Fluorescence detection device and fluorescence detection method

Abstract

A scanning microscope disclosed herein comprises: a pulse light generation device (or a pulse light generation device) that emits excitation light toward a sample, wherein the excitation wavelength of the excitation light is switched for each pulse; and a detector that detects fluorescence from a cell mass S irradiated with the excitation light and outputs a detection signal. The sensor bandwidth of the detector is adjusted so that the detection signal attenuates within a pulse-off period of the excitation light.

Description

Technical Field

[0001] The present disclosure relates to a fluorescence detection device and a fluorescence detection method.

Background Art

[0002] Patent Document 1 discloses a multiphoton microscope that causes multiphoton excitation in which a plurality of excitation photons are simultaneously absorbed when irradiating a fluorescent molecule with excitation light.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] When observing the deep part of a sample with a multiphoton microscope, the laser intensity is increased as it goes deeper. Therefore, there is a high possibility of damaging the sample in deep observation. In order not to cause damage, it is necessary to keep the laser intensity as low as possible even in deep observation. However, when the laser intensity is lowered, the amount of detection signal becomes small, and there is a risk that appropriate observation cannot be performed.

[0005] The present disclosure has been made in view of the above circumstances, and an object thereof is to provide a fluorescence detection device and a fluorescence detection method that improve fluorescence observation accuracy by improving the amount of detection signal.

Means for Solving the Problems

[0006] (1) A fluorescence detection device according to one aspect of the present disclosure comprises an excitation light source that emits excitation light, the excitation wavelength of which is switched with each pulse, toward a sample, and a detection unit that detects fluorescence from a sample irradiated with excitation light and outputs a detection signal, wherein the detection band of the detection unit is adjusted so that the detection signal is attenuated during the pulse-off period of the excitation light.

[0007] In a fluorescence detection device according to one aspect of this disclosure, fluorescence from a sample irradiated with excitation light whose excitation wavelength switches with each pulse is detected by a detection unit, and a detection signal is output from the detection unit. In the fluorescence detection device according to one aspect of this disclosure, the detection bandwidth of the detection unit is adjusted so that the above-mentioned detection signal is attenuated within the pulse-off period of the excitation light. By adjusting the detection bandwidth so that the detection signal is attenuated within the pulse-off period (before the next pulse arrives), different pulses are temporally distinguished from each other, and the mixing of detection signals related to the detection of different pulses is avoided. As a result, a configuration that filters the signal according to wavelength to distinguish different pulses becomes unnecessary, and the reduction in the amount of detection signal that would be a problem if such a configuration were provided can be suppressed. As described above, the fluorescence detection device according to one aspect of this disclosure can improve the fluorescence observation accuracy by improving the amount of detection signal.

[0008] (2) In the fluorescence detection device described in (1) above, the detection band of the detection unit may be adjusted such that the combined time until the fluorescence decays and the time until the detection signal decays is shorter than the repetition interval of the excitation light pulses. With such a configuration, the time until the fluorescence and the detection signal decays can be made shorter than the repetition interval of the pulses, and the temporal separation of different pulses described above can be achieved more reliably.

[0009] (3) In the fluorescence detection device described in (1) or (2) above, the detection unit may detect fluorescence without passing it through a wavelength-separating filter. This avoids the reduction in the amount of detection signal caused by passing it through a filter, and makes it possible to appropriately improve the amount of detection signal.

[0010] (4) In the fluorescence detection device described in any one of the above items (1) to (3), the detection unit may consist of a single photosensor that detects fluorescence corresponding to the excitation wavelength of each pulse. With such a configuration, it is possible to achieve an improved detection signal quantity without providing a configuration that filters the signal according to the wavelength, while also providing a simpler configuration.

[0011] (5) In the fluorescence detection apparatus described in any one of the above items (1) to (4), the excitation light source emits excitation light by switching for each pulse between a first excitation wavelength to stimulate the sample and a second excitation wavelength to observe the effect of the light stimulation of the first excitation wavelength, and the detection unit does not detect light at the fluorescence detection timing from the sample corresponding to the pulse of the first excitation wavelength, but detects light at the fluorescence detection timing from the sample corresponding to the pulse of the second excitation wavelength. With such a configuration, it is possible to detect only the necessary fluorescence while avoiding the detection of stimulating fluorescence (removing stimulating fluorescence), and contribute to improving the detection limit.

[0012] (6) The fluorescence detection device described in any one of the above items (1) to (5) further comprises an A / D converter that converts the detection signal output from the detection unit into a digital signal, and the sampling frequency of the A / D converter may be synchronized with the repetition frequency of the excitation light pulse. With such a configuration, it is easier to obtain the necessary information from the analog signal corresponding to each pulse without providing a high-speed A / D converter, and high-speed signals can be processed at low cost.

[0013] (7) In the fluorescence detection apparatus described in (6) above, the A / D converter may be set to acquire the maximum value of the detection signal. As described above, by synchronizing the frequency and always acquiring the same position of the signal waveform, and further acquiring the maximum value of the detection signal, it is possible to more appropriately acquire the necessary information from the analog signal corresponding to each pulse.

[0014] (8) In the fluorescence detection device described in (6) above, the A / D converter may be set to acquire the integral value of the detection signal. When acquiring the maximum value as described above, there is a risk that the maximum value may not be acquired properly due to fluctuations in the fluorescence signal, for example. In this regard, by acquiring the integral value, the necessary information can be acquired more appropriately from the analog signal corresponding to each pulse.

[0015] (9) A fluorescence detection method according to one aspect of the present disclosure includes emitting excitation light, the excitation wavelength of which is switched with each pulse, toward a sample, and detecting fluorescence from the sample irradiated with the excitation light and outputting a detection signal, wherein the detection bandwidth is adjusted so that the detection signal decays within the pulse-off period of the excitation light. [Effects of the Invention]

[0016] According to one aspect of the present invention, a fluorescence detection device and a fluorescence detection method can be provided that improve fluorescence observation accuracy by improving the amount of detected signal. [Brief explanation of the drawing]

[0017] [Figure 1] Figure 1 is a schematic diagram showing a scanning microscope according to an embodiment. [Figure 2] Figure 2 illustrates the challenges of deep tissue imaging. [Figure 3] Figure 3 illustrates the excitation light and fluorescence of GFP and RFP. [Figure 4] Figure 4(a) illustrates an example of excitation in two different wavelength bands, Figure 4(b) illustrates the fluorescence signal corresponding to the excitation light in Figure 4(a), and Figure 4(c) illustrates the sensor bandwidth. [Figure 5] Figure 5 is a schematic diagram showing an example of the detector configuration for a comparative example. [Figure 6] Figure 6 is a graph illustrating how the entire fluorescence spectrum cannot be obtained due to the use of a bandpass filter. [Figure 7]FIG. 7(a) is a diagram for explaining an example of excitation in two wavelength bands, FIG. 7(b) is a diagram for explaining a fluorescence signal corresponding to the excitation light in FIG. 7(a), and FIG. 7(c) is a diagram for explaining a sensor band. [Figure 8] FIG. 8 is a diagram schematically showing a configuration example of the detector according to the present embodiment. [Figure 9] FIGS. 9(a) and (b) are graphs for explaining the conditions of the sensor band. [Figure 10] FIG. 10(a) is a graph showing a case where the sensor band is sufficiently narrower than the A / D converter band, FIG. 10(b) is a graph showing a case where the sensor band is widened, and FIG. 10(c) is a graph for explaining a method of acquiring only the maximum value of the detection signal. [Figure 11] FIGS. 11(a) and (b) are diagrams for explaining the effect of synchronizing the sampling frequency of the A / D converter with the repetition frequency of the pulse. [Figure 12] FIGS. 12(a) and (b) are graphs for explaining the acquisition of the integrated value of the detection signal. [Figure 13] FIG. 13 is a diagram for explaining a detection method in the case of application to optical stimulation control. [Figure 14] FIG. 14 is a block diagram showing a pulse light generation device. [Figure 15] FIG. 15(a) is a graph showing the time waveform of the ultrashort pulse light output from the oscillator in FIG. 14. FIG. 15(b) is a graph showing the spectrum of the ultrashort pulse light output from the oscillator in FIG. 14. FIG. 15(c) is a graph showing the time waveform of the ultrashort pulse light output from the fiber amplifier in FIG. 14. FIG. 15(d) is a graph showing the spectrum of the ultrashort pulse light output from the fiber amplifier in FIG. 14. [Figure 16] FIG. 16 is a graph showing a specific example of the spectrum of the ultrashort pulse light output from the fiber amplifier in FIG. 14. [Figure 17] FIG. 17(a) is a graph showing the time waveform of the ultrashort pulse light output from the acousto-optic modulator in FIG. 14. FIG. 17(b) is a graph showing the spectrum of the ultrashort pulse light output from the acousto-optic modulator in FIG. 14. [Figure 18] Figure 18(a) is a graph showing the time waveform of the ultrashort pulse light output from the soliton shift fiber in Figure 14. Figure 18(b) is a graph showing the spectrum of the ultrashort pulse light output from the soliton shift fiber in Figure 14. Figure 18(c) is a graph showing the time waveform of the ultrashort pulse light output from the filter in Figure 14. Figure 18(d) is a graph showing the spectrum of the ultrashort pulse light output from the filter in Figure 14. [Figure 19] Figure 19 illustrates chirp pulse amplification. [Figure 20] Figure 20(a) illustrates a method of creating a difference in optical path length by the difference in refractive index when light passes through a material, and Figure 20(b) illustrates a method of creating a difference in optical path length by diffraction. [Figure 21] Figure 21 is a schematic diagram of the configuration related to chirp pulse amplification. [Figure 22] Figure 22 is a schematic diagram of a configuration related to chirp pulse amplification that widens the time width using a single fiber. [Figure 23] Figure 23(a) is a graph showing the characteristics of the stretcher and compressor in the configuration shown in Figure 22, and Figure 23(b) is a graph showing the residual GVD in the configuration shown in Figure 22. [Figure 24] Figure 24 illustrates the generation of second harmonics. [Figure 25] Figure 25 is a graph showing the wavelength conversion efficiency in each wavelength band for each crystal length. [Figure 26] Figure 26 illustrates angular phase matching. [Figure 27] Figure 27 illustrates an example of the configuration of the wavelength conversion unit. [Figure 28] Figure 28 illustrates an example of the configuration of a wavelength conversion unit according to a modified example. [Figure 29] Figure 29 is a block diagram showing another example of a pulsed light generator. [Figure 30] This is a flowchart showing a wavelength tunable method. [Figure 31]Figure 31(a) is a graph illustrating an example of amplifying short-wavelength ultrashort pulse light. Figure 31(b) is a graph illustrating the continuation of Figure 31(a). Figure 31(c) is a graph illustrating the continuation of Figure 31(b). Figure 31(d) is a graph illustrating the continuation of Figure 31(c). [Figure 32] Figure 32(a) is a graph illustrating an example of amplifying long-wavelength ultrashort pulse light. Figure 32(b) is a graph illustrating the continuation of Figure 32(a). Figure 32(c) is a graph illustrating the continuation of Figure 32(b). Figure 32(d) is a graph illustrating the continuation of Figure 32(c). [Figure 33] Figure 33 is a graph showing the relationship between the absorption cross-section, stimulated emission cross-section, and filter transmittance of a Tm fiber and wavelength. [Figure 34] Figure 34 is a schematic diagram showing a modified scanning microscope (fluorescence detector). [Figure 35] Figure 35 shows the time waveforms of each signal. [Figure 36] Figure 36 is a diagram illustrating a delay generator. [Figure 37] Figure 37 is a schematic diagram showing a pulsed light generator (excitation light source) according to another modified example. [Figure 38] Figure 38 shows the time waveforms of each signal. [Modes for carrying out the invention]

[0018] The embodiments will be described in detail below with reference to the drawings. In each drawing, the same or corresponding parts are denoted by the same reference numerals, and redundant descriptions are omitted.

[0019] Figure 1 is a schematic diagram showing a scanning microscope (fluorescence detection device) 1 according to an embodiment. The scanning microscope 1 is, for example, a laser scanning microscope (LSM). Alternatively, the scanning microscope 1 may be a multiphoton microscope, for example, a two-photon microscope. The scanning microscope 1 irradiates a cell aggregate S (sample) with excitation light. The cell aggregate S has been stained, for example, with a fluorescent probe. Therefore, when the cell aggregate S is irradiated with excitation light, the cell aggregate S that have taken up the fluorescent probe emit fluorescence. In the above staining treatment, for example, green fluorescent protein (GFP) and red fluorescent protein (RFP) may be used as fluorescent molecules (fluorescent dyes), and the antibody of the cell aggregate S may be labeled with them. The scanning microscope 1 detects the fluorescence emitted by the cell aggregate S.

[0020] The scanning microscope 1 comprises a pulsed light generator 10 (excitation light source), a dichroic mirror 20, an objective lens 30, a focusing lens 40, a detector (detection unit) 50, an A / D converter 60, and a computer 70.

[0021] The pulsed light generator 10 is, for example, a laser device, and in this embodiment, a femtosecond laser device. The pulsed light generator 10 generates ultrashort pulsed light as excitation light L1. The pulsed light generator 10 performs wavelength conversion for each pulse. The pulsed light generator 10 only needs to have an output mode that performs wavelength conversion for each pulse, and in addition to that output mode, it may also have other output modes such as outputting pulsed light continuously at a single wavelength, outputting multiple pulsed lights for multiple wavelengths that have been converted, or outputting pulsed light at completely random wavelengths and numbers. That is, the pulsed light generator 10 emits excitation light L1, whose excitation wavelength is switched for each pulse, toward the cell mass S. Details of the pulsed light generator 10 will be described later.

[0022] The dichroic mirror 20 is a mirror that reflects light of a specific wavelength while transmitting light of other wavelengths. The dichroic mirror 20 reflects light in the wavelength range that can be used as excitation light L1 towards the objective lens 30. The dichroic mirror 20 also transmits fluorescence L2 from the cell aggregate S, which will be described later.

[0023] The objective lens 30 is positioned opposite the cell aggregate S and is a lens that focuses the excitation light L1, which has entered through the dichroic mirror 20, toward the focal point FP of the cell aggregate S. The fluorescence L2 generated from the focal point FP passes through the objective lens 30 and then through the dichroic mirror 20. The fluorescence L2 that has passed through the dichroic mirror 20 enters the detector 50 via the focusing lens 40. An excitation light cut filter (not shown) may be provided between the focusing lens 40 and the detector 50 to suppress the entry of the excitation light L1 into the detector 50.

[0024] The detector 50 is a photosensor that detects fluorescence L2 from a cell mass S irradiated with excitation light L1 and outputs a detection signal. The detector 50 is an electron tube that houses a photocathode and an electron multiplier unit in a vacuum housing, such as a photomultiplier tube (PMT) or an HPD (Hybrid Photo-Detector). A control circuit (not shown) is connected to the photomultiplier tube, and this control circuit includes a gain adjustment circuit that adjusts the power supplied to the high-voltage power supply and the electron multiplier unit. In this embodiment, the detector 50 is described as a photomultiplier tube, but is not limited to this, and may be a semiconductor photodetector such as a SiPM (Silicon Photomultiplier), an avalanche photodiode (APD), or a photodiode (PD), or an electron tube, image sensor, or camera using them. Details of the detector 50 will be described later.

[0025] The A / D converter 60 converts the detection signal (analog signal) output from the detector 50 into a digital signal and outputs the digital signal. Details of the A / D converter 60 will be described later.

[0026] The computer 70 performs predetermined control and signal processing on the digital signal (fluorescent signal) output from the A / D converter 60. The computer 70 may transfer the signal processing results to an external server, for example, via serial communication.

[0027] Next, we will explain the challenges of observing the deep tissues of cell aggregates S using multiphoton microscopes such as two-photon microscopes (deep imaging), with reference to Figures 2 to 6.

[0028] Figure 2 illustrates the challenges of deep imaging. Typically, when performing deep imaging of a cell aggregate S, the intensity of the excitation light L1 is increased as the focal point FP is set deeper within the cell aggregate S. However, deep observation carries the risk of damaging the sample, making it necessary to keep the laser intensity of the excitation light L1 as low as possible.

[0029] However, in devices that observe fluorescence, it is necessary to separate the fluorescence wavelength band using wavelength-selective filters such as bandpass filters. As a result, the detection signal tends to be small, and in some cases, the laser intensity cannot be sufficiently low from the standpoint of ensuring a sufficient detection signal. The reasons why the detection signal becomes small are explained below.

[0030] Figure 3 illustrates the excitation light and fluorescence of green fluorescent protein (GFP) and red fluorescent protein (RFP). Now, consider the case where GFP and RFP are used as fluorescent molecules (fluorescent dyes) to label an antibody on a cell aggregate S. The horizontal axis in Figure 3 represents wavelength. Figure 3 shows the excitation light L11 and corresponding fluorescence L21 when GFP is excited, and the excitation light L12 and corresponding fluorescence L22 when RFP is excited. The wavelength bands of fluorescence L21 and fluorescence L22 partially overlap.

[0031] Figure 4(a) shows the waveform of the excitation light when excitation light L11 and excitation light L12 are switched for each pulse. As shown in Figure 4(a), waveform 101 corresponding to excitation light L11 and waveform 102 corresponding to excitation light L12 appear sequentially for each pulse. In this case, as shown in Figure 4(b), waveform 201 corresponding to fluorescence L21 and waveform 202 corresponding to fluorescence L22 also appear sequentially in the same manner.

[0032] Figure 4(c) is a diagram illustrating the sensor bandwidth 301 of the detector (photosensor), which is the detection bandwidth of the photodetector. In this embodiment, the detection bandwidth of the photodetector refers to the bandwidth of the output signal extracted from the photodetector as a detection signal. In other words, a wide detection bandwidth means that the bandwidth of the output signal is wide, that is, the time from when a single output signal rises until it decays (the time from the start of the rise to the end of the fall of a single output signal) is short, and a narrow detection bandwidth means that the bandwidth of the output signal is narrow, that is, the time from when a single output signal rises until it decays is long. More specifically, the width and narrowness of the detection bandwidth are determined by the width and narrowness of the frequency bandwidth of the amplifier section in the circuit section of the photodetector that multiplies and converts the current output output from the photosensor into a voltage output. In other words, if the frequency bandwidth of the amplifier section is wide, the detection bandwidth becomes wide, and if the frequency bandwidth of the amplifier section is narrow, the detection bandwidth also becomes narrow. Therefore, in this embodiment, the description regarding the detection bandwidth of the photodetector can be replaced with the frequency bandwidth of the amplifier section that multiplies and converts the current output from the photosensor into a voltage output in the circuit section of the photodetector. Normally, the sensor bandwidth 301 of the detector (photosensor) is set to be much narrower than the repetition interval of the excitation light pulse (see Figures 4(a) to 4(c)). For this reason, it is not possible to temporally distinguish between the fluorescence L21 component and the fluorescence L22 component in the detection signal output from the photosensor (more specifically, the digital signal converted in the subsequent A / D converter). Thus, even when the excitation light L11 and excitation light L12 are switched for each pulse, the sensor bandwidth 301 of the photosensor is narrow, making it impossible to appropriately distinguish and acquire fluorescence L21 corresponding to excitation light L11 and fluorescence L22 corresponding to excitation light L12. Furthermore, even if fluorescence L21 and fluorescence L22 experience intensity fluctuations within a time shorter than the sensor bandwidth 301, this is not reflected in the detection signal, and therefore, this information cannot be acquired.

[0033] To appropriately distinguish and acquire fluorescence L21 and fluorescence L22, it is advisable to adopt a configuration with bandpass filters 503 and 603, as shown in Figure 5. In the example shown in Figure 5, for fluorescence, the dichroic mirror 400 reflects light in the 500 nm wavelength band and transmits light in the 600 nm wavelength band. The light reflected by the dichroic mirror 400 passes through lens 502 and bandpass filter 503 and is incident on photodetector 501. The light detected by photodetector 501 is GFP fluorescence. The light transmitted through the dichroic mirror 400 passes through lens 602 and bandpass filter 603 and is incident on photodetector 601. The light detected by photodetector 601 is RFP fluorescence. Thus, when the sensor bandwidth is narrow and it is not possible to distinguish fluorescence pulse by pulse in time, it is necessary to use a bandpass filter to decompose the fluorescence signal by wavelength.

[0034] Figure 6 is a graph illustrating how the entire fluorescence spectrum cannot be acquired due to the use of bandpass filters 503 and 603. In Figure 6, the horizontal axis represents wavelength and the vertical axis represents intensity. As indicated in the legend in Figure 6, the spectrum of fluorescence L21 of GFP, the wavelength range of bandpass filter 503 for detecting fluorescence L21, the spectrum of fluorescence L22 of RFP, and the wavelength range of bandpass filter 603 for detecting fluorescence L22 are shown, respectively. As shown in Figure 6, the spectrum range detected by photodetectors 501 and 601 is limited by bandpass filter 503 for fluorescence L21 of GFP, and by bandpass filter 603 for fluorescence L22 of RFP, respectively. Thus, because the entire fluorescence spectrum cannot be detected due to the use of bandpass filters 503 and 603, the amount of detected fluorescence signal becomes smaller. Furthermore, the amount of light transmitted through bandpass filters 503 and 603 is also reduced. In such a configuration, the laser intensity of the excitation light needs to be increased to a certain extent in order to secure the amount of fluorescence signal, and as a result, there is a problem that fluorescence generation from outside the focal point cannot be sufficiently suppressed. Furthermore, even if it is possible to distinguish between fluorescence L21 and fluorescence L22, if the sensor bandwidth 301 of the photosensor is narrow, if fluorescence L21 and fluorescence L22 experience intensity fluctuations within a time period shorter than the sensor bandwidth 301, this will not be reflected in the detection signal, and therefore, that information cannot be obtained.

[0035] Therefore, in the scanning microscope 1 according to this embodiment, a method is employed to distinguish fluorescence for each pulse without using a bandpass filter in a configuration that switches the excitation wavelength for each pulse. Specifically, in the scanning microscope 1, the sensor bandwidth (detection bandwidth) of the detector 50 described above is adjusted so that the detection signal (signal output from the detector 50) is attenuated during the pulse-off period of the excitation light L1. The pulse-off period is the off period in pulsed light that is repeatedly switched on and off. Attenuation of the detection signal means, for example, that the detection signal is attenuated to a signal amount of 10% or less of the maximum value.

[0036] As shown in Figure 7(a), waveform 101 corresponding to excitation light L11 and waveform 102 corresponding to excitation light L12 appear sequentially with each pulse, and as shown in Figure 7(b), waveform 201 corresponding to fluorescence L21 corresponding to excitation light L11 and waveform 202 corresponding to fluorescence L22 corresponding to excitation light L12 appear sequentially in a similar manner. Now, as shown in Figure 7(c), the sensor bandwidths 351 and 352 of the detector 50 (photosensor) are set to be sufficiently wide (equally wide) with respect to the repetition interval of the excitation light pulses. In Figure 7(c), sensor bandwidth 351 is the sensor bandwidth corresponding to fluorescence L21, and sensor bandwidth 352 is the sensor bandwidth corresponding to fluorescence L22. In Figure 7(c), sensor bandwidths 351 and 352 are set to have a 1:1 ratio with respect to waveform 201 corresponding to fluorescence L21 and waveform 202 corresponding to fluorescence L22 corresponding to excitation light L12. As shown in Figure 7(c), the sensor bandwidths 351 and 352 of the detector 50 are adjusted so that the detection signal (waveform of sensor bandwidths 351 and 352) attenuates within the pulse-off period of the excitation light. In this way, the detection signal of the detector 50 attenuates before the next pulse is detected.

[0037] Figure 8 is a schematic diagram showing an example of the configuration of the detector 50 according to this embodiment. As shown in Figure 8, the detector 50 is composed of a single photosensor 51 that detects fluorescence corresponding to the excitation wavelength of each pulse (for example, the fluorescence L21 and L22 described above). The photosensor 51 of the detector 50 detects fluorescence without using a bandpass filter for wavelength separation of fluorescence. This configuration without a bandpass filter is possible because the sensor bandwidth of the detector 50 is set so that the detection signal of the detector 50 attenuates within the pulse-off period of the excitation light, and fluorescence related to different pulses is distinguished in time.

[0038] More specifically, the sensor bandwidth of detector 50 is adjusted so that the combined time until fluorescence decays and the detection signal decays is shorter than the repetition interval of the excitation light pulses. Figures 9(a) and 9(b) are graphs illustrating the sensor bandwidth conditions. In Figures 9(a) and 9(b), the horizontal axis represents time and the vertical axis represents intensity. If we let frep be the repetition interval of the excitation light pulses (picking frequency by AOM), T be the fluorescence lifetime (the longest fluorescence lifetime in the case of multiple staining), FWHM be the full width at half maximum of the detection signal, and fsensor be the sensor bandwidth of detector 50 (the signal bandwidth of the sensor), then the following equation (1) can be derived from the above conditions for detector 50. Note that the fluorescence lifetime here is the time until the initial intensity of the afterglow decays to 1 / e. Since 1 / e is approximately 37%, here we use about four times that amount as the time until the fluorescence signal decays. 1 / frep>4T+2FWHM 1 / frep > 4T + 2 × 1 / (2 × fsensor) 1 / fsensor<1 / frep-4T (1) Thus, the sensor bandwidth of the detector 50 is set to fall within a range determined by the pulse repetition interval and fluorescence lifetime.

[0039] Here, if the sensor bandwidth of the detector 50 is sufficiently narrower than the A / D converter bandwidth, the detection signal from the detector 50 can be sampled without problems. That is, as shown in Figure 10(a), if the sensor bandwidth 301 is sufficiently narrower than the A / D converter bandwidth and many sampling points SP can be set for the detection signal, the detection signal can be sampled without problems and the analog signal can be appropriately reproduced from the A / D converter. In this regard, as described above, in the configuration according to this embodiment, as shown in Figure 10(b), the sensor bandwidth 351 of the detector 50 is wide, so if the A / D converter bandwidth remains the same as before, the detection signal from the detector 50 cannot be sampled sufficiently, and there is a risk that the analog signal cannot be appropriately reproduced from the A / D converter. To address this problem, for example, increasing the speed of the A / D converter can be considered, but this is technically difficult and would increase costs. Furthermore, in the scanning microscope 1, it is necessary to construct the overall image after scanning at a high frequency and acquiring a large amount of signals, so an increase in the number of samples due to the speed of the A / D converter means an enormous increase in the amount of data, which is practically fatal.

[0040] In this embodiment, the sampling frequency of the A / D converter 60 is synchronized with the repetition frequency of the excitation light pulse, thereby enabling sampling at the same position in the detection signal. Furthermore, it is preferable to provide an appropriate time delay to the sampling frequency of the A / D converter 60. In this case, the A / D converter 60 may be set to acquire the maximum value of the detection signal. That is, as shown in Figure 10(c), the sampling point SPM of the A / D converter 60 may be set to always be near the maximum value of the detection signal. By synchronizing the frequencies in this way, it is possible to perform sampling on the detector signal with the minimum number of steps without increasing the speed of the A / D converter 60 (in fact, by making it slower than conventional methods), which also contributes to data compression. Note that the maximum value of the detection signal in this embodiment is not limited to the peak value of the pulsed detection signal, but includes a range showing a signal value of 70% or more of the signal value shown by the peak.

[0041] Furthermore, by synchronizing the sampling frequency of the A / D converter 60 with the repetition frequency of the excitation light pulse, random noise (dark / stray light) can be eliminated. As shown in Figure 11(a), if the above-mentioned synchronization is not performed, the detection signal sampled at the sampling point SP will contain a component of random noise D. In this respect, as shown in Figure 11(b), if the above-mentioned synchronization is performed and, for example, only the maximum value of the detection signal is detected at the sampling point SPM, the random noise D is more likely to fall outside the detection range, and the random noise D can be appropriately eliminated. By eliminating random noise, the detection limit can be expanded.

[0042] The A / D converter 60 may be configured to acquire values ​​other than the maximum value of the detection signal, for example, it may be configured to acquire the integral value. Figures 12(a) and 12(b) are graphs illustrating the acquisition of the integral value of the detection signal. As shown in Figure 12(a), when the sensor bandwidth 351 is relatively wide, the jitter (fluctuation) of the fluorescence signal shown by the waveform 201 may have an effect, and the method of acquiring the maximum value of the detection signal described above may result in a large variation in the value. Therefore, for example, a method can be considered in which the detection signal is input to a charge amplifier circuit (not shown), the signal is charged for the time until the fluorescence signal decays, and the integral value is sampled at only one point. In the example shown in Figure 12(b), one point of the charge signal CS related to the integral value obtained by inputting the detection signal to the charge amplifier circuit (not shown) is set as the sampling point SPM. With such a configuration, the effect of jitter (fluctuation) in the fluorescence signal can be suppressed. By sampling as quickly as possible after the fluorescence signal decays, the effects of stray light and the like can be minimized.

[0043] Furthermore, the scanning microscope 1 may be used for observing various types of fluorescence other than those described above, for example, for fluorescence observation with photostimulation control. In this case, the pulsed light generator 10 emits excitation light by switching between a first excitation wavelength to stimulate the cell aggregate S and a second excitation wavelength to observe the effect of the photostimulation of the first excitation wavelength for each pulse. Then, as shown in Figure 13, the detector 50 may not detect light by not performing sampling of the A / D converter at the fluorescence detection timing from the cell aggregate S corresponding to the pulse of the first excitation wavelength (detection timing corresponding to the stimulation excitation), and may only detect light at the fluorescence detection timing from the cell aggregate S corresponding to the pulse of the second excitation wavelength (detection timing corresponding to the observation excitation). With this configuration, fluorescence corresponding to the stimulation excitation can be removed, so that the fluorescence corresponding to the observation excitation is not properly detected because it is buried by the fluorescence corresponding to the stimulation excitation. This can contribute to improving the detection limit.

[0044] Next, the details of the pulsed light generator 10, which switches the excitation wavelength for each pulse, will be explained with reference to Figures 14 to 33.

[0045] As shown in Figure 14, the pulsed light generator 10 of this embodiment generates long-wavelength ultrashort pulsed light (pulsed light) using soliton self-frequency shift (Raman soliton shift). For example, the pulsed light generator 10 is a femtosecond laser device for two-photon microscopy. The pulsed light generator 10 comprises an oscillator 702, a fiber amplifier 703, an acousto-optic modulator 704, a compressor 705, a soliton shift fiber 706, a stretcher fiber 707, a fiber amplifier 708, a compressor 709, and a wavelength conversion unit 710.

[0046] The oscillator 702 constitutes an oscillator that oscillates ultrashort pulse light. As shown in Figure 15(a), the oscillator 702 generates an ultrashort pulse train with a predetermined period F1. The wavelength band of the ultrashort pulse light oscillated from the oscillator 702 may include, for example, 1550 nm. Here, the oscillator 702 oscillates ultrashort pulse light having a spectrum with a first spectral width H1 and a first intensity K1, as shown in Figure 15(b). The oscillator 702 is not particularly limited, and various oscillators can be used.

[0047] The fiber amplifier 703 constitutes an amplification section that broadens the spectrum of the ultrashort pulse light oscillated by the oscillator 702. The fiber amplifier 703 broadens the spectrum of the ultrashort pulse light and increases the output power of the ultrashort pulse light by simillariton amplification. The fiber amplifier 703 is positioned between the oscillator 702 and the soliton-shift fiber 706 in the optical path of the ultrashort pulse light.

[0048] The fiber amplifier 703 includes a fiber amplifier. The fiber amplifier in the fiber amplifier 703 is a normally dispersed fiber, a double-clad fiber co-doped with erbium and ytterbium. That is, the fiber amplifier 703 performs amplification while causing a nonlinear effect using a normally dispersed double-clad fiber without stretching, and obtains ultrashort pulse light as broadband amplifier light. A normally dispersed fiber is a fiber with a negative dispersion parameter D (ps / nm / km). The additives used in the fiber amplifier 703 are not particularly limited, and various additives may be used.

[0049] As shown in Figures 15(c) and 15(d), the fiber amplifier 703 broadens the spectral width of the ultrashort pulse light to a second spectral width H2, which is wider than the first spectral width H1. The fiber amplifier 703 also increases the intensity of the ultrashort pulse light to a second intensity K2, which is higher than the first intensity K1. Specifically, as shown in Figure 16, the fiber amplifier 703 sets the spectral width of the ultrashort pulse light to 100 nm or more. In Figure 16, the horizontal axis represents the wavelength of the ultrashort pulse light, and the vertical axis represents the relative value of the intensity of the ultrashort pulse light relative to a predetermined intensity.

[0050] The acousto-optic modulator 704 constitutes an optical intensity control unit that controls the intensity of ultrashort pulse light pulse by pulse. The acousto-optic modulator 704 is a device that modulates ultrashort pulse light using the power of sound (sound waves), and is called an AOM (Acousto Optic Modulator). In this embodiment, the acousto-optic modulator 704 is placed between the fiber amplifier 703 and the soliton shift fiber 706 in the optical path of the ultrashort pulse light. The acousto-optic modulator 704 may be placed at any position between the oscillator 702 and the soliton shift fiber 706. As shown in Figures 17(a) and 17(b), the acousto-optic modulator 704 controls the intensity of the ultrashort pulse light so that it changes with each pulse. For example, as shown in Figure 17(a), when modulation of intensity M1 and M2 is applied, ultrashort pulse lights LM1 and LM2 corresponding to the intensities applied by M1 and M2 are generated, as shown in Figure 17(b). The intensity modulation range and accuracy of the ultrashort pulse light L(LM1,LM2) depend on the performance of the acousto-optic modulator 704. The intensity of each pulse in the ultrashort pulse light train can be arbitrarily modulated by the acousto-optic modulator 704.

[0051] The compressor 705 constitutes a pulse compression unit that compresses the pulse width of the ultrashort pulse light. In this embodiment, the compressor 705 is positioned between the acousto-optic modulator 704 and the soliton shift fiber 706 in the optical path of the ultrashort pulse light. The compressor 705 may be positioned at any location between the fiber amplifier 703 and the soliton shift fiber 706. Even if the ultrashort pulse light is stretched (for example by several picoseconds) by the fiber amplifier 703, the compressor 705 compresses the pulse width of the ultrashort pulse light and outputs ultrashort pulse light with a pulse width of less than a certain amount (less than 1 picosecond). The compressor 705 is not particularly limited, and various compressors can be used.

[0052] The soliton shift fiber 706 constitutes a modulation section that modulates the wavelength of ultrashort pulse light, whose spectrum has been broadened and output increased by the fiber amplifier 703, using soliton self-frequency shifting. The soliton shift fiber 706 is positioned downstream of the fiber amplifier 703 in the optical path of the ultrashort pulse light. As shown in Figures 18(a) and 18(b), the soliton shift fiber 706 lengthens the wavelength of the ultrashort pulse light and generates soliton S1. The wavelength band of the lengthened ultrashort pulse light L may include, for example, a band of 1800 nm to 2200 nm. The soliton shift fiber 706 can be, for example, a single-mode anomalous dispersion fiber that exhibits anomalous dispersion in the wavelength band of the ultrashort pulse light generated by the fiber amplifier 703. In addition, by controlling the acousto-optic modulator 704, it is also possible to generate solitons with wavelengths different from soliton S1. The wavelength of soliton S is shifted to wavelengths corresponding to the intensities M1 and M2, as shown in Figure 18(d), when modulated with intensities M1 and M2, as shown in Figure 18(c). The shift wavelength range and accuracy of soliton S depend on the performance of the acousto-optic modulator 704. The shift wavelength of each soliton S in the soliton train generated from the pulse train of ultrashort pulse light can be arbitrarily changed by applying intensity modulation to the pulse train with the acousto-optic modulator 704. In the illustrated example, the ultrashort pulse light modulated by soliton self-frequency shift includes a non-soliton component S0 (a component that did not become soliton S1 or S2). A filter (not shown) that cuts out the non-soliton component S0 of the ultrashort pulse light L may be provided downstream of the soliton shift fiber 6. Such a filter may have an OD value of 3 or more.

[0053] The stretcher fiber 707, fiber amplifier 708, and compressor 709 constitute a configuration related to chirp pulse amplification. The configuration related to chirp pulse amplification will be explained with reference to Figures 19 to 23.

[0054] First, let's explain chirp pulse amplification. Figure 19 is a diagram illustrating chirp pulse amplification. When strongly amplifying the output of ultrashort pulse light, a process called chirp pulse amplification is sometimes performed to avoid adverse effects due to nonlinearity caused by excessively high peak power. In chirp pulse amplification, the time width of the ultrashort pulse light is first widened (stretched), then the stretched ultrashort pulse light is amplified, and finally, the amplified ultrashort pulse light is compressed to its original time width (returned to the state before stretching). As shown in Figure 19, in chirp pulse amplification, first, pulse light I, which is the ultrashort pulse light input to the stretcher fiber 707, has its time width widened by the stretcher fiber 707. Then, pulse light II, which is the ultrashort pulse light whose time width has been widened by the stretcher fiber 707, is amplified by the fiber amplifier 708. Finally, pulse light III, which is the ultrashort pulse light amplified by the fiber amplifier 708, is compressed by the compressor 9, and its time width is returned to the state before stretching.

[0055] Next, we will explain two examples of methods for widening (or compressing) the time width in chirp pulse amplification. In both examples, the time width is widened (or compressed) by creating a difference in optical path length. Figure 20(a) illustrates a method of creating a difference in optical path length by the difference in refractive index when passing through a material, and Figure 20(b) illustrates a method of creating a difference in optical path length by diffraction. As shown in Figure 20(a), ultrashort pulse light in which light of multiple wavelengths overlaps with their phases aligned, when it passes through some material, the difference in refractive index for each wavelength causes a difference in the optical path length of each wavelength, and therefore the group delay time of each wavelength changes. By utilizing this, the time width can be widened or, conversely, compressed. The material here is, for example, a fiber. Also, as shown in Figure 20(b), the diffraction grating pair 91,92 (details will be described later) causes a difference in the optical path length of each wavelength, which changes the group delay time of each wavelength. This allows us to either widen or, conversely, compress the time frame.

[0056] In this type of chirp pulse amplification, when returning the expanded time width of ultrashort pulse light to its original state, it is preferable to compress it in the same way as when expanding the time width. From this perspective, it is conceivable to perform stretching and compression in the same way for both. That is, for example, when using a method that creates a difference in the optical path length of each wavelength due to the difference in refractive index by passing the light through a fiber (see Figure 20(a)), both the stretcher and the compressor are made of fiber, and when using a method that creates a difference in the optical path difference of each wavelength using a pair of diffraction gratings (see Figure 20(b)), both the stretcher and the compressor are made of a pair of diffraction gratings. In this way, by performing stretching and compression using the same type of optical element, the characteristics of stretching and compression (details will be described later) will match, and it will be possible to compress the ultrashort pulse light by the same amount as the expanded time width.

[0057] On the other hand, when optimizing stretchers and compressors from the standpoint of stability and throughput, it may be advantageous to use different configurations for the stretcher and compressor. For example, when stretching and compressing with fiber, stretching can be expected to have higher throughput (typically with almost zero loss) compared to a diffraction grating pair, but compression may result in nonlinear effects causing the resulting light to not have the expected characteristics. Specifically, complex dispersion may be carried on the ultrashort pulse light, making it impossible to compress the ultrashort pulse light completely. As a result, the energy that should be concentrated in the main pulse is dispersed to other parts, creating an undesirable state for generating nonlinear effects. On the other hand, when stretching and compressing with a diffraction grating pair, for compression, since the ultrashort pulse light propagates in a spatial system, the beam diameter increases and the energy density is relaxed, so it is expected that nonlinear effects will be less likely to occur. For stretching, a diffraction grating pair typically has a diffraction efficiency of about 90%, so in the case of a diffraction grating pair with multiple diffraction gratings, energy is lost with each diffraction, and the final throughput becomes about 60% of the original energy.

[0058] Considering these factors, it is conceivable to use a fiber as a stretcher and a pair of diffraction gratings as a compressor to ensure throughput while suppressing the occurrence of nonlinear effects. In this configuration, stretching and compression are performed in different ways (using different types of optical elements), it is more difficult to match the characteristics of stretching and compression (details will be described later) compared to a configuration where stretching and compression are performed in the same way (using the same type of optical elements), and it becomes difficult to return the compressed light to the form of the ultrashort pulse light before stretching. This becomes a more serious problem in broadband light sources and broadband tunable light sources. In order to solve this problem, the configuration of the chirp pulse amplification of the pulse light generator 10 according to this embodiment is provided with two fibers (first fiber 71 and second fiber 72) as stretcher fibers 707 as shown in Figure 21.

[0059] Figure 21 is a schematic diagram of the configuration related to chirp pulse amplification of the pulse light generator 10 according to this embodiment. As shown in Figure 21, the pulse light generator 10 includes a stretcher fiber 707 (stretcher), a fiber amplifier 708 (amplification unit), and a compressor 709 as the configuration related to chirp pulse amplification.

[0060] The stretcher fiber 707 is a stretcher that widens the time width of ultrashort pulsed light. The wavelength band of ultrashort pulsed light widened by the stretcher fiber 707 is, for example, 1800 nm to 2200 nm. The stretcher fiber 707 is composed of a first fiber 71 that widens the time width of ultrashort pulsed light with a first characteristic and a second fiber 72 that widens the time width of ultrashort pulsed light with a second characteristic different from that of the first fiber 71. The first fiber 71 and the second fiber 72 are configured to widen the time width of ultrashort pulsed light by creating a difference in the optical path length for each wavelength due to the difference in refractive index of each wavelength when passing ultrashort pulsed light through them. The first fiber 71 widens the time width of ultrashort pulsed light output from the soliton shift fiber 706, for example, including the wavelength band of 1800 nm to 2200 nm, with the first characteristic and outputs it to the second fiber 72. The first fiber 71 may be, for example, a normal dispersion fiber. The second fiber 72 is connected to the first fiber 71 and widens the time width of the ultrashort pulse light input from the first fiber 71 using the second characteristic, and outputs it to the fiber amplifier 708. The second fiber 72 may be, for example, a normally dispersed or anomalously dispersed fiber. Details of the first and second characteristics will be described later.

[0061] The fiber amplifier 708 amplifies (increases the power output of) pulsed light whose time width has been widened by the stretcher fiber 707. The fiber amplifier 708 includes a fiber amplifier. The fiber amplifier of the fiber amplifier 708 is an anomalous dispersion fiber, for example, a thulium-doped fiber. The laser medium added to the fiber of the fiber amplifier 708 is not particularly limited and may be a rare earth element such as ytterbium, erbium, neodymium, or Bi. The wavelength range of the ultrashort pulse light L amplified by the fiber amplifier 708 is, for example, 1800 nm to 2200 nm. The fiber amplifier 708, which can reliably amplify the pulsed light in a broadband wavelength range, comprises, for example, a first fiber amplifier (not shown) having a high gain G1 (not shown) on the first wavelength side, which is the short wavelength side of the ultrashort pulse light L, and a second fiber amplifier (not shown) having a high gain G2 (not shown) on the second wavelength side, which is the long wavelength side. A filter that attenuates the amplified light of noise generated by ASE and soliton self-frequency shift is combined between these amplifiers. As an example, the short wavelength (first wavelength) is a wavelength between 1800 nm and less than 2000 nm, and the long wavelength (second wavelength) is a wavelength between 2000 nm and 2200 nm. For example, in the ultrashort pulse light L containing the wavelength band of 1800nm ​​to 2200nm, the light on the second wavelength side generates amplified light from noise caused by ASE and soliton self-frequency shift during amplification by the first fiber amplifier described above. This not only prevents sufficient amplification in the second fiber amplifier described above, but can also cause parasitic oscillations. However, these effects can be suppressed by the filter provided between the fiber amplifiers, so that the light is reliably amplified and output to the compressor 709.

[0062] The compressor 709 compresses the time width of the ultrashort pulse light amplified by the fiber amplifier 708. The compressor 709 has diffraction grating pairs 91 and 92. The diffraction grating pairs 91 and 92 are configured to compress the time width by changing the group delay time for each wavelength by creating differences in the optical path length for each wavelength. The wavelength range of the ultrashort pulse light for which the diffraction grating pairs 91 and 92 compress the time width is, for example, 1800 nm to 2200 nm.

[0063] Next, we will explain how to determine the first characteristics of the first fiber 71, the second characteristics of the second fiber 72, and the characteristics (compression characteristics) of the diffraction grating pair 91,92.

[0064] The first and second characteristics described above determine how the ultrashort pulse light L spreads. The compression characteristics determine how the ultrashort pulse light is compressed. The first characteristics, the second characteristics, and the compression characteristics are represented, for example, by information related to group velocity dispersion (GVD). Group velocity is the velocity of the wave packet related to the ultrashort pulse light. Group velocity dispersion is the wavelength dispersion of the group velocity. Information related to group velocity dispersion may be the group velocity dispersion itself, or it may be the group delay dispersion (GDD) obtained by multiplying the group velocity dispersion by the length of the medium (fiber length or distance between diffraction grating pairs), or it may be the TOD / GVD ratio, which is the ratio of the group velocity dispersion to the third-order group delay dispersion (TOD) obtained by differentiating the group velocity dispersion with respect to frequency.

[0065] Each of the characteristics described above can be determined, for example, by the following procedure. First, the residual group delay dispersion is derived when, in a configuration without a second fiber 72 as shown in Figure 22, the time width of the ultrashort pulse light is widened by the first fiber 71, the ultrashort pulse light is amplified by the fiber amplifier 708, and the time width of the ultrashort pulse light L is compressed by the diffraction grating pair 91,92. In this case, the first characteristic of the first fiber 71 is the group delay dispersion G1 of the first fiber 71 (see Figure 23(a)), which is determined by arbitrarily determining its fiber length. The characteristic of the diffraction grating pair 91,92 is the group delay dispersion G2 of the diffraction grating pair 91,92 (see Figure 23(a)), which is determined by arbitrarily determining the distance l between the diffraction grating pair 91,92 (see Figure 20(b)). Specifically, the group delay dispersion G2 of the diffraction grating pair 91,92 may be derived by the following equation (2). In equation (2) below, λ is the wavelength, c is the speed of light, d is the lattice constant, l is the distance between the diffraction grating pair 91 and 92, and θ is the angle of incidence of light. Group delay dispersion G2(λ)=-λ3l / πc2d2[1-(λ / d-sinθ)2]...(2)

[0066] Then, the group delay dispersion G3 (see Figure 23(b)) is derived by adding the group delay dispersion G1 of the first fiber 71 and the group delay dispersion G2 of the second fiber 72. In Figures 23(a) and 23(b), the horizontal axis is wavelength and the vertical axis is the value of the group delay dispersion. As shown in Figure 23(b), the value of the residual group delay dispersion G3 is 0 for the portion of the ultrashort pulse light at wavelength X1, indicating that the ultrashort pulse light has been restored to its original state (returned to the state before stretching) and has been properly compressed. On the other hand, as shown in Figure 23(b), the value of the residual group delay dispersion G3 for the portion of the ultrashort pulse light at wavelength X2 is a value that deviates from 0, indicating that the ultrashort pulse light has not been restored to its original state (returned to the state before stretching) and has not been properly compressed.

[0067] Then, with the first characteristic, group delay variance G1, the compression characteristic, group delay variance G2, and the residual group delay variance G3 determined, the second characteristic of the second fiber 72, the group delay variance G4 of the second fiber 72, is set so that the value of the residual group delay variance G3 approaches a desired value (for example, 0). Specifically, the group delay variance G4 may be the residual group delay variance G3 multiplied by -1. The group delay variance G4 is determined, for example, by the fiber length of the second fiber 72.

[0068] As described above, the second characteristic of the second fiber 72 is set such that, in a configuration where the second fiber 72 is absent, the value of the residual group delay dispersion G3 (an index related to group velocity dispersion) for each wavelength remaining in the ultrashort pulse light after the time width has been widened by the first fiber 71 and then compressed by the diffraction grating pair 91,92 approaches a predetermined value, preferably approaching 0.

[0069] Returning to Figure 14, the wavelength conversion unit 710 is configured to perform wavelength conversion on a pulse-by-pulse basis for the ultrashort pulse light output from the compressor 709. The wavelength conversion unit 710 may, for example, perform wavelength conversion to the 900 nm to 1100 nm wavelength range for each pulse.

[0070] The wavelength conversion unit 710 converts, for example, light in the range of 1800 nm to 2200 nm to 900 nm to 1100 nm, which is usable in a two-photon microscope. The wavelength conversion unit 710 may also perform the above wavelength conversion by second harmonic generation, a phenomenon included in nonlinear effects. Second harmonic generation is a phenomenon in which light is converted to half the wavelength (second harmonic) of the original wavelength (fundamental wave).

[0071] This section explains the conversion from the fundamental wave to the second harmonic in second harmonic generation. Figure 24 illustrates the generation of the second harmonic. As shown in Figure 24, when a high-intensity ultrashort pulse light, which is the fundamental wave, is incident on a crystal (nonlinear optical medium) that acts as a wavelength conversion element, the second harmonic, which is the harmonic component of electron oscillation, is generated by polarization oscillation. In detail, the harmonics (second harmonic) are generated at each point in the crystal through which the ultrashort pulse light passes. The sum of these second harmonics generated at each point becomes the final second harmonic.

[0072] In the example shown in Figure 24, the phases of the second harmonics at each point are aligned, but the phases of the second harmonics at each point are not always aligned. Because the reaction occurs sequentially, with the crystal responding to the fundamental wave and the second harmonics being generated as a result, differences in the crystal's response are reflected in the phase of the second harmonics, causing the phases of the second harmonics at each point to be misaligned. When the second harmonics at each point are added together while their phases are not aligned, the harmonics may cancel each other out or superimpose, resulting in the summed second harmonics not being sufficiently large. Therefore, it is important to determine the crystal conditions so that the phases of the second harmonics generated at each point are aligned as much as possible (phase matching). Phase matching is achieved when the refractive index of the fundamental wave and the refractive index perceived by the harmonics are equal.

[0073] Figure 25 is a graph showing the wavelength conversion efficiency in each wavelength band for nonlinear optical crystals such as single-crystal BBO (Beta-BaB2O4) for different crystal lengths. In Figure 25, the horizontal axis represents the wavelength band, and the vertical axis represents the normalized value of the wavelength conversion efficiency in second-harmonic generation (SHG). In Figure 25, the solid line shows the wavelength conversion efficiency in each wavelength band for a relatively thick crystal with a crystal length of 1 mm, and the dashed line shows the wavelength conversion efficiency in each wavelength band for a relatively thin crystal with a crystal length of 0.2 mm. Note that the maximum value of the wavelength conversion efficiency is higher for thicker crystals, but in Figure 25, the wavelength conversion efficiency is normalized, so the upper limit of the wavelength conversion efficiency is 1 for all crystals. As shown in Figure 25, the longer the crystal length (thicker the crystal), the stricter the conditions for the superposition of the second harmonics at each point, and the narrower the wavelength band in which high-efficiency wavelength conversion is possible. Therefore, from the viewpoint of widening the wavelength band that satisfies phase matching, it is advisable to use a relatively thin crystal.

[0074] However, since the intensity of the second harmonic is proportional to the square of the crystal length, it is difficult to increase the wavelength conversion efficiency with relatively thin crystals. The intensity of the second harmonic also depends on the intensity of the fundamental wave, so it is conceivable to increase the intensity of the fundamental wave to ensure the intensity of the second harmonic, but amplifying the fundamental wave intensity to compensate for the effect of the square of the crystal length mentioned above is not advisable from a cost standpoint. It is also conceivable to increase the intensity of the fundamental wave spatially by focusing it on a very small region of the crystal, but focusing the fundamental wave to a small area will also produce third-order and higher nonlinear effects, and these nonlinear effects will inhibit the generation of second harmonics, resulting in a decrease in wavelength conversion efficiency. In addition, if the intensity is high, the crystal itself may be damaged. For the above reasons, it is difficult to sufficiently increase the wavelength conversion efficiency with the relatively thin crystals mentioned above.

[0075] Therefore, in order to achieve broadband phase matching, which is a challenge for thick crystals, while using relatively thick crystals, a configuration can be considered in which a drive system is provided to change the angle of incidence of light in the crystal. Figure 26 is a diagram illustrating angular phase matching. In the configuration shown in Figure 26, a birefringent crystal 500 is prepared, and the refractive index is adjusted by adjusting the angle of incidence of light in the crystal 500 using a drive system (not shown), thereby achieving broadband phase matching.

[0076] However, in a configuration that includes a drive system as shown in Figure 26, it is impossible to drive the drive system for each pulse of ultrashort pulsed light (for example, every tens of MHz), making it difficult to use as a configuration for wavelength conversion of ultrashort pulsed light.

[0077] As described above, conventionally, it has been difficult to perform wavelength conversion with high efficiency in configurations that perform wavelength conversion over a wide bandwidth in a short time, such as when performing wavelength conversion of ultrashort pulse light in which the wavelength is changed with each pulse.

[0078] To address these challenges, the wavelength conversion unit 710 according to this embodiment performs high-efficiency wavelength conversion using a relatively thick crystal, and without using a drive system, it disperses light for each wavelength band and changes the optical path for each wavelength band, thereby performing broadband wavelength conversion even for ultrashort pulse light. A detailed configuration example of the wavelength conversion unit 710 will be described below with reference to Figure 27.

[0079] Figure 27 illustrates an example of the configuration of the wavelength conversion unit 710. As shown in Figure 27, the wavelength conversion unit 710 comprises a diffraction grating (dispersion unit) 1101, a lens (focusing unit) 1102, a wavelength conversion element (wavelength conversion unit) 1103, a lens 1104, and a diffraction grating 1105. The diffraction grating 1101 and lens 1102 are configured as a wavelength-specific optical path setting unit that changes the optical path for each wavelength band so that light of the wavelength band to be converted at each incident position (details will be described later) is incident at each incident position of the wavelength conversion element 1103.

[0080] The diffraction grating 1101 disperses the ultrashort pulsed light output from the compressor 709 into wavelength bands (spectroscopy the ultrashort pulsed light), thereby changing the optical path for each wavelength band. Specifically, the diffraction grating 1101 disperses the ultrashort pulsed light into wavelength bands so that light of the wavelength band to be converted at each incident position of the wavelength conversion element 1103 is incident at each incident position.

[0081] Lens 1102 is a lens that focuses the light from the diffraction grating 1101 to the incident position of the wavelength conversion element 1103. The distance from the diffraction grating 1101 to lens 1102, and the distance from lens 1102 to wavelength conversion element 1103 are both set to match the focal length f of lens 1102, for example, the focal length f of lens 1102.

[0082] The wavelength conversion element 1103 is a crystal in which different wavelength bands are converted depending on the incident position of light. The wavelength conversion element 1103 has a fan-shaped structure so that pseudo-phase matching of different wavelength bands is achieved at each incident position. More specifically, the wavelength conversion element 1103 may be a fan-shaped PPLN (Periodically Poled Lithium Niobate). That is, the wavelength conversion element 103 may be a so-called fan-out PPLN. The crystal length (thickness) of the wavelength conversion element 1103 may be, for example, about 1 mm to 20 mm.

[0083] Pseudo-phase matching is a technique that simulates phase matching by preventing the generation of a desired harmonic among the summed harmonics, or by inverting the phase of the desired harmonic. Here, we will explain an example of pseudo-phase matching in which the phase of the desired harmonic is inverted (the polarity of the crystal is reversed).

[0084] The wavelength conversion element 1103, due to its fan-shaped structure, allows for pseudo-phase matching of any wavelength band at each incident position in the crystal. That is, since the period of the region where the second harmonics cancel each other out differs depending on the wavelength band, by using a crystal with a reversed period according to the fan-shaped structure, the wavelength bands for which pseudo-phase matching is achieved at each incident position in the crystal can be made to differ from one another. In such a wavelength conversion element 1103, since the wavelength band for which pseudo-phase matching is achieved at each incident position can be predetermined, the diffraction grating 1101 and lens 1102 described above can be set so that light of the wavelength band to be converted is incident at each incident position, thereby enabling appropriate wavelength conversion for light of various wavelength bands.

[0085] Lens 1104 is a lens that focuses the light whose wavelength has been converted by the wavelength conversion element 1103 onto the diffraction grating 1101. The distance from the wavelength conversion element 1103 to lens 1104, and the distance from lens 1104 to diffraction grating 1105 are both set to match the focal length f of lenses 1102 and 1104, for example, the focal length f of lenses 1102 and 1104.

[0086] The diffraction grating 1105 outputs ultrashort pulsed light after wavelength conversion, following the dispersion of light by wavelength band by the diffraction grating 1101, and then passing through lens 1102, wavelength conversion element 1103, and lens 1104. The number of grooves in the diffraction grating 1105 may be 2N, which is twice the number of grooves N of the diffraction grating 1101, in order to restore the dispersion of the light after wavelength conversion.

[0087] The configuration of the wavelength conversion unit 710 is not limited to the above embodiment, and may be, for example, the configuration shown in Figure 28. As shown in Figure 28, the modified wavelength conversion unit comprises a lens 1201, a BBO (Beta-BaB2O4) 1202 (first conversion element), a lens 1203, a lens 1204, a BBO 1205 (second conversion element), and a lens 1206. Lens 1201 is a lens that focuses light onto the BBO 1202. Lenses 1203 and 1204 are lenses that focus light from the BBO 1202 onto the BBO 1205. Lens 1206 is a lens that focuses light from the BBO 1205 and emits it.

[0088] BBO1202 and 1205 are nonlinear optical crystals that perform wavelength conversion. BBO1202 is positioned at an angle (phase matching angle) suitable for wavelength conversion in a first wavelength band. BBO1205 is connected in series with BBO1202 and is positioned at an angle (phase matching angle) suitable for wavelength conversion in a second wavelength band different from the first wavelength band.

[0089] With this configuration, a simple setup of connecting BBO1202 and 1205 in series allows for wavelength conversion in the wavelength bands corresponding to each BBO1202 and 1205. In this configuration, wavelength conversion can be performed appropriately for multiple wavelength bands without using a drive system to move the BBO1202 and 1205. Because wavelength conversion can be performed without a drive system, it is possible to perform wavelength conversion appropriately even when performing wavelength conversion in an extremely short time, such as with ultrashort pulse light (when the movement of the wavelength conversion element by a drive system would not be fast enough for wavelength conversion). Furthermore, in this configuration, the shape of BBO1202 and 1205, such as the thickness of BBO1202 and 1205, can be appropriately set to perform wavelength conversion efficiently, thus enabling highly efficient wavelength conversion. As described above, the wavelength conversion unit according to this modified example can perform wavelength conversion efficiently in a configuration that performs broadband wavelength conversion in a short time.

[0090] While it is possible to connect BBO1202 and 1205 in parallel rather than in series, this would result in wavelength bands becoming unusable when dividing the wavelength band. In other words, light in the wavelength band between the first and second wavelength bands would become unusable. Therefore, from the perspective of using a wide bandwidth, it is preferable to connect BBO1202 and 1205 in series. In the example above, there were two types of BBOs, but the number of BBOs may be increased according to the number of wavelength bands to be converted.

[0091] The configuration of the pulsed light generator is not limited to the above-described embodiment; for example, the pulsed light generator 10A shown in Figure 29 may be used. Below, embodiments of the pulsed light generator 10A will be described with reference to Figures 29 to 33.

[0092] As shown in Figure 29, the pulsed light generator 10A comprises an oscillator 702, a first fiber amplifier 753, an acousto-optic modulator 704, a compressor 705, a soliton shift fiber 706, a second fiber amplifier 757, a filter 758, a third fiber amplifier 759, and a wavelength conversion unit 710. Below, the first fiber amplifier 753, the second fiber amplifier 757, the filter 758, and the third fiber amplifier 759, which have a different configuration from the pulsed light generator 10 described above, will be explained.

[0093] The first fiber amplifier 753 constitutes an amplification section that broadens the bandwidth of the spectrum of the ultrashort pulse light oscillated by the oscillator 702. The first fiber amplifier 753 broadens the bandwidth of the spectrum of the ultrashort pulse light and increases the output power of the ultrashort pulse light by similariton amplification. The first fiber amplifier 753 is positioned between the oscillator 702 and the soliton shift fiber 706 in the optical path of the ultrashort pulse light.

[0094] The first fiber amplifier 753 comprises a normally dispersed fiber and an excitation light source. The normally dispersed fiber is a double-clad fiber co-doped with erbium and ytterbium. That is, the first fiber amplifier 753 performs amplification while causing a nonlinear effect using a normally dispersed double-clad fiber without stretching, thereby obtaining ultrashort pulse light as broadband amplifier light. The normally dispersed fiber is a fiber with a negative dispersion parameter D (ps / nm / km). The dopants used in the first fiber amplifier 753 are not particularly limited, and various dopants may be used.

[0095] The second fiber amplifier 757 constitutes the first optical amplifier section, which amplifies ultrashort pulse light whose wavelength is tuned by the soliton-shift fiber 706. The second fiber amplifier 757 is composed of a rare-earth doped fiber and an excitation light source. The rare-earth doped fiber is, for example, a Tm fiber. The excitation light source is, for example, a laser diode that outputs excitation light with a wavelength of 1550 nm. The second fiber amplifier 757 is an amplifier optimized for the short wavelength side. The second fiber amplifier 757 amplifies by the absorption of gain by the ultrashort pulse light, which is the signal light (seed light).

[0096] The second fiber amplifier 757 has a higher gain G1 (see Figure 31(b)) on the first wavelength side, which is the short wavelength side of the ultrashort pulse light, than on the second wavelength side, which is the long wavelength side. Such characteristics of the second fiber amplifier 757 can be achieved, for example, by adjusting the length of the rare earth-doped fiber and the concentration of the added rare earth. For example, the short wavelength (first wavelength) is a wavelength between 1800 nm and 2000 nm, and the long wavelength (second wavelength) is a wavelength between 2000 nm and 2200 nm.

[0097] Filter 758 constitutes a filter section that filters the ultrashort pulse light amplified by the second fiber amplifier 757. Filter 758 attenuates ASE (Amplified Spontaneous Emission). In this embodiment, in addition to attenuating the ASE generated by amplification by the second fiber amplifier 757, filter 758 also attenuates light that has been amplified with noise components generated by amplification by the second fiber amplifier 757, as well as light with wavelengths in the high-gain region. In this embodiment, light with wavelengths in the high-gain region refers to light in a predetermined wavelength band at short wavelengths, which has a particularly high amplification ratio compared to light in other wavelength bands. If this light is not attenuated, only this light will be preferentially amplified, hindering optical amplification over a wide wavelength range. The ASE that filter 758 attenuates here is ASE in a predetermined wavelength band that includes the wavelength range between short and long wavelengths. Light with amplified ASE and noise components not only hinders amplification but can also generate parasitic oscillations and potentially damage the amplifier system. ASE tends to occur, for example, when the wavelength bands of the seed light and the gain do not match, or when the power of the seed light is weak. In addition, in filter 758, pulses in the short wavelength region are absorbed and become the gain of the following pulses in the long wavelength region, and there is also a process of energy transfer from pulses in the gain region to pulses in the non-gain region of the gain fiber. Filter 758 may be one that utilizes an AOM (acousto-optic modulator) or EOM (electro-optic modulator), or a gain fiber such as a Tm fiber, Ho fiber, or Tm / Ho fiber that does not receive excitation light. Furthermore, filter 758 may be a gain fiber in the gain fiber of the second fiber amplifier 757 in the region where the excitation light is attenuated by 20 dB or more.

[0098] The third fiber amplifier 759 constitutes a second optical amplifier that amplifies the ultrashort pulse light filtered by the filter 758. The third fiber amplifier 759 is composed of a rare-earth doped fiber and an excitation light source. The rare-earth doped fiber is, for example, a Tm fiber. The excitation light source is, for example, a laser diode that outputs excitation light with a wavelength of 790 nm or a CW laser that outputs excitation light with a wavelength of 1550 nm. The third fiber amplifier 759 is an amplifier optimized for the long-wavelength side. The third fiber amplifier 759 amplifies by the absorption of gain by the ultrashort pulse light L as the signal light. The third fiber amplifier 759 has a higher gain G2 (see Figure 31(d)) on the second wavelength side, which is the long-wavelength side of the ultrashort pulse light L, than on the first wavelength side, which is the short-wavelength side. Such characteristics of the third fiber amplifier 759 can be achieved, for example, by adjusting the length of the rare-earth doped fiber and the concentration of the added rare earth.

[0099] Next, the wavelength tuning method performed using the pulsed light generator 10A will be explained with reference to the flowchart in Figure 30.

[0100] First, the oscillator 702 oscillates ultrashort pulse light L to generate an ultrashort pulse train with a predetermined period (oscillation step: step S1). The first fiber amplifier 753 increases the output power of the ultrashort pulse light L and broadens the spectrum of the ultrashort pulse light L (amplification step: step S2). The acousto-optic modulator 704 controls the intensity of the ultrashort pulse light L pulse by pulse according to, for example, the required specifications or conditions (step S3). The compressor 705 compresses the time width of the ultrashort pulse light L (step S4). The soliton shift fiber 706 modulates and varies the wavelength of the ultrashort pulse light L with a broadband spectrum using soliton self-frequency shift (step S5).

[0101] Next, the ultrashort pulse light L, whose wavelength was variable in step S5, is amplified by the second fiber amplifier 757 (step S6). In step S6, the ultrashort pulse light L is amplified with a higher gain G1 on the short wavelength side than on the long wavelength side. The ultrashort pulse light L amplified in step S6 is filtered by the filter 758 (step S7). This reduces the ASE in a predetermined wavelength band generated by the amplification by the second fiber amplifier 757, the amplified noise component, and the light in the high-gain region.

[0102] Next, the ultrashort pulse light L filtered in step S7 is amplified by the third fiber amplifier 759 (step S8). In step S8, the ultrashort pulse light L is amplified with a higher gain G2 on the long wavelength side than on the short wavelength side. Then, the wavelength conversion unit 710 converts the wavelength of the amplified ultrashort pulse light L to the target wavelength (step S9).

[0103] In the pulsed light generator 10A described above, for example, when the wavelength of the ultrashort pulse light L is varied to a short wavelength λ1, as shown in Figure 31(a), the following effects are achieved. That is, as shown in Figure 31(b), the ultrashort pulse light L with a short wavelength λ1 is amplified by the second fiber amplifier 757 having a gain G1 without generating ASE (preamplifier). The preamplified ultrashort pulse light L with a short wavelength λ1 passes through the filter 758 without being particularly attenuated, as shown in Figure 31(c). Then, the ultrashort pulse light L that has passed through the filter 758 is further amplified by the third fiber amplifier 759 having a gain G2, as shown in Figure 31(d). In this case, although the gain G2 is higher on the longer wavelength side, the ultrashort pulse light L with a short wavelength λ1 is already amplified by the second fiber amplifier 757 and therefore easily absorbs energy. Consequently, the third fiber amplifier 759 also amplifies it while suppressing the generation of ASE (without worrying about ASE).

[0104] On the other hand, for example, as shown in Figure 32(a), when the wavelength of the ultrashort pulse light L is varied to a longer wavelength λ2, the following effects are achieved. That is, as shown in Figure 32(b), the ultrashort pulse light L with a longer wavelength λ2 is not amplified as much as the ultrashort pulse light L with a shorter wavelength λ1 in the preamplifier of the second fiber amplifier 757 which has a gain G1, and ASE11 is generated. Subsequently, as shown in Figure 32(c), ASE11 is attenuated in the filter 758. Then, the ultrashort pulse light L with a longer wavelength λ2 is amplified by the third fiber amplifier 759 which has a gain G2, without generating ASE. Therefore, according to this embodiment, in a wavelength tuning method in which the wavelength of the output ultrashort pulse light L is variable, it is possible to reliably amplify the ultrashort pulse light L while suppressing ASE11. This is particularly effective when performing optical amplification over a wide wavelength band or when performing optical amplification in a wavelength band where it is difficult to use an optical amplifier suited to the conditions.

[0105] Figure 33 is a graph showing the relationship between the absorption cross-section, stimulated emission cross-section, and transmittance of the filter and wavelength of the Tm fiber. In Figure 33, the horizontal axis represents the wavelength of the ultrashort pulse light L, and the vertical axis represents the absorption cross-section, stimulated emission cross-section, and transmittance of the filter 758 of the Tm fiber. When the Tm fiber amplifies light with a wavelength of 1800 to 2200 nm, it has the absorption cross-section 922 and stimulated emission cross-section 921 shown in Figure 33. In a fiber amplifier system including a Tm fiber, if the fiber length is short, the gain on the long wavelength side is insufficient, making ASE (Acoustic Emission Syndrome) more likely to occur when amplifying long wavelength light. On the other hand, if the fiber length is long, absorption is also large, so the short wavelength light is absorbed too much. Therefore, ASE is more likely to occur when amplifying short wavelength light. In this embodiment, the fiber amplifier including the Tm fiber is arranged in two stages, with a second fiber amplifier 757 and a third fiber amplifier 759 provided. Furthermore, by inserting a filter 758 with a transmittance of 923 (dotted line in the figure) between the second fiber amplifier 757 and the third fiber amplifier 759, ASE is removed, enabling amplification of broadband tunable laser light.

[0106] In this embodiment, broadband ultrashort pulse light L can be amplified without using a separate mechanism. Broadband ultrashort pulse light L can be amplified without utilizing nonlinear optical effects such as Raman amplification. This embodiment can configure a broadband and high-power (e.g., 30 dBm or more) amplifier system. According to this embodiment, parasitic oscillation due to increased ASE can also be prevented. In this embodiment, as described above, a tunable light source is used in a two-photon microscope. In this case, high-power amplification is possible in a broadband wavelength range, and together with the subsequent wavelength conversion system, high-power and multi-wavelength imaging becomes possible in the two-photon microscope.

[0107] Finally, the effects and advantages of the configuration according to this embodiment will be described.

[0108] The scanning microscope 1 includes a pulsed light generator 10 (or pulsed light generator 10A) that emits excitation light with a switching excitation wavelength toward a sample with each pulse, and a detector 50 that detects fluorescence from a cell mass S irradiated with excitation light and outputs a detection signal. The detector 50 has a sensor bandwidth adjusted so that the detection signal attenuates during the pulse-off period of the excitation light.

[0109] In the scanning microscope 1 according to this embodiment, fluorescence from a cell mass S irradiated with excitation light whose excitation wavelength switches with each pulse is detected by the detector 50, and a detection signal is output from the detector 50. In the scanning microscope 1 according to this embodiment, the sensor bandwidth of the detector 50 is adjusted so that the above-mentioned detection signal is attenuated within the pulse-off period of the excitation light. By adjusting the detection bandwidth so that the detection signal is attenuated within the pulse-off period (before the next pulse arrives), different pulses are temporally distinguished from each other, and the mixing of detection signals related to the detection of different pulses is avoided. As a result, a configuration that filters the signal according to wavelength to distinguish different pulses becomes unnecessary, and the reduction in the amount of detection signal that would be a problem if such a configuration were provided can be suppressed. Since the amount of detection signal is secured, it is not necessary to make the laser intensity of the excitation light excessively high, and fluorescence generation from outside the focal point can be appropriately suppressed. As described above, the scanning microscope 1 according to this embodiment can improve the fluorescence observation accuracy by improving the amount of detection signal.

[0110] The intensity distribution of the focused beam with respect to depth does not change even when the laser power is reduced. Therefore, the ratio of fluorescence generated at the focal point to the total amount of fluorescence generated outside the focal point (near the surface) does not change even when the laser power is reduced. To reduce this ratio, the only way is to reduce losses due to absorption and scattering. The reason for wanting to reduce the laser power is to avoid damaging the sample. In deep imaging, laser powers exceeding 100 mW are irradiated onto the sample, but in the case of a mouse brain, damage such as inflammation occurs at several hundred mW. In addition, to separate multicolor fluorescence using multiple filters and prevent spectrum overlap in multiple detectors, 40-50% of the generated fluorescence is discarded. Therefore, the detection efficiency is low. Furthermore, because multiple filters and detectors are arranged, the distance from the objective lens to the detector increases, reducing the detection efficiency of fluorescence scattered inside the sample. The configuration according to this embodiment can be adopted as a solution to these problems.

[0111] In a confocal microscope, fluorescence intensity is proportional to the laser intensity, whereas in a two-photon microscope, fluorescence intensity is proportional to the square of the laser intensity. For example, by reducing the repetition frequency from 80 MHz to 8 MHz, the pulse energy per pulse increases tenfold while the average laser power remains the same, resulting in a 100-fold increase in fluorescence intensity. Therefore, pulse synchronization techniques can be used more effectively with two-photon (multiphoton) microscopes than with confocal microscopes to increase the amount of fluorescence signal.

[0112] The sensor bandwidth of the detector 50 may be adjusted such that the combined time until fluorescence decays and the time until the detection signal decays is shorter than the repetition interval of the excitation light pulses. With such a configuration, the time until fluorescence and the detection signal decays can be made shorter than the repetition interval of the pulses, thereby more reliably achieving the temporal separation of the different pulses described above.

[0113] The detector 50 may detect fluorescence without passing it through a wavelength-separating filter (such as a bandpass filter). This avoids the reduction in the detection signal amount caused by passing it through a filter, and allows for an appropriate improvement in the detection signal amount.

[0114] The detector 50 may consist of a single photosensor 51 that detects fluorescence corresponding to the excitation wavelength of each pulse. With such a configuration, it is possible to achieve an improved detection signal quantity without requiring a configuration to filter the signal according to wavelength, while also achieving a simpler configuration. Furthermore, with only one photosensor, it can be placed as close to the objective lens as possible, thereby improving the detection efficiency of scattered fluorescence.

[0115] The pulsed light generator 10 (or pulsed light generator 10A) emits excitation light by switching between a first excitation wavelength to stimulate the cell aggregate S and a second excitation wavelength to observe the effect of the light stimulation of the first excitation wavelength for each pulse. The detector 50 may not detect light at the fluorescence detection timing from the cell aggregate S corresponding to the pulse of the first excitation wavelength, but may detect light at the fluorescence detection timing from the cell aggregate S corresponding to the pulse of the second excitation wavelength. With this configuration, it is possible to detect only the necessary fluorescence while avoiding the detection of stimulation fluorescence (removing stimulation fluorescence), thereby contributing to an improvement in the detection limit. The pulsed light generator 10 is configured to switch the excitation wavelength for each pulse using pulsed light obtained from the same oscillator, which is more convenient compared to a configuration that switches the excitation wavelength for each pulse using multiple excitation lasers of different wavelengths. In methods using two different lasers, it is difficult to synchronize the frequencies in different resonators, and the interval between pulses gradually shifts over time, which may prevent the desired effect from being fully achieved. However, this problem can be solved by using pulsed light obtained from the same oscillator.

[0116] The scanning microscope 1 further includes an A / D converter 60 that converts the detection signal output from the detector 50 into a digital signal, and the sampling frequency of the A / D converter 60 may be synchronized with the repetition frequency of the excitation light pulse. With such a configuration, it becomes easier to obtain the necessary information from the analog signal corresponding to each pulse without providing a high-speed A / D converter, and high-speed signals can be processed at low cost. Furthermore, an appropriate fixed delay may be provided to enable sampling at an ideal timing.

[0117] The A / D converter 60 may be configured to acquire the maximum value of the detected signal. As described above, by synchronizing the frequency to always acquire the same position in the signal waveform, and further acquiring the maximum value of the detected signal, it is possible to more appropriately obtain the necessary information from the analog signal corresponding to each pulse.

[0118] The A / D converter 60 may be configured to acquire the integral value of the detection signal. When acquiring the maximum value as described above, there is a risk that the maximum value may not be acquired properly due to fluctuations in the fluorescence signal, for example. In this respect, by acquiring the integral value, the necessary information can be acquired more appropriately from the analog signal corresponding to each pulse. The A / D converter may also be configured to acquire the time waveform of the detection signal. In this case, information such as fluorescence lifetime can also be extracted from the time waveform.

[0119] This disclosure is not limited to the embodiments described above. Modified configurations will be described with reference to Figures 34 to 38.

[0120] Figure 34 is a schematic diagram of a modified scanning microscope 2001 (fluorescence detector). In Figure 34, some components are omitted from the illustration for ease of explanation. The scanning microscope 2001 shown in Figure 34 comprises a pulse light generator 2010, a microscope housing 2080, a fluorescence detector 2050, and an A / D converter 2060 (see Figure 35).

[0121] The pulsed light generator 2010 includes an oscillator 2011, a first fiber amplifier / compressor 2012, an AOM (acousto-optic modulator) 2013, a soliton shift fiber 2014, a second fiber amplifier / compressor 2015, and a wavelength conversion unit 2016.

[0122] Oscillator 2011 generates pulsed light. First fiber amplifier / compressor 2012 amplifies (increases power output) the pulsed light output from oscillator 2011 and compresses the pulse duration of the pulsed light. The fiber amplifier may be, for example, a normal dispersion fiber containing erbium. AOM2013 is located downstream of the first fiber amplifier / compressor 2012 and modulates the ultrashort pulsed light using acoustic (sound wave) power. Soliton shift fiber 2014 is located downstream of AOM2013 and modulates the wavelength of the pulsed light using soliton self-frequency shift. Second fiber amplifier / compressor 2015 is located downstream of soliton shift fiber 2014 and amplifies the pulsed light and compresses the pulse duration of the pulsed light. Wavelength conversion unit 2016 performs wavelength conversion, for example, by second harmonic generation (SHG).

[0123] The pulsed light output from the wavelength conversion unit 2016 is irradiated onto the cell aggregate S via the dichroic mirror 2020 contained in the microscope housing 2080. Fluorescence from the cell aggregate S is detected by the fluorescence detector 2050 via the dichroic mirror 2020. The fluorescence detector 2050 may be, for example, a photomultiplier tube or a photodiode (PD). The A / D converter 2060 (see Figure 35) converts the detection signal (analog signal) output from the detector 3050 into a digital signal and outputs the digital signal.

[0124] Here, in addition to the detection signal, a synchronization signal generated by delaying the pulsed light for a predetermined time is input to the A / D converter 2060. Figure 35 shows the time waveforms of each signal. In Figure 35, the horizontal axis represents time and the vertical axis represents intensity. In Figure 35, from top to bottom, the following are shown: "1" pulsed light emitted from oscillator 2011 (for example, light including a wavelength band of 1550 nm), "2" light that has passed through AOM2013, "3" light input to microscope housing 2080, "4" light input to fluorescence detector 2050, "4" signal output from fluorescence detector 2050, and the synchronization signal. The positions of "1" to "4" in Figure 34 correspond to the time waveforms "1" to "4" shown in Figure 35.

[0125] As shown in Figure 36, the synchronization signal may be generated based on a signal obtained by, for example, branching (sampling) a portion of the pulse trains labeled "1," "2," and "3" in Figure 34 and detecting and outputting it with a detector 2055 such as a photodetector. For example, the synchronization signal may be generated by delaying the signal output from the detector 2055 by an arbitrary (predetermined) time with a delay generator 2090. The synchronization signal output from the delay generator 2090 is input to the A / D converter 2060.

[0126] As shown in Figure 35, the A / D converter 2060 receives a synchronization signal generated by delaying the pulse for a predetermined time in addition to the detection signal, and outputs the digital signals of the detection signal and the synchronization signal from the A / D converter 2060. The computer (not shown) samples the signal acquisition from the fluorescence detector 2050 based on the digital signals output from the A / D converter 2060 (the digital signals of the detection signal and the synchronization signal), thereby suppressing noise in the fluorescence detector 2050 and appropriately capturing the peak of the fluorescence signal of the cell aggregate S generated by pulse irradiation.

[0127] Figure 37 is a schematic diagram showing another modified pulsed light generator 3010. The pulsed light generator 3010 includes an oscillator 3011, a first fiber amplifier / compressor 3012, an AOM 3013, a soliton shift fiber 3014, a WDM (Wavelength Division Multiplexing) 3017, second fiber amplifiers / compressors 3015, 3015, wavelength conversion units 3016, 3016, and a WDM 3018. In the configuration shown in Figure 37, the AOM 3013 rapidly modulates the intensity of each pulse, and the soliton shift fiber 3014 modulates the wavelength of each pulse to generate a pulse train. Then, in the WDM 3017, the light of two wavelengths is separated, and each light passes through the second fiber amplifier / compressor 3015 and the wavelength conversion unit 3016 before reaching the WDM 3018. The WDM3018 outputs light by integrating two wavelengths. In this embodiment, the two wavelengths are switched alternately, but the wavelengths do not have to be alternate, and there may be more than two wavelengths.

[0128] Figure 38 shows the time waveforms of each signal. In Figure 38, the horizontal axis represents time and the vertical axis represents intensity. In Figure 38, from top to bottom, the following are shown: "1" pulse light emitted from oscillator 3011 (for example, light including the wavelength band of 1550 nm), "I" modulated signal input to AOM3013, "2" light after passing through AOM3013, "4" light output from WDM3018, signal output from detector (not shown), signal obtained by delaying light of one wavelength (light A), signal obtained by delaying light of the other wavelength (light B), and synchronization signal generated by delaying each excitation wavelength by a different amount.

[0129] As shown in Figure 38, the synchronization signal is generated by delaying pulsed light by different delay amounts for each excitation wavelength. In other words, the synchronization signal is generated by delaying light A by an arbitrary delay amount and delaying light B by a different delay amount than light A. For example, when generating a synchronization signal based on light that has passed through an AOM, the pulse train after integrating different wavelengths will have a shift in pulse period due to the AMP fiber length, refractive index, and optical path length. Therefore, by using a synchronization signal with unequal periods (a synchronization signal delayed by different delay amounts for each wavelength), the peak of the fluorescence signal can be appropriately captured.

[0130] In this case, if we only look at the detection signal and synchronization signal from the detector, it is conceivable that we cannot distinguish which excitation wavelength each sample represents. In this regard, a computer (not shown) can identify the excitation wavelength related to the detection signal based on an identification signal that identifies the excitation wavelength, thereby determining which excitation wavelength each sample represents. In this case, the identification signal may be generated based on a synchronization signal of any of the excitation wavelengths. By controlling the start timing of signal train acquisition using a synchronization signal of any of the excitation wavelengths, the excitation wavelength of each sampled signal can be appropriately identified.

[0131] The A / D converter 3060 may receive a synchronization signal generated by delaying the pulsed light output from the oscillator 3011 before it is input to the first fiber amplifier / compressor 3012. Possible acquisition locations for the pulsed light that generates the synchronization signal include, for example, before the first fiber amplifier / compressor 3012 and after the WDM 3018. The optical pulse is converted into an analog signal by a detector such as a photodetector, a certain threshold is set for the peak value of the analog signal, and the signal is binarized and a delay is added to generate the synchronization signal. However, when the pulsed light is acquired after the WDM 3018, problems arise such as the difficulty in setting the threshold when the difference in light intensity between light A and light B becomes large, potentially causing the delay to shift, and the need to change the threshold for generating the synchronization signal when the output is changed by the amplifier. In this regard, when the pulsed light is acquired before the first fiber amplifier / compressor 3012, although it is necessary to have a different delay for each final output wavelength, the synchronization signal can be generated based on pulsed light with a stable light intensity without being affected by the amplifier. [Explanation of symbols]

[0132] 1,2001…Scanning microscope (fluorescence detector), 10,10A,2010,3010…Pulsed light generator (excitation light source), 50…Detector, 51…Optical sensor, 60,2060,3060…A / D converter, 2050…Fluorescence detector.

Claims

1. An excitation light source emits excitation light towards the sample, with the excitation wavelength switching with each pulse. The system includes a detection unit that detects fluorescence from the sample irradiated with the excitation light and outputs a detection signal, The detection unit is configured such that the detection signal is attenuated during the pulse-off period of the excitation light. The system further includes an A / D converter that converts the detection signal output from the detection unit into a digital signal. In addition to the detection signal, the A / D converter receives a synchronization signal generated by delaying the pulse for a predetermined time. The excitation light source comprises an oscillator that oscillates pulses and an amplifier that amplifies the pulses output from the oscillator. A fluorescence detection device to which a synchronization signal generated by delaying the pulse output from the oscillator before it is input to the amplifier is input to the A / D converter.

2. The fluorescence detection device according to claim 1, wherein the detection band of the detection unit is adjusted such that the time taken for the fluorescence to decay and the time taken for the detection signal to decay are shorter than the repetition interval of the excitation light pulses.

3. The fluorescence detection device according to claim 1 or 2, wherein the detection unit detects the fluorescence without using a wavelength separation filter for the fluorescence.

4. The fluorescence detection device according to claim 1 or 2, wherein the detection unit is composed of a single photosensor that detects all of the fluorescence corresponding to the excitation wavelength of each pulse.

5. The excitation light source emits excitation light by switching between a first excitation wavelength for stimulating the sample and a second excitation wavelength for observing the effect of the optical stimulation of the first excitation wavelength for each pulse. The fluorescence detection apparatus according to claim 1 or 2, wherein the detection unit does not detect light at the fluorescence detection timing from the sample corresponding to the pulse of the first excitation wavelength, but detects light at the fluorescence detection timing from the sample corresponding to the pulse of the second excitation wavelength.

6. The fluorescence detection device according to claim 1 or 2, wherein the sampling frequency of the A / D converter is synchronized with the repetition frequency of the excitation light pulse.

7. The fluorescence detection apparatus according to claim 6, wherein the A / D converter is configured to acquire the maximum value of the detection signal.

8. The fluorescence detection apparatus according to claim 6, wherein the A / D converter is configured to acquire the integral value of the detection signal.

9. The fluorescence detection apparatus according to claim 1, wherein, in addition to the detection signal, a synchronization signal generated by delaying the pulses by a different delay amount for each excitation wavelength is input to the A / D converter.

10. The processing unit further comprises a unit that performs predetermined signal processing on the signal output from the A / D converter. The fluorescence detection apparatus according to claim 9, wherein the processing unit identifies the excitation wavelength related to the detection signal based on an identification signal that identifies the excitation wavelength.

11. This involves emitting excitation light, whose excitation wavelength changes with each pulse, towards the sample, This includes detecting fluorescence from the sample irradiated with the excitation light and outputting a detection signal, The detection bandwidth is adjusted so that the detection signal attenuates within the pulse-off period of the excitation light. The output detection signal is further converted into a digital signal using an A / D converter. In addition to the detection signal, the A / D converter receives a synchronization signal generated by delaying the pulse for a predetermined time. A fluorescence detection method in which a synchronization signal generated by delaying a pulse oscillated by an oscillator before it is amplified by an amplifier is input to the A / D converter.

Images