JP3729043B2 - Fluorescence image detection method, DNA inspection method and apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for detecting fluorescence with high sensitivity and high speed by irradiating excitation light. The present invention also relates to a method and apparatus for DNA testing using this fluorescence detection method.
[0002]
[Prior art]
Conventionally, as a method for detecting weak fluorescence obtained by irradiating excitation light, fluorescence obtained by focusing one laser beam on an object is separated from excitation light and detected by a highly sensitive detection element. In order to detect two-dimensionally, the irradiation excitation light beam is scanned relative to the sample surface. At this time, in order to increase the signal-to-noise ratio of the detection, an aperture is provided at a position conjugate with the object to detect the confocal point. In addition, in order to detect two-dimensionally, a so-called nippo disk with a spiral minute circular opening is used. Uniform illumination light is applied to the disk, and the light from the circular opening is imaged on an asymmetric object and reflected. There has been a method of detecting a confocal image or a fluorescence image in a two-dimensional manner by detecting light or fluorescence as transmitted light through the opening.
[0003]
Further, as a method for inspecting a DNA chip or a DNA microarray, there is a method in which fluorescence is detected using the above-mentioned one spot and DNA is inspected from the fluorescence detection result.
[0004]
[Problems to be solved by the invention]
In any of the above methods, it takes a very long time to detect weak fluorescence two-dimensionally. The method of narrowing down one-beam laser light to an object detects each picture element in time series. Therefore, when the fluorescence is very weak, it takes time to detect one picture element, and the whole two-dimensional image is detected. Takes a lot of time. Even if the excitation light of a 1-beam laser is increased, for example, when detecting 6000 × 6000 picture elements in 1 minute, the detection time per picture element is 1 to 2 μsec in order to detect a faint fluorescent object at high speed. (Microseconds). If detection is attempted in such a short time, the fluorescence detection intensity is set to a wide dynamic range (for example, 2¹⁶) Becomes difficult to detect. Ordinary fluorescence is 10% after receiving excitation light.^-9-10^-FiveSince fluorescence is emitted with a delay of sec, sufficient fluorescence cannot be detected when the fluorescence detection time per pixel is 1 to 2 μsec. On the other hand, in the method using the Nippon disk, light is effectively used only for the ratio of the minute apertures in the uniform light, so that it takes a considerable time to detect the weak fluorescence.
[0005]
The present invention solves these conventional problems and provides a technique for detecting weak fluorescence at high speed, high resolution and high sensitivity with respect to a detection target that requires a large two-dimensional resolution. In addition, this technique is used to provide a technique for examining DNA with high sensitivity and high speed.
[0006]
[Means for Solving the Problems]
The present invention uses the following means as a solution to the above problems.
[0007]
A target object having fluorescence characteristics is irradiated with multi-spot excitation light composed of many M minute spots. The fluorescence from each multi-spot obtained by the excitation light is separated from the excitation light, and the fluorescence image emitted from the target object is detected by a plurality of weak light detection elements capable of photon counting. In this way, even very weak fluorescence can be detected. The photon signal obtained from each detection element is individually photon counted, and the photon count number Npm detected by each detection element is individually stored. Further, the positions of the multi-spot light and the target object are relatively changed by a drive system or the like, and the photon count number of each detector is sequentially stored. Thus, if photon count data is stored and collected over a desired range on the target object, a fluorescence image can be constructed from the collected data.
[0008]
Further, sheet-like excitation light is used instead of the multi-spot light. The fluorescence from the irradiation area having a long and narrow shape obtained by irradiating the sheet-like excitation light is separated from the excitation light, and the fluorescence image emitted from the target object is detected by a plurality of weak light detection elements capable of photon counting. . The photon signal obtained from each detection element is individually photon counted, and the photon count number Npm detected by each detection element is individually stored. The position of the irradiation object having an elongated shape by the sheet-like excitation light and the target object are relatively changed, and the photon count number of each detector is sequentially stored. Thus, if photon count data is stored and collected over a desired range on the target object, a fluorescence image can be constructed from the collected data.
[0009]
In order to detect 6000 × 6000 pixels in the above-mentioned 1 minute by irradiating multi-spot or sheet-like excitation light and simultaneously detecting fluorescence of M picture elements, it takes M to 2 M μsec per picture element. It becomes possible to apply. For example, if M = 50, this time is 50 to 100 μsec, and the above problem of fluorescence generation delay can be solved. In addition, a dynamic range can be secured and highly sensitive detection is possible for the following reasons.
[0010]
That is, if the detection time per picture element is about 1 μsec, the detection time sufficient for the counting time of the photon count necessary for performing high-sensitivity detection is shortened, and it becomes difficult to ensure the dynamic range. This is because the pulse width of the photon pulse signal generated when one photon is detected is several tens of ns, so that only a signal of several tens of pulses can be detected within a time of 1 μsec. Further, even if the fluorescence intensity is detected in an analog manner in such a short time, it is difficult to detect with a sufficient dynamic range secured from the frequency characteristics of the detection circuit.
[0011]
The sample is examined as follows using a sample obtained by binding a DNA fragment to which a fluorescent molecule is added as a target to the corresponding DNA. As an example, when the test target is a DNA chip, first, a target obtained by adding a desired phosphor to a DNA fragment prepared by pretreatment from the DNA to be tested is hybridized to the DNA chip. Fluorescence from each multi-spot obtained by irradiating this hybridized DNA chip to be inspected with multi-spot excitation light consisting of many M minute spots is separated from the excitation light. A fluorescence image emitted from the separated DNA chip is detected by a plurality of weak light detection elements capable of photon counting. The photon signal obtained from each detection element is individually photon counted, and the photon count number Npm detected by each detection element is individually stored. Further, the photon count numbers of the detectors are sequentially stored by relatively changing the positions of the multi-spot light and the DNA chip. Thus, photon count number data is stored and collected over a desired range on the DNA chip, and DNA is examined by constructing a fluorescence image from the collected data.
[0012]
Even if a sheet beam is used in place of the multi-spot light, a DNA test is performed by detecting the fluorescence added to the target DNA in the same manner as described in the fluorescence detection using the sheet beam in the fluorescence detection. Can do.
[0013]
Only the multi-spot image or sheet-like elongated region image passes through the focal point of the fluorescent image from the object surface at the position where the minimum aperture diameter of the multi-spot excitation light or sheet-like excitation light irradiated on the target object is passed. A multi-spot opening or a long and narrow opening is arranged for confocal detection. By doing so, a confocal image having a large signal-to-noise ratio can be obtained. That is, the fluorescence at the multi-spot position to be detected or the fluorescence at the position of the sheet-like beam can be detected with low noise without being affected by the fluorescence noise in the background of the fluorescent material or the like existing in the other optical path. When the sample for fluorescence detection is in the two-dimensional plane, if the multi-spot and the sample are moved relative to each other in this plane and confocal detection is performed, the signal is hardly affected by noise from the phosphor outside the plane. Fluorescence detection or DNA inspection with a large noise-to-noise ratio can be performed.
[0014]
In the fluorescence image detection using the multi-spot light, a relative position change between the multi-spot and the target is performed in at least one of a total of three directions including two directions in the target object plane and a direction perpendicular to the target object plane. In particular, in the case of the above confocal detection, by changing the relative position of the object in the optical axis direction from the minimum aperture position of the multi-spot or the minimum aperture position of the sheet beam, the in-focus position is determined from the change in the fluorescence image intensity obtained. Thus, it is possible to obtain correct fluorescence intensity information, surface unevenness information, or three-dimensional position information that emits fluorescence for an object having an uneven surface from the fluorescence intensity at this position.
[0015]
As indicated by the problems in the case of the conventional one-spot excitation light, it is advantageous that the number of multi-spot lights is large in order to detect a fluorescent image at high speed. When M is 10 or more, the maximum number of photon counts that can be detected within this time is as large as several hundreds, which is effective in realizing high sensitivity and a wide dynamic range. In DNA testing and the like, the effect of using multiple spots is significant because the speed is further increased. More preferably, M is 50 or more. That is, when M is set to 50 or more, for example, 6000 × 6000 picture elements are detected over several tens of μs per picture element, and all pixels can be detected within one minute.
[0016]
The multi-spot is detected by a plurality of detection elements, particularly detection elements arranged on a one- or two-dimensional line, and the detected plurality of information is output as a fluorescence image corresponding to the position information of the object or DNA chip. Therefore, they are arranged on a one- or two-dimensional straight line.
[0017]
Let Td be a period in which the relative positions of the multi-spot excitation light or sheet-like excitation light and the object are changed to move one picture element. At this time, a value obtained by dividing the photon count signal by the pulse width Δt0 when single photons are detected, that is, a value obtained by multiplying Td / Δt0 by a desired coefficient α of 1 or more and 0.1 or more, that is, αTd / Δt0 is obtained. Based on the determination that this value reaches the photon count number Npm, that is, when αTd / Δt0 ≧ Npm, the intensity of the multi-spot excitation light or sheet excitation light is changed. Detection is performed so that the photon count can be performed with a reduced intensity. In other words, photon counting more than α times the value obtained by dividing the time to detect one picture element by the photon pulse time width is impossible in terms of accuracy, so the excitation light can be reduced or the excitation light irradiation time can be shortened. In order to effectively reduce the excitation light intensity or reduce the detection fluorescence, the gain of the detection element is reduced to obtain a photon count value that allows effective photon counting.
[0018]
The multi-spot light or sheet-like excitation light is converted into multicolor light having two or more wavelengths. By doing so, information of fluorescence detection is increased, and the fluorescence detection function and the DNA detection function are improved.
[0019]
Further, the time during which the photon pulse signal is at the high level is added and measured over a time βTd within the detection time Td for one picture element. When this value is Th ′, it is determined whether or not the condition of Th ′ ≧ γβTd is satisfied for a predetermined γ of 1 or less. If this equation is satisfied, the result obtained by integrating the output signal of the detection element with a circuit that integrates in an analog manner is used. In this way, even strong fluorescence can be detected. In other words, even if strong fluorescence that cannot be counted by photons is detected, strong fluorescence can be detected by analog integration. Moreover, if the analog detection time is set within Td, detection can be performed over all picture elements without changing the detection time per picture element. As a result, high-accuracy and high-speed detection can be achieved over a wide dynamic range from weak fluorescence using photon count to strong detection intensity using analog integration detection by this integration circuit. As a detection method in which α is equal to 1, photon counting and analog integration detection are simultaneously performed in parallel, and the obtained results are compared and determined, and either detection result is used.
[0020]
A laser is used as the excitation light source. A multi-spot generation hologram or a sheet beam generation hologram is arranged in the resonator of the laser, and a multi-spot is generated by a powerful laser beam in the laser resonator. Usually, the light energy in the resonator of the laser is 10 times or more of the energy of the emitted beam. In a normal laser, the reflectivity of the mirror of the exit window is about 90%, and the remaining 10% is taken out as the exit light. By making the reflectivity of this exit window 100%, confining the light in the resonator, and placing the hologram in the optical path in this resonator, the hologram diffraction efficiency of about 10% can be used as a multi-spot light or sheet beam can do. As a result, multi-spot light or sheet-like beam with intensity about 10 times that of the conventional arrangement of the hologram in the laser beam path can be obtained, enabling high-speed and high-sensitivity fluorescence detection and DNA inspection. Become.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is an embodiment of the present invention. Laser light sources 211, 212, 213,... Are formed by diffusing light emitted from a semiconductor laser having a wavelength of 635 nm into a round parallel beam. K light beams emitted from the laser light sources are incident on the multi-beam splitter 22 at intervals of a pitch P. The multi-beam splitter 22 arranges K beams with a pitch P, divides each incident beam into two, and forms 2K parallel beams with a pitch P / 2. These beams are elliptically polarized light, linearly polarized light, or circularly polarized light having a major axis in the direction of ± 45 degrees with respect to the yz axis of the illustrated coordinates.
[0022]
The 2K beams are arranged at a pitch P / 2 and enter the multi lens 231. The multi lens 231 is attached to the prism 232. The beam that has passed through the multi-lens passes through the prism 232 and enters the multi-pinhole at the boundary with the prism 233.
[0023]
In the prisms 232 and 233, trapezoidal glasses having slightly different thicknesses are bonded to each other at the bottoms, and the bonded surfaces are polarization beam split surfaces. As a result, the beam that has passed through the prism 232 is converged to each pinhole of the pinhole array 234 having 4K pinholes at a pitch P / 4. The pinhole array has the role of removing stray light that becomes noise other than the pinhole portion and irradiating the DNA chip with a multi-spot with less noise.
[0024]
When passing through the prism 232, light passing through the polarization beam split plane is P-polarized light (linearly polarized light orthogonal to z), and reflected light is S-polarized light (linearly polarized light in the z direction). For this reason, half of the light passing through the multi-spot is P-polarized light and the other half is S-polarized light.
[0025]
Immediately after the pinhole array is a half-wave plate or a quarter-wave plate. The half-wave plate is inclined 22.5 degrees with respect to the P and S polarized light. In the case of a quarter wave plate, it is inclined 45 degrees. The P- and S-polarized light that has passed through such a wave plate after passing through the pinhole is linearly polarized at ± 45 degrees with respect to the z axis in the case of a half-wave plate, and right and right in the case of a quarter-wave plate. It becomes counterclockwise circularly polarized light.
[0026]
The light that has passed through the wave plate 235 is incident on the trapezoidal bonding prism 233 described above (however, the thickness difference of the trapezoidal prism is half that of the trapezoidal bonding prism 232 described above). Since the bonding surface, which is the base of the trapezoid of this prism, is a polarization beam splitting surface, the 4K light beams that have passed through the pinhole array and wave plate are further doubled, and the pitch is P / 8. 8K. For this reason, light is emitted from the prism 233 as if 8K point light sources (secondary point light sources) are located at the pinhole array 234.
[0027]
The light from the 8K secondary point light sources thus produced passes through the lens 24, mirror 25, light amount adjuster 26, wavelength selection beam splitter 30, and high NA objective lens 3, and the detection surface 51 of the DNA chip 5. The 8K point light source images are connected to each other. The light amount adjuster 26 is for adjusting the amount of excitation light for fluorescence detection according to the amount of the phosphor on the detection surface 51. That is, when the concentration of the phosphor is large over the entire DNA chip or the fluorescence detection sample, the driving source 260 is driven so that, for example, 10% of the ND filter 262 is inserted into the excitation light path. However, changing the intensity of all the excitation light beams also narrows the dynamic range of detection, so the intensity of each excitation light multi-spot can be adjusted by an optical modulator array (not shown). .
[0028]
For example, 2^NND filter 261 is 100% (no ND filter), and 262 is 100 × 2^{N / 2}% Transmission, replace the ND filter on the entire screen, detect 2 screens, and combine both screens to 2^NIt is also possible to detect a fluorescent image of the dynamic range. If two screens are detected in this way, twice the time is required, but the present invention using a multi-spot can be detected at a speed five times or more that of the conventional method.
[0029]
The 8K point light source images irradiated on the detection surface of the DNA chip are excited by irradiating target DNA having a fluorescent substance at the tip hybridized with the probe DNA.
[0030]
FIG. 2 shows the relationship between the details of the region 50 where the target DNA is present on the glass 51 of the DNA chip 5 and the multi-spot. Reference numeral 501 denotes a cell. The cells are arranged at a constant pitch in the xy direction. In each cell, a desired DNA fragment is attached to the glass for each address of the sequence, which is also called probe DNA. This probe DNA is usually DNA having the same base sequence in each cell, and different DNA is usually probed in different cells. The target DNA sample solution in which the phosphors are attached to the ends of multiple kinds of DNA fragments collected, purified, and amplified from the specimen of the organism to be tested is flowed to the prepared DNA chip in this manner to cause hybridization. . That is, the target DNA corresponding to the base sequence of the probe DNA of each cell binds, that is, hybridizes. The DNA chip in FIG. 2 is made as described above.
[0031]
Multi-spots 2001, 2002, 2003,..., 20M, which are excitation lights, are arranged at a pitch that is an integral multiple of the cell pitch. In the case of FIG. 2, this integer is 2. The beam diameter of the multi-spot is approximately Δ, and the pitch of the multi-spot is NΔ. In the case of FIG. 2, N = 10, and detection is performed by dividing into 5 × 5 picture elements per cell. The reason for this division is that if a foreign substance or the like adheres to the cell, strong fluorescence is generated by this foreign substance and an error occurs in the detection data. This is to obtain the average fluorescence intensity in the cell excluding the portion.
[0032]
When the DNA chip is scanned in the x direction as shown in FIG. 2A and the multi-spot is scanned to the right end of the cell, it moves in the y direction by one pixel Δ, and then scans in the x direction again. When this is repeated and scanning of N picture elements (10 picture elements in the figure) is completed, the movement is performed in the y direction by MNΔ, and the above operation is repeated to scan all cells. FIG. 3 shows the above operation. Time is taken on the horizontal axis, the upper graph shows the amount of movement of the y stage on the vertical axis, and the lower graph shows the amount of movement of the x stage on the vertical axis. x stage is time t_sAt time t_t-T_sTo return to the original position. Detection is performed only on the outbound path. If the stage accuracy is sufficiently good, detection is possible even on the return path.
[0033]
As shown in FIG. 1, the fluorescence detected by the multi-spot irradiation of the excitation light is received by the objective lens 3 and the fluorescence transmitted through the objective lens is received by the wavelength selection beam splitter 30. Light in the vicinity of 670 nm of the fluorescence wavelength is reflected. The fluorescence reflected here is guided to the detection system. That is, the light passes through the mirrors or wavelength separation splitters 32, 34, 35 and 38 and the imaging lens 33, and forms an image on the light receiving apertures on the multichannel photomultipliers 101 and 102. Each light-receiving aperture is provided with a light shielding plate with a pinhole that is approximately the same diameter as the multi-spot image of the excitation light. Only the fluorescence that passes through this pinhole is detected, and confocal detection is performed. Done.
[0034]
The upper graph of FIG. 4 is an enlarged view of the time axis of the lower graph of FIG. The time ts required for one scan is seen in more detail. The lower graph in FIG. 4 has the same time axis as the upper graph in FIG. 4, and the vertical axis shows the timing of the photon count time by a weak light detector such as Photomal. The meaning of this graph is that photon counting is performed when the signal level is 1, and when it is 0, photon counting is not performed. A signal corresponding to one picture element is detected during one break. When 0 is reached and 1 starts next, the next picture element is detected. L during one scan in the x direction_xSince there are individual picture elements, the detection time per picture element is t_s/ L_xIt becomes as follows. The time for photon counting performed for each picture element needs to correspond to the cell position of the DNA chip. Therefore, the photon count start time is generated based on the signal of the length measuring instrument for measuring the position of the x stage.
[0035]
If the photon count end time of one picture element is determined based on the information of the position measuring device of the stage, when the stage does not move uniformly, the photon count time varies for each picture element. Therefore, it is necessary that the end time be a certain time after the start time. For this reason, the pulses of the high-frequency (frequency νr) pulse generator are counted, counted from the start of photon counting, and the photon counting is terminated when the number of pulses reaches a certain number Nr. That is, photon counting is performed for every picture element over a time of Nr / νr seconds. A control circuit 1 in FIG. 1 is a system for controlling the above operation, and 12 is a PC (personal computer), which controls all movements. A stage 13 generates control and drive signals for the stage 4 and takes in positional information of the stage 4 from the length measuring device.
[0036]
As described above, the fluorescence of the picture elements in the x direction is detected one after another by scanning the stage simultaneously with 64 picture elements. Photon count signal S obtained for each picture element_PMIs detected as shown in FIG. In other words, focusing on one photomultiplier of the multichannel photomultiplier, in the case of weak detection fluorescence, the photomultiplier signal is detected with a low frequency as shown in the upper graph of FIG. A pulse signal is obtained. A pulse signal having a constant width is generated from the rising edge of this signal, or a constant signal level I as shown in the upper graph of FIG._TIn the above case, the signal S in the lower graph of FIG._PCAs shown in FIG. The number of detected photons within the detection time ts / Lx−Δt = Nr / νr can be obtained by counting the pulse shaping signals thus produced with a counter.
[0037]
FIG. 6 shows an example when the detected fluorescence intensity is high. Photomal detection signal S_PMAnd signal S after the comparator_PCIs shown. Since photon pulses are detected with high frequency, two or more pulses are partially overlapped and detected. For this reason, a signal having a wide width after comparison has appeared. S_PC ^′Is the signal S_PCIt is the figure which expanded the time between time A and time B. When the detection signals are detected in a superimposed manner, the signal after the comparison has a wider pulse width. As a result, if the number of pulses after comparison is simply counted, the error becomes large. Therefore, as shown in the bottom graph of FIG._PC ^′And a constant high-frequency pulse signal is counted. This counting signal S_PWCIf you use S_PCThe number of photons can be obtained more accurately than counting the number of pulses of the signal. S_PWCA monovalent function F (S_PWC), It is possible to perform more accurate photon counting.
[0038]
However, the above method is insufficient to realize a wider dynamic range required for DNA testing. That is, in the DNA test, a dynamic range from several photon pulses to hundreds of thousands of pulses is required within the time for detecting one picture element. However, since the width of the photon detection pulse is several tens of ns, the time required for one picture element needs to be several tens of ns × 100,000 = several milliseconds when the entire dynamic range is performed by the photon pulse count.
[0039]
Even if the number of multi-spot excitation light is 64, if the number of picture elements is 6000 × 6000, detection takes almost 30 minutes.
[0040]
FIG. 7 shows an embodiment for solving this problem. An output signal of 101 multichannel photomultiplier is input to the analog switch 1A. The initial state of the analog switch is connected to B, that is, the photon count circuit 1B by the PC 10. This means that photon counting is first performed when fluorescence detection is performed for each picture element. FIG. 8 shows the change with time of the photon count signal. (A) and (b) are cases where the fluorescence is very small and the fluorescence intensity is obtained by photon count detection.
[0041]
FIG. 8A shows a pulse signal S obtained by binarizing a photon pulse signal detected by photomultiplier with a comparator._PCIt is. (B) is a counting signal NS obtained by counting the number of pulses of the pulse signal of (a)._PCIt is. This counting signal NS_PCAfter the start of counting α (t_s/ L_x) Threshold N determined at the end of time₀If it does not exceed, the analog switch keeps the photon count in this state, and the fluorescence detection time t for one picture element_s/ L_xThe photon count is performed during −Δt. Time α (t after the start of counting, which is the time when this determination is made_s/ L_x) May be about 0.01 to 0.3 as α. That is, it is sufficient if the photon count pulse is about 10 times or more during this time. The above-mentioned maximum means that the pulse signal is generated with the highest density within this time.
[0042]
FIGS. 8C to 8D show a case where the fluorescence becomes large and inconvenience occurs when the photon count is performed. That is, two or more photon count pulses are simultaneously or partially overlapped. In this case, the pulse signal S after the comparator_PCAs shown in FIG. 8C, the partially overlapped pulse becomes wider. FIG. 8D shows the count signal NS of the pulse signal of FIG._PCIt is. NS_PCIs α (t_s/ L_x) At the threshold N₀It can be seen that it is not suitable for fluorescence detection by photon counting. At this time, the analog switch of FIG. 7 is connected to C, and the photomultiplier signal is integrated in an analog manner by the circuit 1C. In this way, it is determined whether each pixel is detected by photon count or analog integration, and an optimum method is selected and detected each time.
[0043]
Although not shown in the figure, when the fluorescence becomes larger than that in FIG. 8C, the pulses in FIG. 8C are continued, and the pulse count value becomes several or zero. At this time, as shown in FIG._PCIf the signal Spwc is used to count the signals with rectangular high-frequency pulses having a constant frequency, the pulse count signal NS_PWCIs as shown in FIG. This NS_PWCIf it is determined that this value is larger than the threshold value N1 by using a signal and the analog switch is switched to C, analog detection can be performed without mistake even in the case of stronger fluorescence detection.
[0044]
If switched to analog detection, integration is started with the time point switched by the circuit 1C as the start time point. As shown in FIG. 9 (a), the time α (t_s/ L_x), And integration is started from time t as in the case of photon counting._s/ L_xIntegration is performed up to -Δt, and at this point, the integrated value is AD converted by the circuit 1C. The AD-converted digital information 1C01 is sent to the circuit 100. When the fluorescence intensity is weak, the photon count is continued, and from time 0 to time t_s/ L_xThe photon count signal between −Δt is counted, and this digital information 1B01 is also sent to the circuit 100. The result of the photon count digital signal and the analog integrated digital signal transferred in parallel to the circuit 100 from each channel of the multi-channel photomultiplier 101 as described above is transferred to the personal computer PC 11. Whether the detection in each channel is photon count or analog integration is as described above at time α (t_s/ L_x), It is known at the point of time, and the one suitable for detection by the PC 11 is selected and adopted for each channel. That is, this selection signal is also sent to 1C01, 1C02... In FIG.
[0045]
In the above embodiment, whether to select the photon count or the analog integration determines whether the time α (t_s/ L_x), But both are detected in parallel at the same time from the beginning of detection, and (t_s/ L_x) The selection may be made after −Δt.
[0046]
FIG. 9A shows a change in the integration signal when the analog integration detection is executed. Time at which analog integration detection is started α (t_s/ L_x) To time t_s/ L_xThe integrated value increases with a slope proportional to the intensity of the fluorescence detection light between −Δt and time t_s/ L_xThe integrated value at −Δt is sampled and held and AD converted. Such analog detection makes it possible to capture changes in signal intensity over a dynamic range of nearly two digits.
[0047]
A method for enabling detection over a range of more intense fluorescence detection will be described with reference to FIG. Time t_s/ L_xSince the integrated value is saturated before reaching -Δt, it cannot be detected at this time. However, the high-frequency pulse signal used in FIG.
S_hcThe time α (t_s/ L_x) Saturated time t_FTime until (t_F-Α (t_s/ L_x)) To count, saturation value I_FSince this is known in advance, the detection intensity I can be obtained by the following equation.
I = I_F・ ((1-α) t_s/ L_x-Δt) / (t_F-Α (t_s/ L_x))
By measuring from the start of integration to the saturation time in this way, the dynamic range can be further expanded by nearly one digit.
[0048]
To count such saturation time, the following function may be added to the analog integration circuit shown in FIG. That is, the integration signal of the 1C integration circuit is used as a first input, and a signal level corresponding to a predetermined saturation value is input as a second input to the comparator circuit, so that the change in the signal output result can be detected. This is possible by setting the count to end.
[0049]
As described above, the fluorescence detected by the multi-channel photomultipliers 101 and 102 is detected in parallel in 64 channels, and the signal input in parallel in 100 is converted into a time-series signal and transferred to the PC 11. The x stage in FIG. 1 is scanned, and as shown in FIG. 2, 64 picture elements arranged in the y direction are detected simultaneously and sequentially in the x direction. The xy stage 4 in FIG. 1 is provided with an optical length measuring device (not shown), and the amount of movement of the x and y stages is measured. It is assumed that the dimensions Δx and Δy of the fluorescence detection picture element of the DNA chip are each 2 μm, for example. The start of measurement of each picture element starts from the time when the length measuring device of the x stage measures 2 μm movement.
[0050]
Specific examples of the multi-spot fluorescence detection or multi-spot DNA test of the present invention will be described below. The x direction is a scanning operation, the y direction is a shift by one picture element for each scan of x, and the x is greatly shifted in the y direction every 10 scans. Perform an inspection.
[0051]
When a stage position length measurement signal that starts measurement is detected by scanning in the x direction, first, counting of photons is started. Number of detected picture elements L in the x direction_xIs 6400, the pixel pitch is 2 μm, and the scanning time t of 2 μm is t_sIs 60 μsec. Photomal receives 1 photon and the width of the generated pulse is about 30nsec. Therefore, when photon counting is performed for 40μsec out of 60μsec when the spot passes through one picture element, hundreds to thousands of photons are counted. It becomes possible to count by the method. Since a pulse signal is connected to a signal having a higher intensity, analog integration detection is performed. The dynamic range of this analog detection is about 100. Therefore, if the photon count and the analog integration are switched and used in the above manner, a dynamic range of about 10,000 can be obtained.
[0052]
Furthermore, if the method for obtaining the detection intensity from the time when the analog integration described with reference to FIG. 9B is saturated is used in combination, the dynamic range can be further expanded by one digit. As a result, over a wide dynamic range of approximately 2 16, that is, 65536, that is, from weak detection light of several photons to strong detection light reaching 100,000 photons can be detected in a short time in 40 μsec.
[0053]
When the fluorescence intensity described above is composed of three regions, the photon count region, the analog integration region, and the analog integration saturation time detection region, each boundary corresponds to the boundary region in advance so that the boundaries are smoothly connected. Calibration is performed using intensity detection light.
[0054]
In the embodiment of FIG. 7, the photon count and the analog integration are detected by time division, but it is also possible to always perform both and select a signal to be used according to the intensity of fluorescence detection.
[0055]
The switching of the detection method is not limited to fluorescence detection, and is generally effective when the detection time Tp per pixel is short and limited. That is, when the ratio Nr of Tp and photon detection pulse time width Δt is smaller than the required detection dynamic range Nd,
Tp / Δt = Nr <Nd
In this case, it is possible to detect a wide range from a weak photon count region to a strong light that cannot be counted by photon count.
[0056]
Since 6000 picture elements are scanned in the x direction at a speed of 60 μsec / 1 picture element and 64 picture elements are detected at the same time by skipping 10 picture elements in the y direction, 384,000 pictures are obtained when one scan is completed over 360 msec. Element is detected in the above-mentioned wide dynamic range. When one scan is completed, the stage is moved by one pixel, that is, 2 μm in the y direction, and the adjacent pixel line is detected in the same manner. This operation is repeated 10 times to detect 640 × 6000 picture elements. Even if the acceleration / deceleration time required for returning the stage is added, the detection of the number of picture elements is completed in 4 to 5 seconds.
[0057]
Next, it moves greatly in the y direction as 640 picture elements 2 μm × 640 = 1.28 mm, and the above operation is repeated. If this large movement is repeated 10 times, 6400 × 6000 picture elements will be detected in about 40-50 seconds. That is, an image of 6000 × 6000 picture elements with a resolution of 2 μm, a dynamic range of 2 to the 16th power, and fluorescence detection and weak light detection with a minimum detection intensity of several photons can be realized within 1 minute.
[0058]
In the above embodiment, the detection time required per picture element is short, but when it is larger by one digit or more and the dynamic range of 2 16 is required in the same way, (b) of FIG. The method for detecting the saturation time shown in FIG. If the photon detection pulse width of the photomultiplier can be further shortened, it can be detected by only photon count or a combination of photon count and analog integration detection.
[0059]
It is important to realize a high-sensitivity, high-resolution, wide dynamic range, and high-speed testing device when applying DNA testing, particularly DNA testing using a DNA chip, to applications involving a large number of samples such as mass screening. This can be achieved by using multiple spots or sheet-like beams as described in the above embodiment to detect multiple pixels at the same time, and change the relative position of the excitation light and sample by scanning to detect fluorescence one after another. Became.
[0060]
By simultaneously detecting 64 (M = 64) picture elements of 6000 × 6000 picture elements, detection is performed in a time of (300 microseconds / picture element) or less, that is, an average picture element detection time (300 microseconds / picture element). M) It became possible to detect in the following, and detection became possible in 3 minutes or less per sample. Conventionally, it takes 10 minutes or more to detect the fluorescent image having the number of picture elements, so that the demand for high-speed inspection can be met. Further, by using both photon count and analog integration, detection is possible at a time of 50 microseconds / pixel, or detection is possible in 1 minute or less.
[0061]
Further, by reducing the multi-spot diameter on the sample by fluorescence detection using the above-mentioned multi-spot, a multi-spot excitation light consisting of small spots in which many M diameters are smaller than 3 μm and larger than 0.3 μm, or narrowing width Is irradiated with sheet-like excitation light smaller than 3 μm and larger than 0.3 μm. The fluorescence from each multi-spot or sheet-like irradiation position thus obtained is separated from the excitation light, and a fluorescence image emitted from the sample is detected by a plurality of weak light detection elements. The signals obtained from each detection element are individually stored, the positions of the multi-spot light or sheet-like excitation light and the sample are relatively changed, and the signals are sequentially stored. Thus, signals can be stored and collected over a desired range, and DNA can be examined by constructing a fluorescence image from the collected data. When such a method is used, the resolution of the detected image is increased. If this high resolution is used, for example, fluorescence is added to a desired target DNA as a sample, and the target DNA in the cell is used as a fluorescence image by hybridizing to the corresponding DNA among the single-stranded DNA in the cell. Can be detected and DNA testing of cells is possible.
[0062]
FIG. 10 is a diagram showing an example of fluorescence detection according to the present invention, and is an example of fluorescence detection particularly effective for DNA testing. The same number as FIG. 2 represents the same thing. Instead of the multi-spot light irradiated every 10 picture elements shown in FIG. 2, sheet-like excitation light 2000 is used. The fluorescence obtained by the irradiation of the excitation light is separated from the excitation light in the same manner as the optical system in FIG. 1, and is detected by a one-dimensional photomal array or an ultrasensitive one-dimensional sensor. Excitation with a sheet-like beam is performed simultaneously over, for example, 50 adjacent picture elements in a one-dimensional direction, and is simultaneously detected by a one-dimensional high-sensitivity sensor. By scanning in the x direction, for example, 500 pixels, orthogonal to the sheet-like beam long in the y direction, a fluorescent image of 50 × 500 picture elements is obtained. Next, it moves by 50 picture elements in the y direction, and performs the above-mentioned one scan. If this is repeated 10 times, a fluorescent image of 500 × 500 picture elements is finally obtained. In this embodiment, background noise is slightly larger than in the above-described embodiment in which isolated or separated spot light is irradiated, but fluorescence detection is performed with less noise than in the case of large irradiation in a two-dimensional plane. It is possible. Since the signal obtained from the photomultiplier or the ultra-sensitive one-dimensional sensor is weak light, photon count detection is performed.
[0063]
FIG. 11 shows an embodiment showing a method of generating multi-spot excitation light for fluorescence detection or DNA inspection using fluorescence detection. The laser 27 has a Brewster angle window at both ends of the laser tube 271 in which a laser medium is sealed, and the laser beam emitted from the laser tube is folded by 100% reflection mirrors 272 and 273 at both ends. The laser beam reciprocating between the resonator mirrors is usually more than 10 times the energy of the beam that is taken out of the exit window from the light energy in the laser resonator and taken out of the resonator. When the multi-spot or sheet beam generating hologram plate 270 is inserted into the resonator, the hologram 2701 is pre-made so as to generate a multi-spot or sheet-like beam. A spot 2702 is reproduced. The laser beam irradiated on the hologram forms a multi-spot image with a diffraction efficiency of about 10%. The remaining light is transmitted as zero-order light as it is and reciprocates in the resonator. Since this multi-spot is exactly the same spot as the multi-spot light immediately after passing through the mask-shaped circular aperture array described in FIG. 1, it is possible to irradiate the inspection target with the lens 24 using the configuration of FIG. It becomes possible. As a result, multi-spot light or sheet-like beam with intensity about 10 times that of the conventional arrangement of the hologram in the laser beam path can be obtained, enabling high-speed and high-sensitivity fluorescence detection and DNA inspection. Become.
[0064]
In the above embodiment, the case of one wavelength as the excitation light has been described. However, the present invention is also effective when two wavelengths or three or more wavelengths are used as the excitation light. In this case, it takes a very long time to detect at a plurality of wavelengths in the past, but if photon count detection is performed using the multi-spot light of the present invention, it can be detected with high sensitivity and high speed.
[0065]
【The invention's effect】
As described above, according to the present invention, a weak fluorescence detection object capable of obtaining a photon pulse of several counts from one picture element, in particular, a DNA microarray used in DNA testing can be detected at high speed with high resolution and high sensitivity. It became possible. Also, analog integration detection is possible, and detection with a very wide dynamic range is possible. That is, for example, a fluorescence image of 6000 × 6000 picture elements is 2¹⁶It is also possible to detect with a detection time of 1 minute in the dynamic range. As a result, it becomes possible to detect a test object such as a DNA chip in an inspection time of about 1 minute, and it is possible to accurately test a large amount of sample samples collected in a mass screening performed widely. It becomes possible.
[0066]
The present invention is not only applied to DNA testing, but is widely used for fluorescence detection microscopy. The time required for observation and inspection, which previously required a long time, is greatly reduced. In other words, it has become possible to detect even faint light in which only a few photons are detected in a short time. Moreover, it has become possible to detect with high accuracy in a short time over a wide range from weak light to strong detection light.
[Brief description of the drawings]
FIG. 1 is a perspective view showing a schematic configuration of a DNA chip inspection apparatus according to the present invention.
FIG. 2 is a plan view of a cell showing the relationship between a cell of a DNA chip according to the present invention and excitation light.
FIG. 3 is a diagram illustrating scanning in the xy direction of a DNA chip according to the present invention.
FIG. 4 is a diagram showing the relationship between scanning in the x direction of a DNA chip according to the present invention and photon detection.
FIG. 5 is a diagram showing a photon count signal of faint light.
FIG. 6 is a diagram showing a photon pulse signal in the case of strong detection light.
FIG. 7 is a block diagram showing a circuit configuration for detecting a wide dynamic range according to the present invention.
FIG. 8 is a diagram illustrating a method for determining weak detection light and strong detection light according to the present invention.
FIG. 9 is a diagram illustrating a method for detecting a stronger detection light according to the present invention.
FIG. 10 is a plan view of a cell of a DNA chip showing a case where sheet-like excitation light according to the present invention is used.
FIG. 11 is a perspective view of a part of an optical system showing an example in which a multi-spot according to the present invention is formed by a hologram in a laser resonator.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Control circuit 1A ... Analog switch 1B ... Photon count circuit 211, 212, 213 ... Laser light source 22 ... Multi beam splitter 231 ... Multi lens 232 ... Prism 3 ... Objective lens 4 ... xy stage 5 ... Sample 33 ... Lens 36, 37 ... Interference filters 101, 102 ... Multi-anode photomultiplier

Claims (21)

Irradiating excitation light to the target object having fluorescent properties, the fluorescence emitted from the object is separated from the excitation light, the position of the target object with the excitation light detecting the fluorescence said separated by the irradiation A fluorescence image detection method for obtaining a fluorescence image from fluorescence data detected over a desired range on the target object by relatively changing In the step of detecting the separated fluorescence, the signal processing is executed by dividing into either the photon count processing or the analog integration processing according to the signal of the initial predetermined time among the signals obtained by detecting the fluorescence. A fluorescent image detection method characterized by that . The fluorescent image detection method according to claim 1, wherein the target object is divided into a plurality of regions, and each of a part of the plurality of regions is simultaneously irradiated with spot-like excitation light. . 3. The fluorescence image according to claim 2, wherein a fluorescence image generated from each region irradiated with the spot-like excitation light of the target object is detected by a detection element corresponding to each region. Detection method. The relative position change between the excitation light and the target is at least one of a total of three directions, ie, the direction in the target plane and the direction perpendicular to the target plane, thereby detecting a two-dimensional or three-dimensional fluorescent image. The fluorescent image detection method according to claim 1 . The fluorescent image detection method according to claim 1, wherein the excitation light is multicolor light having two or more wavelengths . Irradiating a sample comprising a plurality of target DNAs formed by binding a DNA fragment to which a fluorescent molecule has been added to the corresponding DNA, with excitation light, and separating the fluorescence generated from the target DNA from the excitation light by the irradiation; Detecting the separated fluorescence is performed over a desired range on the sample by relatively changing the positions of the excitation light and the sample, and the detected fluorescence is detected over the desired range on the sample. A DNA inspection method for constructing a fluorescence image from the data and performing a DNA inspection based on the fluorescence image, wherein an initial predetermined signal in the signal obtained by detecting the fluorescence in the step of detecting the separated fluorescence DNA processing method characterized in that signal processing is executed by dividing into photon counting processing or analog integration processing according to the time signal of . 7. The DNA testing method according to claim 6, wherein spot-like excitation light is simultaneously irradiated onto each of a part of the target DNAs of the sample .   The DNA test method according to claim 7, wherein the number of the partial target DNAs simultaneously irradiated with the spot-like excitation light is 10 or more. 8. The DNA according to claim 7, wherein a fluorescence image generated from each target DNA irradiated with the spot-like excitation light of the sample is detected by a detection element corresponding to each target DNA. Inspection method . 7. The DNA according to claim 6, wherein, in the step of detecting the separated fluorescence, an analog integration saturation time detection process for obtaining a detection intensity from a time at which the analog integration is saturated is executed when the analog integration is saturated. Inspection method . By changing the position of the excitation light and the sample relative to each other in at least one of a total of three directions, ie, a direction in the object plane and a direction perpendicular to the target object plane, two-dimensional or three-dimensional The DNA test method according to claim 6, wherein a fluorescent image is detected . The DNA test method according to claim 6, wherein the excitation light is multicolor light having two or more wavelengths . 7. The DNA testing method according to claim 6, wherein the sample is a DNA chip . A table means on which a sample having a plurality of target DNAs formed by binding a DNA fragment to which a fluorescent molecule has been added to the corresponding DNA, can be moved in a plane;
Excitation light irradiation means for irradiating the sample placed on the table means with excitation light;
Fluorescence separation means for separating fluorescence generated from the target DNA by irradiating excitation light with the excitation light irradiation means from the excitation light;
Fluorescence detection means for detecting fluorescence separated by the fluorescence separation means;
Control means for controlling the driving of the table means so as to perform excitation light irradiation with the excitation light irradiation means and detection of fluorescence with the fluorescence detection means over a desired range on the sample;
A fluorescence image forming means for forming a fluorescence image from fluorescence data detected by the fluorescence detection means over a desired range on the sample while controlling the driving of the table means by the control means;
A DNA testing apparatus comprising testing means for performing DNA testing based on the fluorescence image constructed by the fluorescence image construction means,
The fluorescence detection means includes a detector, a photon count processing circuit, and an analog integration processing circuit, and the detection is performed according to a signal at an initial predetermined time among signals obtained by detecting fluorescence with the detector. A DNA test apparatus characterized in that a signal obtained by detecting fluorescence with a vessel is separated into a processing circuit of either the photon count processing circuit or the analog integration processing circuit and signal processing is executed . 15. The DNA testing apparatus according to claim 14, wherein the excitation light irradiating means simultaneously irradiates each of a part of target DNAs of the sample with spot-like excitation light . The DNA test apparatus according to claim 15, wherein the number of target DNAs in a part of the plurality of target DNAs of the sample irradiated simultaneously with the spot-like excitation light is 10 or more . The detector of the fluorescence detection means has a plurality of detection elements, and images of fluorescence generated from the respective target DNAs simultaneously irradiated with the spot-like excitation light by the excitation light irradiation means are applied to the respective target DNAs. 15. The DNA test apparatus according to claim 14, wherein detection is performed by the corresponding detection element . The fluorescence detection means further includes an analog integration saturation time detection processing circuit for obtaining a detection intensity from a time at which the analog integration is saturated, and detects an initial predetermined time of signals obtained by detecting fluorescence with the detector. The signal obtained by detecting the fluorescence with the detector according to the signal is divided into the processing circuit of the photon count processing circuit, the analog integration processing circuit, or the analog integration saturation time detection processing circuit, and the signal processing is executed. 15. The DNA testing apparatus according to claim 14, wherein The control means controls the driving of the table means to relatively change the positions of the excitation light and the sample in at least one of a total of three directions, ie, a direction in the object plane and a direction perpendicular to the object plane. The DNA test apparatus according to claim 14, wherein the fluorescent image constructing means constructs a two-dimensional or three-dimensional fluorescent image . The DNA test apparatus according to claim 14, wherein the excitation light irradiation means irradiates the sample with multicolor light having two or more wavelengths as excitation light . 15. The DNA test apparatus according to claim 14, wherein the sample is a DNA chip .

Images