loop process. When there are m circular images, since one image has already been selected as the reference image, the number of loop processes is m-1. In step S405, the control PC 111 selects the next circular image to be combined (the next index circular image) with respect to the selected reference image or the circular image after the other images are combined.

In step S406, the control PC 111 aligns the annular image using the information of the overlapping pixel region of the two annular images to be combined. Specifically, the alignment shifts one of the circular images in the X and Y directions so that the images of the superimposed regions in the two circular images are compared and the correlation coefficient is the highest. At this time, the shift parameter used for the alignment is applied to the entire region of the circular image to be combined. In step S407, the control PC 111 combines the two annular images aligned in step S406.

In step S408, the control PC 111 determines whether or not the index of the Loop Image (j') is m-1, and if it is m-1, the process is exited if all the loop processing is completed, and the image is displayed. End the generation flow. However, when it is determined that the m-1 is not reached and the entire loop processing is not completed, the control PC 111 returns the flow to step S404, and the circular image of the next index is transferred to the combined circular image. Repeat the alignment and join processing loop. By repeating this process, the control PC 111 enlarges the circular image. When the processing of all loops is completed, OCT image data is assigned to all the pixels on the display screen, and a flat image of the fundus can be generated.

By executing the process shown in the flowchart shown in FIG. 6, individual loops are combined using the representative values of each measurement point, and a plane image is generated using the representative values. As a result, the drawing positional relationship of each rounded square loop on the fundus can be obtained. After the processing, three-dimensional tomographic information of the fundus is constructed from a plurality of tomographic information sequences obtained in advance based on the obtained drawing positional relationship. Specifically, each of the plurality of tomographic information sequences obtained from each loop is arranged so as to correspond to the pixels on the display image according to the drawing positional relationship. At that time, based on the index of each loop assigned in the joining process, the arrangement corresponding to the display pixel of the one-dimensional data string which is the tomographic information string at each measurement point is performed. At the same time, according to the process of step S303 described above, the interpolation process for the missing portion of the data string is performed. By going through the above operations, the three-dimensional tomographic information of the fundus is constructed from the tomographic information obtained by scanning the fundus with the measurement light so as to draw a lisage-like figure.

As described above, the control PC 111 acquires a plurality of tomographic information sequences included in each of a plurality of different rounded square loops in which the measurement light is drawn on the fundus by the X galvano scanner and the Y galvano scanner. Further, the control PC 111 sets a plurality of acquired tomographic information sequences into one group, and obtains an average value (average information) for each of the acquired tomographic information sequences. Further, the control PC 111 groups each of these series of average value groups (one loop) by using the addition result of a series of average values in one loop corresponding to the acquired plurality of tomographic information sequences. After that, the control PC 111 assigns each of the average values included in the same group after grouping to the pixels in the display means. Next, the control PC 111 acquires the positional relationship of each rounded square loop based on the average value assigned to the pixels. After that, the control PC 111 arranges a plurality of tomographic information sequences together with each pixel on the display screen based on the acquired positional relationship, and generates three-dimensional tomographic information of the fundus. The above processing is executed by making the control PC 111 function as an image generation means.

Here, the average values included in each average value group are added and the results are compared, and it is determined that the average values are included in the same group when the difference in the addition results is equal to or less than a predetermined threshold value. However, the grouping method is not limited to this, and other known methods may be used as long as the tomographic information group sequences obtained by each loop are compared and grouping can be performed based on the magnitude of the difference.

As described above, the imaging device as one embodiment of the present invention includes an X galvano scanner as a first scanning means and a Y galvano scanner as a second scanning means. The X galvano scanner reciprocally scans the measurement light in the X direction, which is the first direction, in the first cycle on the fundus of the eye to be inspected, which is the object to be inspected. Further, the Y galvano scanner reciprocally scans the measurement light in the Y direction, which is the second direction, in the second cycle on the fundus. Here, the first cycle and the second cycle are determined by the above-mentioned relationship. Further, the control PC 111, which is a control means, delays the scanning of the measurement light by the other galvano scanner by a predetermined time with respect to the scanning of the measurement light by one galvano scanner of the X galvano scanner and the Y galvano scanner. Further, the control PC 111 functions as an image generation means, and generates an image of the fundus using information such as a brightness value obtained from the return light from the fundus.

[Second Embodiment]
In the first embodiment described above, the three-dimensional tomographic information of the fundus is obtained by scanning the fundus so as to draw a lisage-like figure with the measurement light. However, in the Lissajous figure composed of a composite of a plurality of ellipses and the Lissajous figure described in the first embodiment, the line spacing drawn in the central portion and the peripheral portion is different, and the line spacing becomes coarse in the central portion and in the peripheral portion. Become dense. This means that when tomographic information such as the retina is acquired by a resage scan, the tomographic information acquired from the central portion of interest is reduced, and the image quality of the central portion of the generated image is deteriorated. Therefore, it is conceivable that the waveform of the drive voltage applied to each of the two galvano scanners is a triangular wave so that the measurement light draws an image similar to a Lissajous figure at linear and even intervals.

When a triangular wave is used as the drive voltage waveform, the two galvano scanners scan the measured light on the fundus at a constant velocity. As a result, the line drawn by the measurement light becomes linear, and the drawn loop becomes rectangular. In the Lisage-like figure obtained by synthesizing such a loop, the line spacing between the peripheral portion and the central portion becomes even.

Here, when a voltage is applied to the galvano scanner as a triangular wave, the galvano scanner suddenly changes the direction of rotation at the turning point of the drive voltage at which the drive voltage changes from an increase to a decrease or a decrease to an increase. The acquisition of image information such as tomographic information at equal intervals by using the above-mentioned triangular wave is based on the premise that the galvano scanner changes the rotation direction at a constant speed and at a constant position even at this turning point.

However, when changing the direction of rotational movement, it is necessary to pay attention to the deceleration that copes with the inertia acting in the direction of movement and the acceleration that is required after changing the direction. That is, in the galvano scanner, the execution of almost instantaneous deceleration and acceleration for constant velocity scanning and the accuracy of the instantaneous deceleration to maintain the accuracy of the irradiation position of the measurement light at the folded portion at a predetermined timing. Execution is required. However, it is not easy to satisfy this premise in the galvano scanner generally used in the OCT apparatus, and the ideal turning point operation cannot be performed, and the irradiation position accuracy of the measurement light fluctuates.

Therefore, in the second embodiment, the basic waveform is a triangular wave as the drive waveform, and the folded portion of the waveform is connected by a sine waveform or the like. As described above, in the folded portion which is the deceleration region of the scanner, a drive waveform having a smooth or gradual waveform change is applied instead of a triangular wave in which a steep waveform change occurs. As a result, the measurement light gradually changes the scanning direction on the fundus. By using such a drive waveform, it is possible to slow down the change in acceleration of the folded portion and scan the measurement light without deteriorating the position accuracy of the scanner. The folded portion described here is the above-mentioned folding point and its vicinity where the change in voltage changes from an increase to a decrease or a decrease to an increase in the drive waveform, and is a square annular corner portion described above in the scanning locus. Corresponds to the point where the scanning direction of the measurement light changes and its vicinity.

Hereinafter, the second embodiment will be described. Since the configuration of the OCT apparatus and the image generation method used in the second embodiment are the same as those in the first embodiment described above, detailed description thereof will be omitted here.

<Imaging method>
As described above, in the present embodiment, regarding the drive waveform used for driving both galvano scanners, a triangular wave is used as a basic waveform, and a folded portion of the waveform is connected by a sine waveform or the like. In this case as well, the X galvano scanner and the Y galvano scanner are set to be driven in the same period T, and the phase of the X galvano scanner is delayed so that the phases of the two triangular waves are shifted every cycle. As in the first embodiment, the phase delay can be performed every time the sex number of a predetermined period elapses.

ただし、０≦ｔ＜Ｔであり、ｔ＝（Ｌ−１）Ｔ+ｔである。 When the phase delay is not taken into consideration, the driving waveform of the triangular wave that is the reference of the galvano scanner can be expressed by the following equation 6.

However, 0 ≦ t <T, and t = (L-1) T + t.

式７において、ｇ（ｔ_ｎ）はガルバノスキャナにおける正負の折り返し部において三角波に接続する折り返し用の駆動波形であり、該折り返し用駆動波形に推移してからｔ_ｎ秒が経過した時点の軌跡の位置を示す。Ａ_ｒは折り返しのｓｉｎ波形の振幅、Ｔ_ｒは折り返しに要する時間である。折り返し用の駆動波形を接続した結果、得られる駆動波形を図１１に示す。 Here, f (t) indicates the position of the locus at the time when t seconds have elapsed from the start of driving the galvano scanner, A is the amplitude of the driving waveform of the galvano scanner, and L is the time t in the triangular wave which is each driving waveform. It is an index showing the number of cycles of the point. Further, a sine waveform as shown in Equation 7 is connected to the folded portion of the triangular wave to moderate the change in acceleration of the folded portion.

In Equation 7, g (t _n ) is a folding drive waveform connected to a triangular wave at the positive and negative folding portions of the galvano scanner, and is a locus at a time when _{t n seconds have elapsed since the transition to the folding drive waveform.} Indicates the position. _Ar is the amplitude of the sine wave of _{folding, and Tr} is the time required for folding. FIG. 11 shows a drive waveform obtained as a result of connecting the drive waveform for folding back.

In the example of the drive waveform shown in FIG. 11, A _x = A _y = 4 [V], T _x = T _y = 8 [ms], _Ar = 1 [V], _Tr = 2 [ms]. be. Further, in the example of the drive waveform shown in FIG. 11, the phase of the X galvano scanner is delayed by 1 [ms] for each cycle. In addition, the galvano scanner scans linearly in the section from -4 [V] to +4 [V], and a sine wave is connected to the folded portion of the triangular wave to form a gently changing waveform. ..

FIG. 12 shows a locus drawn by the measurement light when the galvano scanner is driven by the drive waveform shown in FIG. 11 and the fundus is scanned by the measurement light. When each of the loops, which are the loci drawn by the measurement light, is plotted on the display screen of the display device 112, a total of 12 patterns of rounded square loops 1201 are obtained. By assembling these loops, it is possible to obtain a lisage-like figure 1202 in which the measurement light does not suddenly change in the scanning direction at the folded portion. As in the first embodiment, here, it is assumed that a series of measurement optical scans is completed by drawing 12 patterns of the loop, but in reality, more patterns such as 1024 patterns are drawn. There is. In this case, by repeatedly giving a delay corresponding to 1/1024 of one cycle of one galvano scanner during scanning of any galvano scanner, a denser lisage-like figure can be drawn.

Here, the sample point S (x,) when the sample is performed at the sample intervals of 1.0 [ms] and 0.5 [ms] on the scanning locus when the Lissajous figure 1202 is drawn with the measurement light y) is shown in FIG. In the figure, the sampling points 1304 arranged at any of the intersections of the scanning lines 1303 are indicated by black squares. An example of sampling at a sample interval of 1.0 [ms] is shown in FIG. 13 (a). In this case, assuming that the sampling point corresponding to the display screen is S (x, y), (x, y) = (2n, 2n), where n = -2, -1, 0, 1, 2, 5 × A total of 25 pixels are sampled four times each on a frame of 5 square arrays. Further, when the sample interval is 0.5 [ms], as shown in FIG. 13 (b), (x, y) = (2n + 1,2n + 1), where n = -2, -1, 0, 1 Sampling points with 4x4 pixels are added. As a result, sampling is performed on a total of 41 pixels from the frame of the rhombus array. The square array described here is sampled in a square shape of, for example, (0,0), (0,2), (2,2), (2,0) on the display screen in FIG. 13A. An array in which points are placed. The rhombus array refers to an array in which sampling points are arranged in the rhombus shapes of (0,0), (-1,1), (2,0), and (1,1) on the display screen.

In the case of the sampling points of the rhombus array, the pixel value at the center coordinate of the rhombus may be generated from the neighboring pixel values in the same manner as the pixel value rearrangement by using the pixel value interpolation method for the unarranged pixels described above. .. By performing the interpolation, as shown in FIG. 13C, a 9 × 9 square array can be obtained as the pixels used for display, and it is possible to improve the apparent resolution. Alternatively, for example, by setting the delay phase given to the galvanometer mirror Y for each cycle to 0.5 [ms], the center coordinates of the sampling points in the rhombic array shown in FIG. 13B are also set as the sampling points, and the center is the center. The pixel value of the coordinates may be actually sampled. However, when actually performing a sample with the delay phase halved in this way, it should be noted that the measurement time is doubled because all 24 patterns of rounded and square loci are required.

In the present embodiment, the measurement light is scanned at a constant velocity in the region where the galvano scanner is driven by the triangular wave. This area is defined as the second area. Further, the acceleration / deceleration region in which the galvano scanner is driven by the sine waveform, which is the folding drive waveform for causing the galvano scanner to make a gradual change in the scanning direction, is defined as the first region. According to the scanning method in the present embodiment, the _{region defined by the amplitudes A x} and A _y , that is, the region within ± 4 [V] is imaged with a uniform density in both the X galvano scanner and the Y galvano scanner. Corresponds to two areas. Outside this region, the regions of voltages 4 to 5 [V] and -5 to -4 [V] in the drive waveform, where the direction in which the scanning line of the measurement light is drawn gradually changes, correspond to the first region. do.

In this case, the phase delay section is provided in the first region. In this phase delay section, in order to reduce the influence of deterioration of scanning position accuracy in the folded portion, it is desirable to stop imaging in the region. Alternatively, it is preferable not to acquire tomographic information on the locus passing during the transition between the loops from the drawing state of one loop to the drawing state of the next loop. Even in the explanation of the sampling point using FIG. 13 described above, it is assumed that the tomographic information during the transition of the loop is not acquired.

However, as an image displayed for reference at the time of diagnosis, such as a simple display image, an area including this first area may be displayed. In this case, it is preferable to continuously acquire the tomographic information even in the phase delay section and generate an image. In this case, the generated image will be slightly distorted around the frame in which the sampling points shown in FIG. 13 are arranged. However, for example, the amplitude A = ± of the drive waveform in the imaging region of the 9 × 9 array defined by the amplitude A = ± 4 [V] of the drive waveform of each of the X galvano scanner and the Y galvano scanner of FIG. It can be slightly expanded to the imaging region of the 11 × 11 array defined by 5 [V].

By executing the above processing, in the constant velocity region, an image of the fundus can be generated from the interference signal obtained by scanning like a lisage. In the second region, the interference signal can be acquired from the scanning lines that are linear and the imaging points are arranged without being biased, so that the interpolation process in the image generation processing can be remarkably reduced. Further, in the second region, the intervals between the scanning lines are equal in the vicinity of the center and the vicinity of the periphery, and an equal number of imaging points can be obtained, which is advantageous from the viewpoint of positioning the circular image. Further, in the second region, the interval between the same loops of measurement points is also constant, and the number of pixels requiring interpolation described above is evenly arranged in the entire second region, so that the partial resolution is increased. Deterioration can also be avoided.

When generating an additional image or an OCTA image, as described in the first embodiment, the measurement light is scanned so as to repeatedly draw the same loop to acquire tomographic information in the same cross section, and the same coordinates are obtained. The sampling of the pixels will be repeated. Therefore, instead of adding a delay to the phase of the X galvanometer mirror for each cycle, a delay is added for each of a plurality of predetermined cycles (here, M cycle is assumed). Therefore, when the locus is traced while repeatedly scanning the entire locus (loop), m data sets obtained from each pixel when the repeat loop is not performed can be obtained for M times, which is the number of repetitions. .. Therefore, for example, when generating an additional image equivalent to N images, the whole locus may be repeated as an integer p = N / (M × m) times. Here, the integer p is selected in a timely manner depending on the state of fixation and tremor of the eye to be inspected and the amount of the number of additions N.

In the case of OCTA, the number of pixels for which motion contrast (decorrelation value in the present embodiment) is obtained, that is, the number of repetition loops M is usually determined according to the desired image quality. At this time, at p = 1 that minimizes the measurement time, m × (M-1) decorrelation values, that is, (M-1) decorrelation values can be obtained in m sets, which is the number of data sets. become. Therefore, in the OCTA apparatus that scans the measurement light as in the second embodiment, it is desirable to obtain a more stable OCTA image with less noise by averaging the decorrelation values of this m set.

That is, in the OCTA apparatus, the control PC 111 scans the measurement light by both galvano scanners for a plurality of cycles before delaying the scanning of the measurement light of the Y galvano scanner by a predetermined time with respect to the scanning of the measurement light of the X galvano scanner. Repeat for minutes. Then, the control PC 111 generates an image by using the information such as the luminance value obtained from the same cross section in the eye to be inspected by repeating the scanning of the pre-measurement light for a plurality of cycles.

As shown in FIG. 14, the phase difference of 1/2 of the delay time at the start of scanning so that the locus drawn at L = 1 or 7 does not become a reciprocating linear locus so that the sampling points are arranged. May be added as the initial phase of the X galvano mirror. By giving the initial phase in this way, the measurement light draws a Lissajous figure 1402 that does not draw a reciprocating straight line, and the imaging point density can be increased.

In the present embodiment, the drive waveform at that time is a sine wave different from the triangular wave so that each galvano scanner gently inverts at the folded portion where the scanning direction is reversed by each galvano scanner. As a result, the operation of the galvano scanner can be stabilized, and the accuracy of the irradiation position of the measurement light when scanning the object to be inspected with the measurement light can be maintained. Further, it is not necessary to give a load such as sudden deceleration and sudden acceleration to the operation of the galvano scanner, and it is not necessary to add a special configuration for this purpose.

In the second embodiment, an example in which a sine wave is connected to the folded portion of the triangular wave is described. However, the connected waveform is not limited to this, and if the folded portion of the drive waveform is composed of a smooth curve and the drive waveform is formed so that the direction of the scanning line gradually changes at the corner of the loop. good. Further, in the embodiment described above, in the second region, the drive waveform applied to both scanners also changes linearly so that the X galvano scanner and the Y galvano scanner scan the measurement light at a constant speed. It is supposed to be. However, the scanning mode of the measurement light in the second region is not limited to the linear scanning mode at a constant speed. If the scanning locus at the time of changing the scanning direction in the first region is smoothly connected to the scanning locus in the second region, and the change in scanning direction satisfies the change in scanning direction that can be handled by the galvano scanner. It may be a scanning locus that draws a gentle curve. That is, as described above, the change in the scanning speed of the measurement light in each of the X-galvano scanner and the Y-galvano scanner is set to be smaller in the first region than in the second region, and corresponds to the first region. The change in the drive waveform applied to both galvano scanners does not have to be abrupt.

Further, in the above-described embodiment, a case where a galvano scanner is used for each of the X scanner and the Y scanner is described. However, the scanner to be used is not limited to the galvano scanner, and a scanner generally driven according to the applied drive waveform can be used. Further, in the above-described embodiment, the case where the OCT device is used as the photographing device for scanning the measurement light is described. However, a scanning laser ophthalmoscope (SLO) for obtaining image information by scanning light on an object to be inspected such as an eye, and an adaptive optics-SLO scanning device for adaptive optics. : AO-SLO) or the like. In SLO and the like, the brightness value and the like are obtained from the reflected light of the light applied to the fundus, and the fundus plane image is obtained from the brightness value and the like. In this case, noise can be reduced and a clearer image can be obtained by adding and averaging the luminance values obtained from the same scanning line.

In this case, instead of the tomographic information sequence acquired by the OCT apparatus, the data (information) corresponding to the above-mentioned average value is directly obtained as the brightness value. That is, in these ophthalmoscopes, by drawing a plurality of different loops on the fundus, a data group consisting of a plurality of data is obtained from each of the different loops. Then, the data groups (loops) are grouped by comparing the data groups of each loop (the group in which the data obtained from the individual sampling points are arranged in a loop), and the data groups are grouped into the same group by the grouping. Each of the information in the (loop) is assigned to the pixels in the display means.

In this case, the comparison of the data groups is performed by quantifying the information (data) included in the data group as a luminance value or the like and adding the respective numerical values. Although it depends on the number of loops drawn to generate one fundus image, for example, when the number of loops is 500 or more, the drawing position of each loop is almost the same. Therefore, the added value such as the luminance value acquired in one loop and the added value such as the luminance value acquired in the loop drawn next are almost the same value. That is, by comparing the added values, it is possible to determine whether or not two consecutive loops are drawn adjacent to each other on the fundus.

If the eye to be inspected moves due to fixation and fine movement while drawing a continuous loop, the sampling points will shift, so that the obtained luminance value and the like will differ, and the added value will differ. By comparing and referring to the added values in this way, it is possible to know whether or not the eye to be inspected has moved during scanning of light. The movement of the eye to be inspected may stop after moving, or may return to a state close to the original state after moving. When the movement is stopped once, the added values obtained from each of the loops drawn during the stop show values close to each other. Therefore, it can be inferred that the loops showing values close to each other are drawn when the eyes are in the same state, and these loops are drawn so as to be arranged on the fundus as expected. Therefore, it is assumed that these loops are drawn with the eye to be inspected in the same state, and the data group obtained from each of the loops belongs to the same group as a data group that does not need to consider the correction of the drawing position. That is, the above-mentioned grouping is an operation of dividing the groups drawn by a series of measurement optical scans on the fundus into groups presumed to be drawn when the fundus is in the same state.

After that, the image generation means aligns each of the circular images by shifting one of the circular images on the image display screen so that the correlation coefficient of the corresponding pixel in the circular image generated from each of the loops becomes high. I do. As a result, it is possible to correct the positional deviation of the drawing position of the scanning line due to the measurement light caused by the fixed vision fine movement of the eye to be inspected. By synthesizing the aligned annular images obtained by the operation, a fundus image of the eye to be inspected can be generated. By performing the grouping as described above, the data can be rearranged by performing the alignment operation between the divided groups. The above operation is the same as the operation performed when the above-mentioned three-dimensional tomographic information is acquired by OCT, the average value of the tomographic information sequences arranged in the depth direction is taken, and this is used as the representative value at the sampling point. .. That is, as described above, the present invention is generally applicable to an apparatus that obtains image information by scanning light on an object to be inspected.

Further, in the above-described embodiment, the fundus of the eye 120 to be inspected has been described as an example of the object to be inspected, but the object to be inspected is not limited to this. For example, the object to be inspected may be the anterior segment of the eye to be inspected 120, or the skin or organ of the subject to be inspected. In this case, the above-mentioned imaging device can be used not only as an ophthalmic device but also as an imaging device for a medical device such as an endoscope.

(Other Examples)
The present invention supplies a program that realizes one or more functions of the above-described embodiment to a system or device via a network or storage medium, and one or more processors in the computer of the system or device reads and executes the program. It can also be realized by the processing to be performed. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-described embodiments. The present invention also includes inventions modified to the extent not contrary to the gist of the present invention, and inventions equivalent to the present invention. In addition, the above-described embodiments and modifications thereof can be appropriately combined as long as they do not contradict the gist of the present invention.

101 Light source 102, 108 Fiber collimator 103 Beam splitter 104 Scanning optical system 105 Eyepiece system 106 Dispersion compensation glass 107 Reference mirror 109 Control device 110 Detection unit 111 Control PC
112 Display device 120 Eye to be inspected (object to be inspected)

Claims (16)

A first scanning means for reciprocating light in a first direction in a first cycle on an object to be inspected, and a first scanning means.
A second scanning means that reciprocates the light in a second direction different from the first direction in the second cycle on the object to be inspected.
A control means for delaying the scanning of the light of the second scanning means by a predetermined time with respect to the scanning of the light of the first scanning means.
An image generating means for generating an image of the inspected object by using the information obtained from the return light of the light from the inspected object is provided .
The control means delays the predetermined time when the drive waveform applied to the second scanning means has the maximum amplitude after the light is scanned by the second scanning means for a predetermined cycle. imaging device according to claim Rukoto applied to the second scanning means. The imaging apparatus according to claim 1, wherein the first cycle is 1, 1.5, or twice that of the second cycle. The imaging device according to claim 1 or 2, wherein the control means delays scanning of the light of the second scanning means by a predetermined time at predetermined time intervals. A drive waveform applied to the first scanning means when the first scanning means scans the light, and applied to the second scanning means when the second scanning means scans the light. The imaging apparatus according to any one of claims 1 to 3 , wherein the drive waveform to be driven is a cosine wave. When the light is scanned by the first scanning means and the second scanning means, the control means determines the scanning speed of the light scanned by the first scanning means and the second scanning means. Controlling the first scanning means and the second scanning means so as to have a first region to be changed and a second region in which the change in the scanning speed of the light is smaller than the first region. The imaging apparatus according to any one of claims 1 to 3 , wherein the imaging apparatus is characterized. When the control means scans the light by the first scanning means and the second scanning means, the first scanning means that the light is scanned at a constant speed in the second region. The imaging apparatus according to claim 5 , wherein the means and the second scanning means are controlled. In the first region, the control means gradually decelerates from the scanning speed of the second region, gradually accelerates after the scanning direction of the light changes, and returns to the scanning speed of the second region. The image pickup apparatus according to claim 5 or 6 , wherein the first scanning means and the second scanning means are controlled. The imaging device according to any one of claims 5 to 7 , wherein the control means gives the second scanning means a delay of the predetermined time in the first region. Further provided with a display means for displaying the generated image,
The image generation means has a plurality of measurements arranged in individual loops by drawing a plurality of different loops of the light on the object to be inspected by the first scanning means and the second scanning means. The imaging apparatus according to any one of claims 1 to 8 , wherein an image of the object to be inspected is generated by allocating a plurality of the information obtained from the points to corresponding pixels in the display means. .. The image generation means shifts one of the annular images on the image display screen so that the correlation coefficient of the corresponding pixel in the annular image generated from each of the loops is high, so that each of the annular images The imaging apparatus according to claim 9 , wherein an image of the object to be inspected is generated by performing alignment and synthesizing the aligned annular images. Said first scanning means and the second scanning means each comprise a galvanometer scanner, any one of claims 1 to 10 wherein the first direction and the second direction and wherein the mutually orthogonal The imaging apparatus according to. The light is measurement light divided from the light emitted from the low coherence light source.
The image generating means includes the return light from the inspected object and the emitted light of the measurement light scanned on the inspected object via the first scanning means and the second scanning means. The imaging apparatus according to any one of claims 1 to 11 , wherein an image of the object to be inspected is generated by using the interference light obtained by combining the reference light which is further divided. The control means performs a reciprocating scan of the first scanning means before delaying the scanning of the measurement light of the second scanning means by the predetermined time with respect to the scanning of the light of the first scanning means. Repeat for multiple cycles,
The image generation means generates an image by using the information from the same cross section in the object to be inspected obtained by using the interference light obtained by repeating the scanning of the measurement light for a plurality of cycles. The imaging apparatus according to claim 12. The imaging device according to any one of claims 1 to 13 , wherein the object to be inspected is an eye. A first scanning means for reciprocating light in a first direction on an object to be inspected at a predetermined cycle, and a first scanning means.
In an imaging apparatus having a second scanning means for reciprocating scanning the light in a second direction different from the first direction at a cycle that is an integral multiple of the predetermined cycle on the object to be inspected.
A step of delaying the scanning of the light of the second scanning means by a predetermined time with respect to the scanning of the light of the first scanning means.
See containing and a step of generating an image of the inspection object by using the information obtained from the return light from the object to be inspected of the light,
In the delay step, when the drive waveform applied to the second scanning means has the maximum amplitude after the light is scanned by the second scanning means for a predetermined cycle, the delay is delayed by the predetermined time. Is given to the second scanning means. A program comprising causing a computer to execute each step of the imaging method according to claim 15.

Images