JPH09210907A - Scanning fluorescent sensing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately measure quickly the distribution of fluorescent beams emitted by a fluorescent substance contained in a specimen by solving both problems of the pulse pair resolution of a photo-sensor and the quantum noise. SOLUTION: When an exciting light is cast onto a specimen 6 to be scanned by a scanning mechanism 84, a fluorescent beam is emitted and fed to a photo- sensor 10, and photo-electrons in the number in compliance with the photo-electron number distribution according to the light quantity are emitted, and a multiplexed current signal is emitted. The current signal is integrated for a certain time by an integrator 20 to be converted into a voltage signal, which is converted into digital values by an A/D converter 50, and the wave-height distribution of voltage signals is prepared by a histo-memory 60 for each irradiating position. The photo-electron number distribution in case the photo-sensor 10 receives fluorescent beam is presumed for each irradiating position and the light quantity of the fluorescent beam is determined on the basis of the crest distribution for each position, single photo-electron event crest distribution, and k-photo-electron event crest distribution, and the light quantity distribution of the fluorescent beam on the specimen 86 is determined.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the field of biochemistry, for example, by measuring the amount of fluorescence emitted from a sample excited by excitation light while scanning the excitation light or the sample, and measuring The present invention relates to a scanning fluorescence detection device that quantitatively measures the distribution of the number of fluorescent molecules.

[0002]

2. Description of the Related Art Conventionally, laser microscopes, flow cytometry, capillary electrophoresis, optical probe microscopes, etc. have been used as devices for quantitatively measuring the distribution of fluorescent substances (or fluorescent molecules) in a sample using a scanning means. , Is widely used in various fields. Such a scanning fluorescence detection device has a scanning means, scans the excitation light for exciting the fluorescent substance in the sample, scans the excitation light source that outputs the excitation light, or the sample I am scanning. By combining such a scanning means and a means for detecting the fluorescence generated from the sample, the fluorescence generation distribution of the sample or the change in the fluorescent light amount over time with respect to the scanning is measured.

In the fluorescence detection device having such a scanning means, there is a problem in that the scanning range of the sample for which the fluorescence generation distribution is to be obtained and the scanning time required for measuring the fluorescence generation distribution over the entire scanning range are in balance. Become. That is, it is desired to measure a wide scanning range with a short scanning time, but since the two are contradictory, it is necessary to compromise the scanning range and the scanning time at appropriate points.

For example, in the case of a laser microscope, the fluorescence generated from each position of the sample is measured by scanning the laser light to construct an image showing the fluorescence generation distribution. The fluorescence measurement time per pixel of the image is about 1 μsec to 10 μsec, which is short. This means that fluorescent detection is performed inefficiently in consideration of the time taken to scan the entire scanning range.

Particularly, when the fluorescence is weak, the S / N ratio per pixel becomes poor, which is a serious problem. When fluorescence is detected by a photodetector in a direct current manner, the S / N ratio is extremely deteriorated, and thus the amount of fluorescence light cannot be measured. Therefore, in order to improve the S / N ratio, fluorescence is measured by a so-called photon counting method, which counts the number of pulses output from the photodetector corresponding to each incident photon. . Even in the fluorescence measurement by the photon counting method, since the fluorescence detection time per pixel is short, the fluorescence distribution is measured by performing scanning many times in order to improve the S / N ratio.

[0006]

However, even in this case, when the fluorescence emitted from the sample is weak, the following problems still remain.

In the light quantity measurement by the photon counting method, the pulse pair resolution which is a characteristic of the photodetector becomes a problem. That is, even if two or more photons are continuously incident on the photodetector at approximately the same time or at short time intervals, the photodetector may output only one pulse. This means that two or more incident photons are counted as only one, and counting is therefore lost, so that the amount of incident light cannot be accurately measured. This phenomenon is likely to occur when the sample is irradiated with excitation light having a large amount of light and fluorescence having a large amount of light is generated.

In order to prevent such counting of incident photons from occurring, the amount of excitation light and the amount of fluorescent light may be reduced. However, when the amount of fluorescent light becomes smaller, the probability that the photodetector will detect fluorescence is now lower, and because the fluorescence detection time is short, the measured value obtained by the photodetector becomes a quantum noise. And the S / N ratio deteriorates. Then, in order to reduce the influence of this quantum noise and improve the S / N ratio, it is necessary to increase the number of scans over a long period of time.

As described above, in order to measure the fluorescence generation distribution of the entire scanning range of the sample in a short time, it is desired to perform the fluorescence measurement per pixel in an extremely short time. If the amount of excitation light is large, the problem of the pulse pair resolution of the photodetector occurs, while if the amount of excitation light is small, the problem of quantum noise occurs. Considering the problems of pulse pair resolution and quantum noise, it is necessary to make a compromise at an appropriate place to determine the pumping light amount. However, it is difficult to determine the appropriate amount of pumping light, and even with the amount of pumping light that is compromised, the problems of pulse pair resolution and quantum noise still exist.

The present invention has been made in order to solve the above problems, and eliminates the problems of both pulse pair resolution and quantum noise of a photodetector, and eliminates the fluorescence generated from a fluorescent substance contained in a sample. An object of the present invention is to provide a scanning fluorescence detection device capable of accurately measuring distribution in a short time.

[0011]

The scanning fluorescence detection apparatus according to the present invention comprises (1) an excitation light source for outputting excitation light, and (2) at least one of a sample irradiated with excitation light and excitation light. By scanning in a direction substantially perpendicular to the optical axis of the excitation light, and irradiating each irradiation position on the sample with the excitation light for a predetermined time or more, and a scanning unit that outputs an irradiation position signal according to the irradiation position, ( 3) A photodetector that receives fluorescence emitted from each irradiation position, emits a number of photoelectrons according to the photoelectron number distribution according to the amount of fluorescence, and multiplies the photoelectrons to output a current signal. ) At each irradiation position, integrating means for integrating the current signal for a certain period of time to convert it into a voltage signal,
(5) AD conversion means for outputting a digital value corresponding to a voltage signal after a lapse of a fixed time at each irradiation position,
(6) An irradiation position signal and a digital value are input, a predetermined value is cumulatively added to the addresses corresponding to the irradiation position and the digital value, and a wave height distribution generation unit that generates a wave height distribution of the voltage signal for each irradiation position, 7) 1 photoelectron emitted by photodetector
In the case of a single photoelectron event wave height distribution generating means for obtaining the wave height distribution generated by the wave height distribution generating means as a single photoelectron event wave height distribution, (8) the number k of photoelectrons emitted by the photodetector is 2 K-photoelectron event wave height distribution in each case of above and a predetermined number or less, k-photoelectron event wave height distribution generation means for calculating by recursive convolution calculation based on the single photoelectron event wave height distribution, and (9) At each irradiation position, based on the pulse height distribution generated by the pulse height distribution generating means when the fluorescence enters the photodetector, the single photoelectron event pulse height distribution, and the k-photoelectron event pulse height distribution for each number k of photoelectrons. , Photoelectron number distribution estimation means for obtaining the amount of fluorescent light by estimating the photoelectron number distribution when fluorescence enters the photodetector, and (10) Photoelectron number distribution estimation for each irradiation position Fluorescence intensity distribution generating means for generating a fluorescence intensity distribution in the sample based on the fluorescence intensity obtained by the means.

The device operates as follows. The excitation light output from the excitation light source is sequentially scanned and irradiated at each irradiation position on the sample by the scanning unit, and an irradiation position signal indicating the irradiation position is output from the scanning unit.
The fluorescence generated from the sample is incident on the photodetector, and the number of photoelectrons according to the photoelectron number distribution corresponding to the amount of light is emitted and multiplied to output a current signal. The current signal is integrated by the integrating means for a certain period of time and converted into a voltage signal, and the voltage signal is converted into a digital value by the AD converting means. The irradiation position signal and the digital value are input to the wave height distribution generating means, and the wave height distribution of the voltage signal is generated for each irradiation position. Based on the wave height distribution for each irradiation position, the single photoelectron event wave height distribution obtained by the single photoelectron event wave height distribution generation means, and the k-photoelectron event wave height distribution generated by the k-photoelectron event wave height distribution generation means The photoelectron number distribution estimating means estimates the photoelectron number distribution when the photodetector receives the fluorescence generated at each irradiation position, and obtains the amount of the fluorescence. Then, the fluorescence light amount distribution generation means obtains the fluorescence light amount distribution in the sample.

As a photodetector, (1) a photoelectric conversion surface that emits a number of photoelectrons according to the photoelectron number distribution according to the amount of received fluorescence, and (2) a reverse bias voltage is applied between the anode and the cathode. The applied portion is set to a potential higher than the potential of the photoelectric conversion surface, and the electron-hole pair generated by inputting photoelectrons is avalanche multiplied and avalanche multiplied. Avalanche photodiode that outputs a current signal according to the number of electron-hole pairs,
(3) A vacuum container having an incident window for transmitting fluorescence and having a photoelectric conversion surface and an avalanche photodiode inside is preferably used.

The photoelectron number distribution estimating means may estimate the photoelectron number distribution by the maximum likelihood method. Moreover, a Poisson distribution may be assumed as the photoelectron number distribution, or an arbitrary distribution may be assumed.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the description of the drawings, the same elements will be denoted by the same reference symbols, without redundant description.

First, the structure of the scanning fluorescence detection apparatus according to the present invention will be described. FIG. 1 is a configuration diagram of a scanning fluorescence detection apparatus according to the present invention.

The scanning fluorescence detection apparatus according to this embodiment is
(1) An excitation optical system including an excitation light source 80 that outputs excitation light and a lens 82 that collects the excitation light and irradiates the sample 86 with the lens 82; and (2) the sample 86 in a direction orthogonal to the optical axis of the excitation light. Scanning mechanism 84 for scanning and XY stage 8 on which these are mounted
(3) a condensing optical system including an objective lens 90 and a barrier filter 92 for guiding fluorescence generated from the fluorescent substance in the sample 86; (4) photodetector 10 and integrator 2
0, an amplifier 30, a sample hold circuit 40, an AD converter 50, a hist memory 60, and a calculation unit 70, and a fluorescence detection system for measuring the fluorescence output from the focusing optical system,
(5) Control for controlling the scanning mechanism 84, generating signals for controlling the integrator 20, the sample-hold circuit 40, the AD converter 50, and the hist memory 60 in synchronization with the scanning, and outputting the number of scanning times N And a portion 94.

The excitation light emitted from the excitation light source 80 is condensed by the lens 82 and applied to the sample 86 containing the fluorescent substance. As this excitation light source, for example, a xenon lamp or a laser light source is used. When a xenon lamp is used, a bandpass filter that transmits only a predetermined wavelength band of excitation light is inserted in the optical path between the excitation light source 80 and the sample 86. The excitation light may be continuous light or pulsed light. Hereinafter, the excitation light will be described as continuous light. The lens 80 collects the excitation light and collects it at a predetermined position on the surface of the sample 86 to efficiently excite the fluorescent substance. The excitation light source 80 has a sufficiently small diameter. It is not necessary when outputting.

The sample 86 irradiated with the excitation light is placed on the XY stage 88 via the scanning mechanism 84.
The scanning mechanism 84 uses, for example, a piezo actuator that uses a piezo element in which distortion is caused by an applied voltage, and moves the sample 86 from the side of the sample 86. When scanning on the XY plane, the piezo actuators are provided in the X and Y directions, respectively. Further, the XY stage 88 includes a scanning mechanism 84 and a sample 86.
A hole is provided in the vicinity of the center, and the fluorescence generated in the sample 86 and directed downward passes through the hole.

These XY stage 88 and scanning mechanism 8
4 can be moved in a direction orthogonal to the optical axis of the excitation light under the control of the control unit 94. XY stage 8
Reference numeral 8 is for roughly determining the region for observing the fluorescence distribution generated from the sample 86, and the scanning mechanism 84 finely adjusts the observation region and scans the sample 86 so that each position in the observation region is scanned. The excitation light is sequentially emitted. The scanning of the sample 86 is not limited to one time, and may be performed a plurality of times, and the number of times of scanning N is counted by the control unit 94.
In addition to the XY stage 88, a Z stage (not shown) that moves the sample 86 in the direction along the optical axis of the excitation light.
May be provided.

The fluorescence generated from the fluorescent substance contained in the sample 86 when the sample 86 is irradiated with the excitation light is the objective lens 90.
Then, the light enters the photodetector 10 through the barrier filter 92. The objective lens 90 is for guiding only the fluorescence generated in the visual field region of the sample 86 at a certain time during scanning to the photodetector 10, and measures the spot diameter of the excitation light with which the sample 86 is irradiated and the measurement of this device. It determines the resolution. Further, the barrier filter 92 transmits fluorescence, but blocks the excitation light and scattered light transmitted through the sample 86.

The excitation light source 80, the lens 82, the objective lens 90, the barrier filter 92, and the photodetector 10 are integrated and the relative positional relationship is fixed. The XY stage 88 can move in the XY directions with respect to these integrated parts, and the scanning mechanism 84 can scan the sample 86 in the XY directions with respect to the XY stage 88.

The photodetector 10 that receives the fluorescent light that has passed through the barrier filter 92 emits a number of photoelectrons according to the photoelectron number distribution corresponding to the amount of light of the incident light flux, and multiplies the photoelectrons to generate a current signal. It is what is output. For example, a photomultiplier tube or an avalanche photodiode (hereinafter referred to as A
Photodetector using PD (Reference: Shawn J. Fag
en, "Vacuum avalanche photodiodes can count single
photons ", Laser Focus World, Nov. (1993) pp.125-1
32) is used. In particular, a photodetector using APD is preferably used. A cross-sectional view of a photodetector using this APD is shown in FIG.

The photodetector 10 using this APD has an entrance window 1 formed in a part of a vacuum container 11 whose inside is kept vacuum.
2 is provided, and the incident light flux A passes through the incident window 12 and reaches the photoelectric conversion surface 13. Photoelectric conversion surface 13
Is, for example, −10 with respect to the anode 16 of the APD 15.
Since a high voltage of kV to −15 kV is applied by the high-voltage power supply 19, when the incident light flux A enters the photoelectric conversion surface 13, the number of photoelectrons B according to the photoelectron number distribution according to the light quantity of the incident light flux A is obtained. Is released. Then, the photoelectrons B are accelerated by the electric field between the photoelectric conversion surface 13 and the APD 15, are focused by the focusing electrode 14 having an opening at the center and set to a predetermined potential, and enter the APD 15.

In this APD 15, a reverse bias voltage (for example, +2.5 kV) is applied between the anode 16 and the cathode 17 by the reverse bias power source 18, and the potential of the anode 16 facing the photoelectric conversion surface 13 is The potential is set higher than the potential of the photoelectric conversion surface 13. This AP
When the photoelectron B collides with D15, a pair of electrons and holes are generated for every 3.6 eV of energy lost by the photoelectrons in the APD 15 due to the ionization action, and this electron-hole pair is avalanche multiplied in the APD 15. , And is output as a current signal between the anode terminal 16a and the cathode terminal 17a. However, the energy lost by photoelectrons in the APD 15 is not a constant value but follows a certain distribution.
The multiplication factor of D15 does not have a constant value but follows a certain multiplication factor distribution, so that the magnitude of the current signal output by the incidence of one photoelectron also has a certain distribution.

Therefore, when the photodetector 10 measures a light flux having a constant light quantity many times, the number distribution of photoelectrons emitted from the photoelectric conversion surface 13 (photoelectron number distribution) is around a certain average value according to the light quantity. The current signal output from the photodetector 10 has a wider distribution, and the current signal output from the photodetector 10 has a wider distribution according to the distribution of the number of electron-hole pairs output from the APD 15 by one photoelectron.

The photodetector 10 using this APD has low noise and high sensitivity as compared with a photomultiplier tube and can count single photons. Furthermore, in the wave height distribution of the current signal obtained by measuring the light flux of a constant light quantity many times,
k-photoelectron event, that is, an event in which the number of photoelectrons emitted when a light beam is incident on the photoelectric conversion surface 13 is k (k = 1, 2, 3,
...) It has an excellent feature that the peaks corresponding to each can be identified. Since it has such characteristics, it is suitable as a photodetector used in the scanning fluorescence detection apparatus according to the present invention.

The integrator 20, which receives the current signal output from the photodetector 10, integrates the current signal for a certain period of time and converts it into a voltage signal. In the integrator 20, a switch 21 and a capacitor 22 are provided in parallel between the anode terminal 16a of the photodetector 10 and the ground terminal.
The switch 21 opens and closes according to a gate signal output from the control unit 94 in synchronization with the scanning of the scanning mechanism 84.
The capacitor 22 integrates the current signal output from the photodetector 10 only while the switch 21 is open, and the potential as the integration result appears at one terminal (point P in the figure). When the switch 21 is closed, the electric charge accumulated in the capacitor 22 is discharged and the potential at the point P becomes the ground potential.

The potential at the point P, which is the result of integrating the current signal in the integrator 20, is amplified by the amplifier 30 so as to have an appropriate amplitude when the sample-hold circuit 40 and the AD converter 50 in the subsequent stage operate. The amplified voltage signal is sampled and held by the sample-hold circuit 40, and is further converted by the AD converter 50 into a peak value which is a digital value corresponding to the voltage signal.
The sample hold circuit 40 and the AD converter 50 also operate under the control of the signal output from the control unit 94. Further, the digital value output from the AD converter 50 may be stored in a frame memory (not shown) and may be displayed on a monitor television (not shown). In this case, in the scanning range of the sample 86. The fluorescence emission distribution can be observed in real time.

The hist memory (wave height distribution generating means) 60 for inputting the peak value and the irradiation position signal representing the excitation light irradiation position of the sample 86 output from the control unit 94 is the sample 8
Every time fluorescence measurement is performed at each irradiation position of 6, a predetermined value (for example, 1) is cumulatively added to the address corresponding to the peak value and the irradiation position signal. By cumulatively adding to the hist memory 60 for multiple scans of the sample 86, A
The peak value distribution of the voltage signal output from the D converter 50 is obtained for each irradiation position of the sample 86. When the scanning range is wide, the number of irradiation positions is large, and the wave height distribution is wide, the hist memory 60 needs a large address space. In such a case, for example, a memory corresponding to each irradiation position may be provided and the memory may be switched each time the irradiation position moves.

Sample 8 generated in this hist memory 60
6, the wave height distribution at each irradiation position and the number of scans N of the sample 86
Based on the above, the calculation unit 70 causes the distribution of the number of photoelectrons emitted from the photoelectric conversion surface 13 of the photodetector 10 (photoelectron number distribution).
And the amount of fluorescence is obtained based on the estimated photoelectron number distribution. For example, a computer is used as the arithmetic unit 70. Details will be described later.

Next, the switch 21, the sample hold circuit 40, the AD converter 50, and the hist memory 6 described above.
The operation of each of 0 and the control unit 94 will be described in detail. FIG. 3 is an explanatory diagram of each signal that controls the operation of each unit of the scanning fluorescence detection apparatus. The signals for controlling the operations of the above-mentioned respective parts of the photodetector 10 and the scanning mechanism 84 are
It is output from the control unit 94.

The scanning control signal (FIG. 3A) for controlling the scanning mechanism 84 to scan the sample 86 changes its voltage value stepwise. The scanning control signal is applied to the piezo actuator of the scanning mechanism 84, and the amount of distortion corresponding to the voltage value is generated. As a result, sample 8
In 6, the excitation light is irradiated to each irradiation position corresponding to the voltage value of the scanning control signal for a certain period of time. The scanning mechanism 84
However, when the sample 86 is scanned in two directions, the X and Y directions, a scan control signal is applied to each of the piezo actuators provided in each direction.

The gate signal (FIG. 3 (b)) for controlling the opening / closing operation of the switch 21 has a pulse shape of a constant width Tw which is output in synchronization with the change timing of the value of the scanning control signal. The rising time of this signal may be the same as the change time of the scan control signal, or may be after a predetermined time has elapsed from the change time of the scan control signal. During a period when this signal is at a high level (a period having a pulse width Tw), the switch 21 is opened, the current signal output from the photodetector 10 is integrated, and the potential which is the integration result appears at the point P.
While the signal is at the low level, the switch 21 is closed and the potential at the point P becomes the ground potential. In this way, for each point where the sample 86 is irradiated with the excitation light, the current signal output from the photodetector 10 is integrated according to the amount of fluorescence emitted from that point.

The sample hold signal (FIG. 3 (c)) for controlling the operation of the sample hold circuit 40 is a pulse signal having a rising edge after a certain time Dw has elapsed from the rising time of the gate signal. At the rising time of this signal, the sample hold circuit 40 samples and holds the voltage value output by the potential of the point P amplified by the amplifier 30. What is important here is that Dw is a constant value equal to or less than Tw. That is, in the sample-hold circuit 40, the potential of the point P at the time point when the integrator 20 starts the integration according to the gate signal and a certain time Dw has elapsed is the amplifier 3.
The voltage value amplified and output by 0 is acquired. And
The sample hold circuit 40 holds the sampled value while the sample hold signal is at a high level. Therefore, the sample hold signal needs to be kept at high level at least until the digital conversion operation by the AD converter 50 is completed, and the sample hold circuit 40 needs to hold the sampling value.

The AD start signal (FIG. 3 (d)) for controlling the digital conversion operation in the AD converter 50 is settling time (starting the sample and hold operation in the sample and hold circuit 40 from the rising time of the sample and hold signal. It is a pulse signal having a rising edge after the elapse of the time required until the output value is determined). At the rising time of this signal, the AD converter 50 converts the voltage value output from the sample hold circuit 40 into a peak value (FIG. 3E) which is a digital value.

The control unit 94 also outputs an irradiation position signal (FIG. 3 (f)) which is a digital value corresponding to the scanning control signal applied to the scanning mechanism 84. This signal is, for example, A
The scan control signal at the rising time of the D start signal is digitally converted. In this case, the peak value output from the AD converter 50 and the irradiation position signal output from the control signal 94 are updated at the same timing.

The hist memory 60, to which the peak value and the irradiation position signal are input, cumulatively adds a predetermined value to the address corresponding to both the peak value and the irradiation position signal. Then, by scanning the sample 86 many times, a peak height distribution corresponding to each position of the excitation light irradiation of the sample 86 is obtained in the hist memory 60.

Next, a method for estimating the distribution of the number of photoelectrons emitted by the incident light flux entering the photoelectric conversion surface 13 of the photodetector 10 will be described in detail. This estimation is performed in the calculation unit 70.

Prior to the estimation, first, the single photoelectron event wave height distribution generating means 71 of the arithmetic unit 70 acquires the single photoelectron event wave height distribution generated in the hist memory 60 in the case of a single photoelectron event. Here, a single photoelectron event refers to an event in which a light beam is incident on the photoelectric conversion surface 13 of the photodetector 10 and one photoelectron is emitted, and a single photoelectron event wave height distribution refers to the event in which the event occurs. This is a wave height distribution generated by the history memory 60 in the case of the above. That is, the single photoelectron event wave height distribution is such that one photoelectron emitted from the photoelectric conversion surface 13 is A
The distribution of the number of electron-hole pairs output from the APD 15 upon entering the PD 15 is shown. This single photoelectron event wave height distribution is p ₁ (h) as a function of the wave height h.
Expressed by h is an output value from the AD converter 50. It should be noted that this single photoelectron event wave height distribution is obtained by making a very weak light beam incident on the photodetector 10 such that the probability of occurrence of a single photoelectron event is overwhelmingly dominant.

Since the noise of the amplifier system is superposed on the low peak portion of p ₁ (h), the actual data cannot be used. Therefore, in the low crest portion, the actual data is not used, and the extrapolated data is obtained from data of an adjacent crest region where noise is not superimposed. Also, p ₁ (h) is

[Equation 1] Standardize with h _max is the maximum peak value determined by the number of bits of the AD converter 50 that converts the peak value into a digital value.

Subsequently, the k-photoelectron event wave height distribution generating means 72 of the arithmetic unit 70 is operated by the k-photoelectron event wave height distribution, that is,
The wave height distribution p _k (h) generated in the hist memory 60 when a light beam is incident on the photoelectric conversion surface 13 of the photodetector 10 and k photoelectrons are emitted,

[Equation 2] (K = 2, 3,..., K). K is 2 or more. When an arbitrary distribution is assumed when estimating the photoelectron number distribution, the peak value giving the peak value of p ₁ (h) is set to h _peak1 , and K = h _max /
_Let h _peak1 . h _max = 4095, h _peak1 = 400
Then, K is about 10. When a Poisson distribution is assumed as the photoelectron number distribution, K is h _max /
about two to three times h _peak1 , that is, h _max = 40
If 95, h _peak1 = 400, K should be about 30. As described above, the wave height distributions p _k-1 (h) and p ₁ (h)
The reason that the wave height distribution p _k (h) can be _obtained from the convolution calculation with is that the wave height distribution p _k (h) is avalanche when k photoelectrons emitted from the photoelectric conversion surface 13 are incident on the APD 15. It is based on the fact that it represents the distribution of the number of multiplied electron-hole pairs.

The photodetector 10, the integrator 20, the amplifier 30, the sample hold circuit 40, and the AD converter 50.
If the spread of the wave height distribution due to the noise that occurs in P is not negligible, the wave height distribution p ₁ (h) can be calculated by deconvolution calculation to remove the influence of the noise from the wave height distribution p ₁ (h). Each _k (h) is calculated, and then the influence of noise is superimposed on each wave height distribution p _k (h) by convolution calculation (k =
2,3, ..., K).

As described above, the wave height distribution p _k (h) (k = 1, k = 1, considering the influence of noise peculiar to the photometric device)
2,3, ..., K) are prepared prior to the estimation of the photoelectron number distribution. FIG. 4 is a diagram showing the respective wave height distributions p _k (h) of the _k -photoelectron events thus obtained.

Then, the photoelectron number distribution estimating means 73 of the arithmetic unit 70 receives the luminous flux to be measured by the photodetector 10 to obtain the wave height distribution n (h) generated in the hist memory 60, and this wave height distribution. From n (h) and the wave height distribution p _k (h) in the above-mentioned single photoelectron event and k-photoelectron event, respectively, the average value of the number of photoelectrons emitted from the photoelectric conversion surface 13 by receiving the measured light beam, that is, The photoelectron number distribution is estimated as follows.

This wave height distribution n (h) does not need to be standardized, and may be the value as it is stored in the hist memory 60.
That is, the peak height distribution n (h) represents the distribution of the events for which the peak value h was obtained. When estimating the distribution of the number of photoelectrons emitted from the photoelectric conversion surface 13 of the photodetector 10 after receiving the incident light flux (photoelectron number distribution), for example, the maximum likelihood method is used. That is, the probability of each k-photoelectron event occurring is _defined by q _k (k = 0,
1,2, ..., K)

(Equation 3) Q _{k at} which the log likelihood represented by the following expression is maximized is obtained, and this is set as the estimated photoelectron number distribution. Here, N is the number of measurements and h _min is the minimum peak value h that can be used for analysis. When the peak value h is small, noise due to the photodetector 10, the amplifier 30, etc. is superimposed and cannot be used for analysis. Therefore, only the _peak value h _min or higher is analyzed. Also, p (h) and p _ND are

(Equation 4)

(Equation 5) It is. This p (h) is the k-photoelectron event (k = 1,2,3, ..., K)
It represents the occurrence probability distribution of the peak value h in consideration of each occurrence probability (photoelectron number distribution). To obtain q _k that maximizes the log-likelihood of equation (3), a numerical calculation method such as the quasi-Newton method used for the optimization problem is applied.

When the Poisson distribution is assumed as the photoelectron number distribution, the probability q _k that each k-photoelectron event occurs
Is

(Equation 6) It is represented by Here, λ is the average value of the number of photoelectrons emitted from the photoelectric conversion surface 13. In this case, finding the photoelectron number distribution that maximizes the log likelihood means finding the λ value that maximizes the log likelihood, and can be found by numerical calculation such as the golden section method. In the following, a Poisson distribution is assumed as the photoelectron number distribution. FIG. 5 shows a wave height distribution (broken line) generated in the hist memory 60 by receiving the measured light flux by the photodetector 10 and λ estimated by the maximum likelihood method.
An example with a wave height distribution (solid line) calculated based on the values is shown. Both show good agreement in the range above the _peak value h _min (= 150). The average value λ of the number of photoelectrons estimated at this time was 1.03.

Then, the fluorescent light quantity distribution generating means 74 of the calculation unit 70 of the sample 86 is calculated based on the average value λ of the number of photoelectrons for each excitation light irradiation position of the sample 86, that is, the fluorescent light quantity, obtained as described above. The fluorescence emission distribution in the scanning range is obtained, a fluorescent substance distribution image is formed by making the amount of fluorescent light correspond to, for example, the luminance, and the image is displayed on the image display unit (not shown).

The present invention is not limited to the above embodiment, and various modifications are possible. For example, in the above embodiment, the pumping light output from the pumping light source is described as continuous light, but the pumping light may be pulsed light.
When the pumping light is pulsed light, the timing of each signal in FIG. 3 is synchronized with the pumping pulsed light generation timing, for example. That is, the excitation light source generates one excitation pulse after a lapse of a certain time after the time when the scan control signal changes. The time D from the rising time of the gate signal to the rising time of the sample hold signal
The w, that is, the integration time of the integrator is, for example, about 4 to 5 times the fluorescence lifetime. Alternatively, the number of times of scanning can be reduced by generating excitation light with a constant pulse number of 2 or more within the time Dw.

The excitation light may be scanned instead of scanning the sample. In this case, the sample is fixedly mounted on the XY stage, and the excitation light scanning means is provided between the excitation light source and the sample.

Further, by irradiating excitation light from above the sample,
Instead of detecting the fluorescence generated toward the bottom of the sample, the fluorescence generated toward the top of the sample may be detected. In this case, the optical system is equipped with a dichroic mirror, and the excitation light output from the excitation light source is reflected by the dichroic mirror and then irradiated onto the sample, and the fluorescence emitted from the sample passes through the dichroic mirror and is detected by the photodetector. To be done.

The most suitable method for estimating the photoelectron number distribution is the maximum likelihood method, but other methods such as the least squares method can also be used.

[0053]

As described above in detail, in the scanning fluorescence detection apparatus according to the present invention, the excitation light output from the excitation light source is sequentially scanned and irradiated by the scanning means at each irradiation position on the sample, An irradiation position signal indicating the irradiation position is output from the scanning means. The fluorescence generated from the sample is incident on the photodetector, and the number of photoelectrons is emitted according to the photoelectron number distribution according to the amount of light, and the photoelectrons are multiplied and a current signal is output. The current signal is integrated by the integrating means for a certain period of time and converted into a voltage signal, and the voltage signal is converted into a digital value by the AD converting means. The irradiation position signal and the digital value are input to the wave height distribution generation means, and the wave height distribution of the voltage signal is generated for each irradiation position. Based on the pulse height distribution for each irradiation position, the single photoelectron event pulse height distribution obtained by the single photoelectron event pulse height distribution generation means, and the k-photoelectron event pulse height distribution generated by the k-photoelectron event pulse height distribution generation means Then, the photoelectron number distribution estimation means estimates the photoelectron number distribution when the photodetector receives the fluorescence generated at each irradiation position, and obtains the amount of the fluorescence. Then, the fluorescence light amount distribution generation means obtains the fluorescence light amount distribution in the sample. Here, as the photodetector, one using an avalanche photodiode is preferably used, and the photoelectron number distribution estimating means determines the photoelectron number distribution by the maximum likelihood method assuming that the photoelectron number distribution is Poisson distribution. presume.

With the above-mentioned structure, both the pulse pair resolution and quantum noise of the photodetector can be eliminated, and the fluorescence distribution generated from the fluorescent substance contained in the sample can be accurately measured in a short time. Can be measured.
Further, since it is not necessary to consider these problems, it becomes easy to set the light amount of the excitation light. Therefore, for example, in the field of biochemistry or the like, it can be suitably used as an apparatus for measuring the amount of fluorescence emitted from a sample excited by excitation light and quantitatively measuring the distribution of the number of fluorescent molecules in the sample.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a scanning fluorescence detection apparatus according to the present invention.

FIG. 2 is a sectional view of a photodetector using an avalanche photodiode (APD).

FIG. 3 is an explanatory diagram of each signal that controls the operation of each unit of the photodetector.

FIG. 4 is a diagram showing a wave height distribution p _k (h) of a _k -photoelectron event.

FIG. 5 is a diagram showing a wave height distribution (broken line) obtained by actually measuring a light flux that is a measurement target, and a wave height distribution (solid line) calculated based on a λ value estimated by the maximum likelihood method. is there.

[Explanation of symbols]

10 photodetector, 11 vacuum chamber, 12 entrance window, 13
... Photoelectric conversion surface, 14 ... Focusing electrode, 15 ... Avalanche photodiode (APD), 16 ... Anode, 17 ... Cathode, 18 ... Reverse bias power supply, 19 ... High voltage power supply, 20
... integrator, 21 ... switch, 22 ... capacitor, 30 ...
Amplifier, 40 ... Sample and hold circuit, 50 ... AD converter, 60 ... Hist memory, 70 ... Arithmetic section, 71 ... Single photoelectron event wave height distribution generating means, 72 ... k-photoelectron event wave height distribution generating means, 73 ... Photoelectron number distribution Estimating means, 74 ... Fluorescent light amount distribution generating means, 80 ... Excitation light source, 82 ... Lens, 8
4 ... Scanning mechanism, 86 ... Sample, 88 ... XY stage, 90
... objective lens, 92 ... barrier filter, 94 ... control unit.

Claims (4)

[Claims] 1. An excitation light source that outputs excitation light, and at least one of a sample irradiated with the excitation light and the excitation light are scanned in a direction substantially perpendicular to an optical axis of the excitation light, While irradiating each of the irradiation positions on the sample with the excitation light for a certain period of time or more, a scanning unit that outputs an irradiation position signal corresponding to the irradiation position, and a fluorescence generated from each of the irradiation positions is received, and the amount of the fluorescence light A photodetector that emits a number of photoelectrons according to the photoelectron number distribution according to, and outputs a current signal by multiplying the photoelectrons, and at each of the irradiation positions, integrates the current signal for a certain period of time to obtain a voltage. An integrating means for converting into a signal; an AD converting means for outputting a digital value corresponding to the voltage signal after the elapse of the fixed time at each of the irradiation positions; A wave height distribution generating means for inputting a digital value, cumulatively adding a predetermined value to an address corresponding to the irradiation position and the digital value, and generating a wave height distribution of the voltage signal for each irradiation position; and the photodetector. A single photoelectron event wave height distribution generating means for acquiring the wave height distribution generated by the wave height distribution generating means as a single photoelectron event wave height distribution when there is one photoelectron emitted by the photodetector; A k-photoelectron event wave height distribution in which the k-photoelectron event wave height distribution in each case where the number k of photoelectrons is 2 or more and not more than a predetermined number is recursively calculated by convolution calculation based on the single photoelectron event wave height distribution. Generating means, and at each of the irradiation positions, the fluorescence is incident on the photodetector and the wave height distribution generated by the wave height distribution generating means and the single photoelectron event wave height. The number k of the cloth and the photoelectron
Based on the k-photoelectron event wave height distribution in each case, the photoelectron number distribution estimating means for determining the light quantity of the fluorescence by estimating the photoelectron number distribution when the fluorescence is incident on the photodetector, And a fluorescence light amount distribution generation unit that generates a light amount distribution of the fluorescence in the sample based on the light amount of the fluorescence obtained by the photoelectron number distribution estimation unit for each of the irradiation positions. Fluorescence detector. 2. The photodetector is applied with a reverse bias voltage between a photoelectric conversion surface that emits a number of photoelectrons according to the photoelectron number distribution according to the amount of the received fluorescent light, and an anode and a cathode. And, the portion facing the photoelectric conversion surface is set to a higher potential than the potential of the photoelectric conversion surface, avalanche multiplication of electron-hole pairs generated by inputting the photoelectrons,
An avalanche photodiode that outputs the current signal according to the number of avalanche-multiplied electron-hole pairs, and a vacuum that includes the photoelectric conversion surface and the avalanche photodiode that includes an incident window that transmits the fluorescence. The scanning fluorescence detection apparatus according to claim 1, further comprising: a container. 3. The scanning fluorescence detection apparatus according to claim 1, wherein the photoelectron number distribution estimating means estimates the photoelectron number distribution by a maximum likelihood method. 4. The scanning fluorescence detection apparatus according to claim 1, wherein the photoelectron number distribution estimating means estimates the photoelectron number distribution on the assumption that the photoelectron number distribution is Poisson distribution.