KR102351249B1 - Method and apparatus for multiplexted imaging of spectrally-similar fluorophors - Google Patents

Abstract

We disclose multiplexed fluorescent imaging, which is essential to find out how various biomolecules are spatially distributed in cells or tissues. According to the present disclosure, ten or more different biomolecular images can be obtained through one labeling and imaging by newly designing a fluorescent material selection, detection wavelength band, and signal separation algorithm. The present disclosure is a blind separation technique that separates images without the emission spectrum of a fluorescent material. In this technology, 4 pairs of fluorescent materials are used, and each pair is composed of two fluorescent materials with overlapping emission spectra. Each pair of fluorescent materials is strongly excited by only one excitation laser. Two images with different detection wavelength bands are obtained for each pair, and the two images are separated without information on the emission spectrum of a fluorescent substance through mutual information minimization. These two images can also be separated through Gram-Schmidt orthogonalization and fluorescence measurement-based separation. This signal separation is repeated for each fluorophore pair. In addition, two large stoke's shift fluorescent materials that emit light in a wavelength band that do not overlap with the emission spectra of the above eight fluorescent materials are added, so that a total of 10 or more fluorescent materials can be used simultaneously.

Description

METHOD AND APPARATUS FOR MULTIPLEXTED IMAGING OF SPECTRALLY-SIMILAR FLUOROPHORS

The present disclosure relates to molecular multiplex imaging methods and devices.

In the past several decades, fluorescence imaging techniques have been widely used in biological research and medical diagnosis to observe biological materials inside biological samples. The fluorescence imaging technique is a technique for indirectly observing the biological material inside the sample by imaging the light emitted from the fluorescent material after labeling the biological material inside the sample with a fluorescent material. A fluorescent material absorbs light, is excited, and then emits light again. At this time, it emits light with a longer wavelength than the absorbed light. For example, the fluorescent material absorbs light in a specific wavelength band (eg, 350 to 400 nm) and emits light in a specific wavelength band (eg, 400 to 600 nm). An excitation spectrum indicating the degree of excitation of a fluorescent material for each wavelength is called an excitation spectrum, and a measurement indicating the intensity of light emitted by each wavelength is called an emission spectrum.

The problem to be solved is to provide a material multi-imaging method and apparatus.

A method of operating an electronic device according to the present disclosure includes: acquiring at least one image for a plurality of fluorescent materials each labeled with different biomaterials; and using the obtained image for each of the biomaterials. and separating the images into images, wherein at least any two of the fluorescent materials may each exhibit emission spectra of a similar wavelength band.

An electronic device according to the present disclosure includes a memory and a processor connected to the memory and configured to execute at least one instruction stored in the memory, wherein the processor includes a plurality of fluorescent materials each labeled with different biomaterials. for materials, acquiring at least one image, and separating the acquired image into images for each of the biomaterials, wherein at least any two of the fluorescent materials have emission spectra of similar wavelength bands each can be represented.

According to the present disclosure, image separation accuracy can be greatly improved, and multi-imaging speed can be increased.

According to the present disclosure, 8 or more fluorescent substances (up to 10 implemented in the present disclosure) can be observed simultaneously in one dyeing, so that the time taken for the entire imaging can be shortened several times. The present disclosure can be combined with the repeat staining technique. In this case, since the number of biomolecules that can be imaged at one time increases from 3 to 8 (or 10), the number of repeat staining can be reduced by more than three times. If the number of repeated staining increases, since there is a cumbersome need to register multiple images with each other, reducing the number of repeated staining has the advantage that the overall imaging difficulty decreases. Although it is difficult to obtain the distribution of multi-molecules in three dimensions using this technique, 10 types of biomolecules can be imaged in three dimensions without image matching. According to the present disclosure, since it is not based on the emission spectrum of the fluorescent material, there is no need for calibration or calibration for each microscope or sample. Since only 8 emission filters are required to simultaneously image 10 fluorescent substances, the present disclosure can be implemented with a fluorescence filter type general confocal microscope or a low-cost simple fluorescence microscope.

According to the present disclosure, imaging time is shortened because fewer images than ICA or NMF are required to simultaneously image the same fluorescent material. In addition, compared to ICA or NMF, since only a small number of elements need to be inferred, the accuracy is much higher, and it is possible to separate 8 or more fluorescent substances with high accuracy at the same time from a tissue in which cells are densely packed. If two large stoke's shift fluorescent materials are additionally used, it is possible to image a total of 10 fluorescent materials simultaneously. According to the present disclosure, since the number of fluorescent materials and the number of images to be obtained are the same, the imaging acquisition speed is fast, a spectral detector is not required, and signal separation errors are different because signal separation occurs independently for each pair of fluorescent materials. Because it does not propagate in pairs, and only one element is inferred from millions of pixels, its accuracy is much higher than that of conventional blind separation techniques.

1 is an example of a fluorescent material selection in multimolecular imaging using a general fluorescent material.
2 is a block diagram of an electronic device according to various embodiments of the present disclosure;
3 is a flowchart illustrating a method for multi-molecular imaging of an electronic device according to various embodiments of the present disclosure.
FIG. 4A is a flowchart of the image acquisition step of FIG. 3 .
Figure 4b is a flow chart of the image separation step of Figure 3;
5 is a view for explaining a method for selecting a fluorescent material according to an exemplary embodiment.
6 is a view for explaining a molecular multi-imaging method according to an embodiment.
7 is an example of multiple molecular imaging for each laser according to an embodiment, verification thereof, and a result combined with signal amplification technology.
8 is a result of using an antibody produced in the same host according to an embodiment, a result of application to a cleared tissue, and a result of simultaneous imaging of mRNA and protein.
9 is a result of simultaneous imaging of 8 to 10 proteins according to an embodiment.
10 is a multi-molecular imaging result of a pathological sample according to an exemplary embodiment.
11 is a multi-molecular imaging result using a bandpass filter without a spectral detector according to an exemplary embodiment.
12 is a validation result of antibody complex staining.
13 is a result of verifying the presence or absence of crosstalk between antibody complexes.
14 is a validation result of signal separation by minimizing the amount of mutual information in a sample stained using an antibody complex.
15 is a result of verifying compatibility between the present disclosure and an expansion microscopy (ExM).
16 is a graph of linearity of signal brightness according to laser exposure time.
17 is a graph of linearity of signal brightness according to laser intensity.
18 is a signal separation result based on measurement of a fluorescent material.
19 is a diagram for explaining application of the present disclosure to another multimolecular imaging technique (t-CyCIF).
20 is a signal separation result through Gram-Schmidt orthogonalization of the present disclosure.

Hereinafter, with reference to the accompanying drawings, the embodiments of the present disclosure will be described in detail so that those of ordinary skill in the art to which the present disclosure pertains can easily implement them. However, the present disclosure may be implemented in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present disclosure in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part "includes" a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated. In addition, terms such as “…unit”, “…group”, and “module” described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software or a combination of hardware and software. have.

In the description, an operating subject may be omitted, but the method described in the present disclosure may be implemented in an electronic device, for example, a device including a computing device and a fluorescence microscope.

In general, when biomolecules inside a biological sample are labeled with a fluorescent material and imaged with a fluorescence microscope, the process is as follows.

Light is emitted from a light source, and the light source may be a lamp emitting white light or a laser or LED emitting light in a specific wavelength band. This light passes through an excitation filter that transmits only the light of the desired wavelength band, and only the light of the corresponding wavelength band passes through. The light passing through the excitation filter is reflected by a dichroic mirror, rises upward, and proceeds toward the objective lens of the microscope. After that, it is irradiated to the sample (sample) through the objective lens.

The fluorescent materials inside the sample absorb light of the irradiated wavelength (eg, 350 - 400 nm) and emit light in the 400 - 700 nm wavelength range, which passes through the objective lens again and is emitted through a dichroic mirror. It is transmitted to an emission filter (a dichroic mirror has a characteristic of reflecting light in a specific wavelength band and passing the rest of the light through).

At this time, in addition to the light emitted by the fluorescent material from the sample, autofluorescence of the sample itself is simultaneously emitted. This autofluorescence has a characteristic of being emitted in a very wide wavelength band. Therefore, the light emitted from the sample is a mixture of the light emitted by the fluorescent material and the autofluorescence.

The light emitted from the sample is passed through an emission filter that transmits only light in a specific wavelength band. Using an emission filter that passes only a wavelength of 400 to 700 nm, light except 400 to 700 nm is removed from the light emitted from the sample. Only light in this wavelength band can be transmitted. Instead of the emission filter, various types of devices that selectively transmit, reflect, or refract a specific wavelength band to select only light in a specific wavelength band may be used.

The light passing through the emission filter is sent to an eyepiece or a detector to be expressed as a digital image. As an example of the fluorescence imaging thus obtained, a specific protein (actin) in a sample (cultured cell) may be labeled with a fluorescent material and then imaged with a fluorescence microscope.

1 is an example of a fluorescent material selection in multimolecular imaging using a general fluorescent material. Referring to FIG. 1 , in order to simultaneously observe several biomolecules in one sample, it is necessary to selectively obtain images of each fluorescent material after labeling several biomolecules with different fluorescent materials. For this, it is necessary to use a fluorescent material whose excitation spectrum and emission spectrum do not overlap. In order to use up to four fluorescent materials simultaneously using a commonly used light source in a wavelength range of 400 to 650 nm, as shown in (a), a fluorescent material strongly excited at 405 nm is used. A fluorescent substance, a fluorescent substance strongly excited at 488 nm, a fluorescent substance strongly excited at 560 nm, and a fluorescent substance strongly excited at 640 nm can be used.

Referring to (a), it can be seen that the fluorescent material Alexa 405 is strongly excited only with a 405 nm laser and not with other lasers. It can be seen that other fluorescent materials (Alexa 488, Alexa 546, and Alexa 647) are also strongly excited only by one excitation laser. Each of the four fluorescent materials is excited with a specific laser and emits light with a spectrum as shown in (b).

Referring to (b), the spectrum (green, yellow, red, blue squares) of the emission filter indicates the wavelength of light transmitted by each emission filter. The light emitted from the fluorescent material is expressed as a digital image through the detector after autofluorescence is removed as much as possible through the emission filter for each fluorescent material.

To list the process of multi-molecular imaging with four fluorescent materials, a 405 nm laser is irradiated to the sample to excite Alexa 405, the light emitted from this fluorescent material is filtered through emission filter 1, and then imaged with a detector. After that, the 405 nm laser is turned off and the 488 nm laser is irradiated to the sample to excite Alexa 488, and the light emitted from this fluorescent material is filtered through emission filter 2 and imaged with a detector. After that, the 488 nm laser is turned off and the 560 nm laser is irradiated to the sample to excite Alexa 546, and the light emitted from this fluorescent material is filtered through the emission filter 3 and imaged with a detector. After that, the 560nm laser is turned off and the 640nm laser is irradiated to the sample to excite Alexa 647, and the light emitted from this fluorescent material is filtered through emission filter 4 and imaged with a detector.

As shown in (b), the emission spectrum of fluorescent materials has a width of about 100 nm, so more than five fluorescent materials cannot be used at the same time. When five fluorescent substances are used at the same time, emission spectra of the fluorescent substances start to overlap. For example, when CF 594 fluorescent material is added as shown in (c), the emission spectrum of CF 594 fluorescent material is located behind the Alexa 546 emission spectrum. The CF 594 fluorescent material, like Alexa 546, is strongly excited with a 560 nm laser. Therefore, when two proteins in the sample are labeled with Alexa 546 and CF 594, respectively, and irradiated with a 560 nm laser, both fluorescent materials strongly emit light. However, in this case, it is not possible to separate the two signals separately. When emission filter 3 shown in (c) is used, the signal of CF 594 is partially mixed (red circle), and when emission filter 5 is used, the signal of Alexa 546 is partially mixed ( Blue circle) There is a problem of mixing. Therefore, current multi-molecular imaging has limitations in that only up to four fluorescent materials can be used at the same time, and five or more fluorescent characters cannot be used at the same time.

In accordance with the recent development of digital pathology, various attempts are being made to develop more precise diagnostic methods by simultaneously acquiring and analyzing more biomolecular information from a single sample. In addition, in the case of immunotherapeutic drugs, a third-generation anticancer drug that has recently been spotlighted, research results have been published that show that the reactivity to specific therapeutic agents differs depending on the type of cell type and type of immune cells present in the patient's cancer tissue. is becoming Therefore, the need to simultaneously image more biomolecules in one sample is emerging. However, there is a limit in that only up to four fluorescent materials can be used at the same time when a light source in a wavelength range of 400 to 650 nm, which is commonly used, is used. In order to overcome this limitation, various technologies have been developed, which can be divided into three (repeated staining technique, signal separation technique after spectral imaging, and blind signal separation technique).

First, in the multi-round staining technique, biomolecules inside a sample are labeled with 3 or 4 fluorescent materials, respectively, and imaged, and then the fluorescent material is inactivated or fluorescent material is removed from the biomolecules through chemical treatment. take it off After that, other biomolecules are again labeled with the same 3 or 4 fluorescent substances and imaged. By repeating this method, it is possible to simultaneously observe dozens of biomolecules in one biological sample. However, it takes a long time because fluorescent labeling has to be repeated, and there is a problem in that the sample is damaged during the chemical treatment process.

Second, after spectral imaging and signal unmixing, several biomolecules are individually labeled with several fluorescent substances with overlapping emission spectra, and then, several fluorescent substances are simultaneously excited. After that, images of the sample are obtained in several detection wavelength bands, and the obtained images are separated into images of each fluorescent material based on the information on the relative intensity of each fluorescent material for each wavelength band. For example, after each of three biomolecules in a sample are labeled with three fluorescent substances (ECFP, EGFP, EYFP) with overlapping emission spectra, images are obtained at 12 detection wavelength bands, and from these 12 images, Images can be separated. ECFP starts emitting light at 450 nm and is strongest at 475 nm, EGFP starts emitting light at 470 nm and emits light most strongly at 510 nm, and EYFP starts emitting light at 500 nm and emits light most strongly at 530 nm. emits the strongest light. If the emission spectrum of each fluorescent substance is known, it is possible to obtain a total of 12 images in 12 wavelength bands (from 1 to 12) and then unmix the images of the three fluorescent substances based on the emission intensity of each fluorescent substance. . If the process of separating the signals of the three fluorescent materials is expressed mathematically, it can be expressed as [Equation 1] below (assuming that the resolution of the image obtained through the above process is 1024 x 1024 pixels).

Left-hand matrix (IMG): Each row represents an image (eg IMG1, IMG2, … IMG12) obtained in each fluorescence detection wavelength band, and each column represents the absolute brightness value for each pixel in each image (eg IMG1 ₁ is the first of IMG1) the brightness value of a pixel). When the resolution of the image is 1024 x 1024, the matrix IMG has a total of 1,048,576 (= 1,024 x 1,024) columns.

Right side first matrix (M): Relative brightness of each fluorescent material in each detection wavelength band

Right-hand second matrix (F): Each row of this matrix is an image composed of only each fluorescent material, F1 is an image composed only of the signal of the first fluorescent material (ECFP), and each column is the absolute brightness value of each pixel, just like the matrix IMG. indicates

IMG1,2,3,4,… 12: Images from different wavelength bands

IMG1 ₁ : Value of the first pixel of IMG1

IMG1 _1,048,576 : The value of the last pixel of IMG1 (1024 x 1024 = 1,048,576)

F1,2,3: Images consisting of only three fluorescent substances (ECFP, EGFP, EYFP) signals

F1 ₁ : Value of the first pixel of F1

F1 _1,048,576 : The value of the last pixel of F1 (1024 x 1024 = 1,048,576)

Black: known values (M matrix) or measured values (IMG matrix)

Red: values to find out (F matrix)

As such, in the signal separation technique after spectral imaging, when the intensity of each fluorescent material is known in each detection wavelength band, an image can be obtained in 12 detection wavelength bands, and an F matrix can be obtained through IMG/M. That is, this technique works only when the M matrix is known, which varies depending on the microscope and the chemical environment of the sample. The M matrix refers to the intensity of each fluorescent material in each detection wavelength band, and it varies depending on the optical properties inside the microscope, the sensitivity of the detector for each wavelength, and the chemical composition of the sample. Therefore, it is inconvenient to separately measure the corresponding matrix for each microscope and sample, so it is not widely used for actual tissue imaging.

Lastly, thirdly, blind unmixing is a method of inferring an M matrix and an F matrix simultaneously without knowing the M matrix. For this, Independent Component Analysis (ICA) or Non-Negative Matrix Analysis (Non- Negative Matrix Factorization (NMF) has been used. However, as shown in Equation 2 below, the M matrix has 36 elements and the F matrix has 3,145,728 (=3 x 1024 x 1024) elements, so it is difficult to accurately infer 3 million elements at the same time. very difficult

Black: measured values (IMG matrix)

Red: values to find out (M matrix, F matrix)

As a result, the conventional blind signal separation has been used very limitedly because its accuracy is greatly reduced. In addition, the existing ICA and NMF have a condition that the number of images required for separation (the number of rows of the IMG matrix) must be greater than the number of fluorescent materials (the number of rows of the F matrix). That is, there is a disadvantage that it takes a long time because 9 or more images need to be obtained in order to image 8 fluorescent materials at the same time.

The present disclosure (Process of ultra-multiplexed Imaging of biomoleCules viA the unmixing of the Signals of Spectrally Overlapping fluorophores, PICASSO) solves the drawbacks of the prior art, and is one of blind signal separation techniques. However, this technique can simultaneously image and separate 8 fluorescent materials with much higher accuracy than conventional blind signal separation techniques, with only 8 imaging times. To this end, the present disclosure includes a fluorescent material signal separation by selecting a fluorescent material, selecting a fluorescence detection wavelength band, and minimizing the amount of mutual information.

The present disclosure greatly improves the image separation accuracy by significantly reducing the number of elements to be inferred at the same time compared to the existing ICA or NMF-based techniques through the above three techniques.

For example, when it is necessary to separate signals of two fluorescent materials with overlapping emission spectra using ICA or NMF, imaging is performed in at least three different wavelength bands as shown in Equation 3 below, and then M using the ICA or NMF algorithm We need to infer the matrix and the F matrix at the same time.

However, since the M matrix has 6 elements and the F matrix has 2,097,152 elements, more than 2 million elements must be inferred at the same time, so the accuracy is poor (the numbers in red indicate the elements to be inferred) .

In contrast, in the present disclosure, the above problem can be simplified as shown in [Equation 4] below through a strategy for selecting a fluorescence material and a strategy for selecting a fluorescence detection wavelength band.

As can be seen above, in the present disclosure, the image IMG1 obtained in the first detection wavelength band through selection of the fluorescent material and the fluorescence detection wavelength band includes only F1, which is the signal of the first fluorescent material, so that the relationship IMG1 = F1 is established. IMG2 can be expressed as the sum of F1 and F2 (IMG2 = F1 x α + F2). Since F1 is the same as the measured value IMG1, if only α is accurately inferred, F2 can be calculated (F2 = IMG2 - α x IMG1). . In the present disclosure, the value of α can be accurately inferred using mutual information minimization between IMG1 and IMG2.

That is, if the accuracy of the method using ICA or NMF is low because millions of elements (M matrix and F matrix) must be inferred simultaneously using ICA or NMF, the present disclosure provides a fluorescence material selection strategy and a fluorescence detection wavelength band selection strategy By designing an experiment so that images can be separated if only one element (α) is accurately inferred through

When three or more fluorescent substances need to be imaged simultaneously, instead of using three fluorescent substances with overlapping emission spectra, multiple pairs of fluorescent substances with overlapping emission spectra can be used. For example, in general, when three fluorescent materials with overlapping emission spectra are separated using ICA or NMF, the M matrix and the F matrix must be inferred at the same time by obtaining images in at least four wavelength bands.

As shown in Equation 5 above, it can be seen that the size of the M matrix increases exponentially as the number of fluorescent substances with overlapping emission spectra increases, and when two fluorescent substances with overlapping emission spectra are separated, the size of the M matrix is 3 x 2 = 6, and the size of the M matrix is 4 x 3 = 12 when three fluorescent materials with overlapping emission spectra are separated. In the case of separating N fluorescent substances with overlapping emission spectra as described above, the size of the M matrix ^{increases by N 2} , and when more fluorescent substances are separated, the mathematical complexity of the matrix increases rapidly. For example, the process of separating eight fluorescent materials having overlapping emission spectra with each other by ICA or NMF is mathematically expressed as follows [Equation 6].

In the present disclosure, in order to avoid such a problem, the visible light region is divided into four regions, and only two fluorescent materials having overlapping emission spectra are used in each region. The visible light region is divided into a 405 nm region, a 488 nm region, a 560 nm region, and a 640 nm region, and only two fluorescent materials with overlapping emission spectra in each region are used. As a result, when a total of eight fluorescent materials are used simultaneously, it is expressed as in [Equation 7] below.

In this way, the problem of simultaneously using eight fluorescent substances and separating the signals can be simplified to the process of inferring the value of α four times.

2 is a block diagram of an electronic device 200 according to various embodiments of the present disclosure.

Referring to FIG. 2 , the electronic device 200 according to various embodiments of the present disclosure includes at least one of a detector 210 , an input module 220 , an output module 230 , a memory 240 , and a processor 250 . may contain one. In some embodiments, at least one of the components of the electronic device 200 may be omitted or one or more other components may be added to the electronic device 200 .

The detector 210 may take an image of the sample. In this case, the detector 210 may be installed at a predetermined position of the electronic device 200 to capture an image. For example, the detector 210 includes at least one of a scientific complementary metal-oxide-semiconductor (sCMOS) camera, a photo multiplier tube (PMT), or other equipment capable of measuring light intensity and expressing it as an image. can do.

The input module 220 may receive a command or data to be used for at least one of the components of the electronic device 200 from the outside of the electronic device 200 . In this case, the input module 220 may include at least one of an input device and a reception device. For example, the input device may include at least one of a microphone, a mouse, and a keyboard. In some embodiments, the input device may include at least one of a touch circuitry configured to sense a touch or a sensor circuit configured to measure the intensity of a force generated by the touch. The reception device may include at least one of a wireless reception device and a wired reception device.

The output module 230 may provide information to the outside of the electronic device 200 . In this case, the output module 230 may include at least one of a display device and a transmission device. For example, the display device may include at least one of a display, a hologram device, and a projector. In some embodiments, the display device may be implemented as a touch screen by being assembled with at least one of a touch circuit and a sensor circuit of the input module 220 . The transmission device may include at least one of a wireless transmission device and a wired transmission device.

According to an embodiment, the receiving device and the transmitting device may be integrated into one communication module. The communication module may support communication between the electronic device 200 and an external device (not shown). Such a communication module may include at least one of a wireless communication module and a wired communication module. In this case, the wireless communication module may be formed of at least one of a wireless reception device and a wireless transmission device. In addition, the wireless communication module may support at least one of a long-distance communication method and a short-distance communication method. The short-range communication method may include, for example, at least one of Bluetooth, WiFi direct, and infrared data association (IrDA). The wireless communication module may communicate in a telecommunication manner through a network, and the network may include, for example, at least one of a cellular network, the Internet, or a computer network such as a local area network (LAN) or a wide area network (WAN). can Meanwhile, the wired communication module may be formed of at least one of a wired reception device and a wired transmission device.

The memory 240 may store at least one of a program or data used by at least one of the components of the electronic device 200 . For example, the memory 240 may include at least one of a volatile memory and a non-volatile memory.

The processor 250 may execute a program in the memory 240 to control at least one of the components of the electronic device 200 , and may process data or perform an operation. The processor 250 may be configured to acquire at least one image for a plurality of fluorescent materials each labeled with different biomolecules, and to separate the obtained image into images for each of the biomolecules. . At this time, at least any two of the fluorescent materials may exhibit emission spectra of similar wavelength bands, respectively. In addition, each wavelength band includes a first detection wavelength band corresponding to a part of any one of the emission spectra, and a second detection wavelength band in which at least a part of any one of the emission spectra and at least a part of the other of the emission spectra are overlapped. may include

According to various embodiments, the processor 250 may acquire a first image in a first detection wavelength band and acquire a second image in a second detection wavelength band. At this time, the processor 250 generates an image of at least two emission spectra of the fluorescent materials as light in a wavelength band is irradiated to the fluorescent materials, and generates a first image and a second image from the generated image. Each can be obtained. According to an embodiment, the first image and the second image may be respectively obtained by a band filter. According to another embodiment, the first image and the second image may be respectively acquired by a spectral detector.

According to various embodiments, the processor 250 may acquire the third image from the second image based on the first image. In this case, the processor 250 may obtain the third image by minimizing information shared between the first image and the second image from the second image. For example, the processor 250 predicts a ratio between the brightness of any one of the fluorescent materials in the first detection wavelength band and the brightness of any one of the fluorescent materials in the second detection wavelength band, and based on the ratio , a third image may be obtained from the second image. As an example, the processor 250 may be configured to obtain the third image by subtracting a result of multiplying the ratio by the first image from the second image.

Through this, the processor 250 may detect the first image as any one of the biomolecules and detect the third image as the other image of the biomolecules.

According to the present disclosure, the following advantages can be obtained compared to the existing multimolecular simultaneous imaging technology.

Compared to the repeated staining technique, in the present disclosure, it generally takes from 1 hour to several hours to dye a sample, and the present disclosure observes 8 or more fluorescent substances (up to 10 in the present disclosure) at the same time in one dyeing. This can shorten the entire imaging time several times. The present disclosure can be combined with the repeat staining technique. In this case, since the number of biomolecules that can be imaged at one time increases from 3 to 8 (or 10), the number of repeated staining can be reduced by three times or more. If the number of repeated staining increases, there is a cumbersome need to register multiple images with each other. Therefore, as the number of repeated staining is reduced, the difficulty of the entire imaging decreases.

Due to the above image registration problem, it is difficult to obtain the distribution of multi-molecules in three dimensions with the conventional repetitive staining technique.

Compared to the signal separation technique after spectral imaging, since the present disclosure is not based on the emission spectrum of a fluorescent material, there is no need for microscopic or sample-by-sample calibration or calibration. For spectral imaging, an expensive equipment called a spectral detector is generally used, but a microscope equipped with a spectral detector is not available in general biological laboratories and hospitals. Since only 8 emission filters are required to simultaneously image 10 fluorescent substances, the present disclosure can be implemented with a fluorescence filter type general confocal microscope or a low-cost simple fluorescence microscope.

Compared to the blind signal separation technique, the present disclosure requires fewer images than ICA or NMF to simultaneously image the same fluorescent material, so the imaging time is shortened. Compared to ICA or NMF, since only a small number of elements need to be inferred, the accuracy is much higher, and it is possible to separate 8 or more fluorescent substances with high accuracy at the same time from a tissue in which cells are densely packed. If two large stoke's shift fluorescent materials are additionally used, it is possible to image a total of 10 fluorescent materials simultaneously.

3 is a flowchart of a molecular multi-imaging method of the electronic device 200 according to various embodiments of the present disclosure. FIG. 4A is a flowchart of the image acquisition step of FIG. 3 , and FIG. 4B is a flowchart of the image separation step of FIG. 3 . 5 is a view for explaining a method for selecting a fluorescent material according to an exemplary embodiment. 6 is a view for explaining a molecular multi-imaging method according to an embodiment.

Referring to FIG. 3 , the electronic device 200 may acquire at least one image of a plurality of fluorescent materials in step 310 . In this case, a plurality of fluorescent materials may be respectively labeled with different biomolecules. This will be described later in more detail with reference to FIG. 4A.

Referring to FIG. 4A , the electronic device 200 may select fluorescent materials of each wavelength band in operation 411 . Through this, at least any two of the fluorescent materials may respectively exhibit emission spectra of a similar wavelength band. For this, at least two fluorescent materials may be selected from among numerous existing fluorescent materials or to be newly synthesized in the future according to the fluorescent material selection method of the present technology.

In the present disclosure, four types of excitation lasers in a wavelength range of 400 to 650 nm, which are commonly used visible light regions, are used, and a list of fluorescent materials that are strongly excited only by one excitation laser among the four excitation lasers is prepared. For example, the CF405S and ATTO390 can be used because they are strongly excited only with the 405 nm excitation laser among the four excitation lasers, but the CF440 may not be included in the list because it can be excited with the 405 nm excitation laser and 488 nm excitation laser.

Since it is better to separate the signals of two fluorescent materials as the emission spectra do not overlap, two fluorescent materials with which emission spectra do not overlap as much as possible may be selected, but two fluorescent materials that differ by more than a reference value may be selected. From the list of fluorescent substances for each excitation laser, two fluorescent substances having a maximum wavelength of an emission spectrum for each laser having a difference of more than a reference value may be selected.

In Fig. 5, looking at how the fluorescent material of the 405 nm excitation laser is selected, the three fluorescent materials Alexa 405, CF405S, and ATTO390 are all strongly excited by the 405 nm laser, but not by the other lasers. Alexa 405 and CF405S have a maximum emission spectrum difference of 10nm, whereas CF405S and ATTO390 have a maximum emission spectrum more than 50nm apart. Therefore, in the present disclosure, CF405S and ATTO390 may be selected (Alexa405 and ATTO390 or Alexa 405 and CF405S may be selected). Examples of fluorescent material pairs for each excitation laser selected according to the above conditions are shown in Table 1 below. In addition, other fluorescent material pairs are possible.

Two additional large stoke's shift fluorescent materials (CF405L, ATTO490LS) can be used for the 405nm excitation laser and the 488nm excitation laser. Large-stoke's shift fluorescent material refers to a fluorescent material with a large difference of about 100 nm between the excitation spectrum and the emission spectrum. Due to this characteristic, it can be used simultaneously with the above 8 fluorescent materials. Large-stoke's shift fluorescent materials that can be excited with a 560 nm excitation laser and a 640 nm excitation laser can also be used simultaneously. In this case, a total of 12 fluorescent materials can be used simultaneously.

The electronic device 200 may select a plurality of detection wavelength bands within each wavelength band in step 413 . At this time, as shown in FIG. 6A , each detection wavelength band includes a first detection wavelength band corresponding to a part of any one of the emission spectra, and at least a part of any one of the emission spectra and the other of the emission spectra. A second detection wavelength band in which at least one portion overlaps may be included. In this case, the electronic device 200 may select the first detection wavelength band so that only the signal of the first fluorescent material among the fluorescent materials corresponding to each wavelength band is detected. Fig. 6 (a) is an emission spectrum (solid line and shaded region) and two detection wavelength bands (dashed line) of two fluorescent materials with overlapping emission spectra used.

The electronic device 200 may acquire images of detection wavelength bands, respectively, in operation 415 . In this case, the electronic device 200 may acquire a first image in the first detection wavelength band and acquire a second image in the second detection wavelength band. Through this, the electronic device 200 may acquire a first image with respect to any one of the fluorescent materials of a similar wavelength band and acquire a second image with respect to all of the fluorescent materials of a similar wavelength band. Thereafter, the electronic device 200 may return to FIG. 3 and proceed to step 320 .

According to one embodiment, four pairs of fluorescent materials are used, and each pair of fluorescent materials with overlapping emission spectra are strongly excited by only one of the four standard excitation lasers (405, 488, 560, 640 nm). Therefore, since only the signals of two fluorescent materials need to be separated for each excitation laser, the process is simpler than in the case of separating signals of multiple fluorescent materials with overlapping emission spectra.

When two biomolecules inside the sample are labeled with two fluorescent materials, respectively, and then irradiated with a 405 nm laser, the first fluorescent material marked in blue and the second fluorescent material marked in green emit light at the same time in Fig. 6(a). . The image IMG1 obtained by filtering this light with a transmission filter that transmits the first detection wavelength band indicated by the blue dotted line contains only the signal of the first fluorescent material, so the relation IMG1 = F1 is established. After that, the image IMG2 obtained by filtering the light emitted from the sample with a transmission filter that transmits the second detection wavelength band indicated by the green dotted line contains the signals of both fluorescent materials. After labeling two biomolecules inside a living tissue with two fluorescent substances, respectively, two images are obtained at two detection wavelength bands. In this case, an image may be obtained in each detection wavelength band using a spectral detector, or a band-pass filter may be used. First, the first detection wavelength band is a wavelength band in front of the emission spectrum of the first fluorescent material, and the image IMG1 obtained in this wavelength band contains only the signal F1 of the first fluorescent material. The second wavelength band is a wavelength band at which signals of both fluorescent materials are detected, and the image IMG2 obtained in this wavelength band includes both the signal F1 of the first fluorescent material and the signal F2 of the second fluorescent material. Therefore, the relationship between IMG1, IMG2, F1, and F2 shown in FIG. 6 (a) above is as follows [Equation 8].

α is the ratio of the brightness in the second detection wavelength band to the brightness in the first detection wavelength band of the first fluorescent material. As can be seen from [Equation 8], if only this value of α is accurately known, F2 can be calculated by completely removing the signal of F1 (=IMG1) from IMG2. Here, in Fig. 6 (a), α is 'the area of the blue shade included in the green dotted line / the area of the blue shade included in the blue dotted line'.

Referring back to FIG. 3 , the electronic device 200 may separate the image acquired in step 320 into images of each of the biomolecules. This will be described later in more detail with reference to FIG. 4B .

Referring to FIG. 4B , in operation 421 , the electronic device 200 may infer α for mutual information minimization (MI minimization) between images obtained in detection wavelength bands. In this case, the electronic device 200 may infer α based on the first and second images.

In order to infer the value of α, it is possible to use the minimization of the amount of mutual information. Mutual information amount (MI) is the amount of information shared between two variables, and the MI of two independent random variables is 0. When F1 is completely removed from IMG2, it is assumed that the resulting (IMG2-αx F1) and IMG1 become independent variables and the amount of mutual information is minimized. Therefore, we find the α value that can minimize the amount of mutual information between IMG1 and (IMG2-αx F1), and calculate F2 from this value. Since this method finds one variable α from IMG1 and IMG2 having millions of pixel values, its accuracy is very high compared to the conventional blind separation technique. In addition, in the present disclosure, as two additional methods for finding α, a fluorescence measurement-based separation method that directly measures an M matrix and another blind separation method, Gram-Schmidt orthogonalization, may be used.

The electronic device 200 may detect images of each of the biomolecules based on α in step 423 . In this case, the electronic device 200 may acquire the third image from the second image based on α. Here, the electronic device 200 may obtain the third image by subtracting a result obtained by multiplying α by the first image from the second image. Through this, information shared between the first image and the second image from the second image is minimized, and thus the third image can be obtained. Accordingly, the first image may be detected as one image among biomolecules, and the third image may be detected as another image among biomolecules.

6(b) is an image before and after signal separation when two biomolecules labeled with two fluorescent substances with overlapping emission spectra are imaged in two different detection wavelength bands as shown in FIG. 6(a).

6 (c)-(e) are examples of signal separation for an image composed of 4 pixels (top row), assuming that α is 1.2. The number in the square is the absolute brightness of each pixel. (c) is an image obtained in the first detection wavelength band, and (d) is an image obtained in the second detection wavelength band. (e) is an image of the second fluorophore after signal separation. The α value (1.2) found through mutual information minimization is multiplied by the image obtained in the first detection wavelength band, and then subtracted from the image obtained in the second detection wavelength band. For example, as shown in FIG. 6(a) , the blue-colored portion and the green-colored portion are emission spectra of the two fluorescent materials, and the blue dotted line and the green dotted line indicate two fluorescence detection wavelength bands.

A method of simultaneously imaging eight fluorescent materials with four excitation lasers (405, 488, 560, and 640 nm) using the above method and separating the signals can be expressed as [Equation 9] below.

As can be seen from [Equation 9], the process of simultaneously using eight fluorescent materials and separating the signals has been simplified to a process of inferring one variable α for each wavelength four times. In addition, even if an error occurs in one α inference, the error does not affect fluorescent materials in other wavelength bands, but has the advantage that it affects only one fluorescent material in the same wavelength band (e.g., when an error occurs in _{α 405 inference)} case, only F2 will be inaccurate, F1, F3-8 will not be affected). In the present disclosure, in addition to the above eight fluorescent substances, two large stoke's shift fluorescent substances having a large difference between the excitation spectrum and the emission spectrum were additionally used. Through this, we succeeded in using 10 fluorescent materials simultaneously with 4 general excitation lasers. If two additional large-stoke's shift fluorescent materials capable of strong excitation with 560 nm and 640 nm lasers are additionally used, it is also possible to use a total of 12 fluorescent materials simultaneously.

6(b) shows a process in which two biomolecules in a biological sample are labeled with two fluorescent materials with overlapping emission spectra, and IMG1 and IMG2 are obtained in two fluorescence detection wavelength bands, and then images F1 and F2 of the two biomolecules are separated. shows As shown in FIG. 6(b), after labeling two biomolecules (square dotted line, circular dotted line) inside a living tissue with two fluorescent materials, two images were obtained in the two fluorescence detection wavelength bands shown in FIG. 6(a). When done, IMG1 becomes an image showing a circular protein, and IMG2 shows both a circular and a square biomolecule.

By applying the mutual information minimization to the two images IMG1 (=F1) and IMG2 obtained in this way, α is calculated, and F2 is separated (unmixed) from IMG2 using this. The final results, F1 and F2, are images of circular proteins and rectangular biomolecules, respectively.

In the two image separation steps, α is inferred from IMG1 (=F1) and IMG2 through mutual information minimization.

The amount of mutual information is a value derived from information theory, and the amount of mutual information between two variables refers to the total amount of information shared by the two variables. Therefore, the amount of mutual information between two random variables is zero.

Since digital images are discrete variables, the amount of mutual information between two images can be calculated using the following relational expression. In [Equation 10], p _X (x) and p _Y (y) correspond to the probability distribution function (or histogram) of each image, and p _(X,Y) (x,y) is the combined probability distribution of the two images Corresponds to a function (or combined histogram).

In the present disclosure, when there are two images (IMG1, IMG2), the loss function is defined as follows [Equation 11] by utilizing the amount of mutual information.

Accordingly, the loss function becomes a one-dimensional function having α as an independent variable, and an optimization method such as gradient descent can be used to find the α value having the minimum L(α) value. In other words, the more accurately IMG1 is subtracted from IMG2 (the more accurately α is inferred), the information of biomolecule 1 (the original biomolecule in FIG. The amount of mutual information with IMG1 will be reduced. Using this, it is possible to find the α value that minimizes the amount of mutual information between IMG1 and (IMG2-αx F1).

If existing blind signal separation techniques focused on dividing a given fluorescent material, the present disclosure dramatically reduced the number of elements that must be inferred for blind unmixing through selection of a specific fluorescent material and fluorescence detection wavelength band selection of the present disclosure, After that, it can be accurately inferred by minimizing the amount of mutual information.

As a result of the present disclosure, it is possible to obtain a result of simultaneously observing 8 proteins in a tissue, which has never been achieved by conventional blind signal separation techniques. In addition, existing blind signal separation techniques have to acquire more images than the number of fluorescent materials to be separated, or introduce conditions or assumptions for protein expression (eg, only one protein is expressed in one pixel). The present technology operates without these assumptions or conditions, and requires only the same number of images as the fluorescent material to be separated. In the present disclosure, since image registration is not required to obtain images of 10 biomolecules, it is possible to obtain multi-molecular three-dimensional images. Therefore, it has a great effect compared to the repeated staining technique in which it is difficult to obtain a three-dimensional image due to the difficulty of image registration.

It can be seen that the scope of biomolecules to which the present disclosure can be applied is diverse, and the present technology works in cultured cells, animal tissue sections, and clinical samples. The present disclosure can simultaneously image multiple types of antibody-labeled protein or DNA-labeled mRNA. The present disclosure is not limited to the types of samples and biomolecules listed above, and can be used to image any biomolecules that can be labeled with a fluorescent material from all types of samples in which a fluorescent material can be used.

In order to observe biomolecules inside a tissue with a fluorescence microscope, it is necessary to label specific biomolecules inside a living tissue with a fluorescent material. In general, an antibody having a fluorescent substance attached thereto is used to observe a protein, and an oligonucleotide having a fluorescent substance attached thereto can be used when observing mRNA. In the present disclosure, antibodies were used to observe proteins and oligonucleotides were used to observe mRNA, but the present disclosure is a technique for classifying fluorescent substances. can be applied

The apparatus of the present disclosure can be implemented as a point-scanning confocal microscopy equipped with a spectral detector and a filter-type spinning-disk confocal microscopy. The present disclosure may be implemented with a filter-type point-scan confocal microscope, a filter-type fluorescence microscope (widefield microscopy), and any microscope capable of selectively detecting only light in a special wavelength band.

In the present disclosure, a total of 10 fluorescent materials were simultaneously imaged using 8 fluorescent materials with overlapping emission spectra and two large stoke's shift fluorescent materials. The present disclosure is not limited to these 10 fluorescent materials, and any fluorescent material that meets the fluorescent material selection strategy of the present disclosure may be used. In the present disclosure, the signal of the YFP protein and the signal of the fluorescent material of CF488A were also successfully separated. As such, it is possible to separate signals of several fluorescent proteins or fluorescent proteins and fluorescent materials with overlapping emission spectra. In the present disclosure, when the spectral detector is used, 10 fluorescence detection wavelength bands are used. In addition to this wavelength band, any wavelength band that meets the fluorescence detection wavelength band selection strategy of the present disclosure may be used. In the present disclosure, when a filter-type microscope is used, eight emission filters are used, and all emission filters conforming to the fluorescence detection wavelength band selection strategy of the present disclosure as well as these filters may be used. The present disclosure can be implemented using any equipment capable of reflecting, transmitting, or selectively detecting light in a specific wavelength band in addition to the spectral detector or emission filter. In addition, the present disclosure can be implemented with various light sources, not only lasers or LEDs that emit light of a specific wavelength, but also emit light in a wide wavelength band such as a white laser, a metal halide lamp, and a mercury lamp, and only light of a specific wavelength is selectively selected among them. It can be implemented using any device that can select and irradiate a sample.

There are various techniques that can be combined with the present disclosure. In the present disclosure, it was shown that more fluorescent materials can be used in one staining and imaging round by applying the present disclosure to t-CyCIF (Tissue-based cyclic immunofluorescence) technology, which is one of the repeated staining techniques. It can be applied to any dyeing technique used. In addition, the present disclosure shows that a signal of a specific fluorescent material can be amplified by introducing a fluorescence signal amplification technique to the present technology, but all existing fluorescence amplification techniques can be introduced in the present disclosure.

In the present disclosure, in addition to the method using the mutual information amount minimization, various methods for inferring the α value may be used and combined. The first of these examples is the α measurement method through the measurement of a fluorescence substance. The fluorescence substance used is made into a solution, the brightness of the fluorescence substance solution is measured under the same microscope as the microscope for imaging the actual sample, and the α value is based on the measurement value. is calculated to separate IMG1 and IMG2. The second example is the separation method through Gram-Schmidt Orthogonalization. By Gram-Schmidt Orthogonalization of IMG1 and IMG2, IMG2 can be divided into components perpendicular to IMG1 (=F1) and horizontal components.

In the following, the present disclosure (PICASSO: Process of ultra-multiplexed Imaging of biomoleCules viA the unmixing of the Signals of Spectrally Overlapping fluorophores) will be described.

Multiplexed fluorescent imaging is an essential technique to find out how various biomolecules are spatially distributed in cells or tissues. However, due to overlapping emission spectra of fluorescent materials, only up to four biomolecules can be used simultaneously. The present disclosure provides a technology that can solve this problem, Process of ultra-multiplexed Imaging of biomoleCules viA the unmixing of the Signals of Spectrally Overlapping fluorophores (PICASSO). Unlike the existing spectral unmixing technology, this technology enables the simultaneous use of four or more fluorescent substances without the need to know or measure the spectral information of the fluorescent material in advance. According to the present disclosure, 10 types of biomolecules in the rat brain can be observed simultaneously with one imaging. In addition, the present disclosure can be successfully applied to a clinical sample (FFPE; formalin-fixed paraffin-embedded). Furthermore, it can be easily applied with other multi-molecular imaging technologies (tissue-based cyclic immunofluorescence, t-CyCIF), thereby providing higher multi-molecular imaging capabilities.

In the present disclosure, by newly designing a fluorescent material selection, detection wavelength band, and signal separation algorithm, 10 fluorescent imaging can be used simultaneously with one labeling and imaging. The present disclosure is a blind separation technique that separates images without the emission spectrum of a fluorescent material. In this technology, 4 pairs of fluorescent materials are used, and each pair is composed of two fluorescent materials with overlapping emission spectra. Each pair of fluorescent materials is strongly excited by only one excitation laser. Two images with different detection wavelength bands are obtained for each pair, and the two images are separated without information on the emission spectrum of a fluorescent substance through mutual information minimization. This signal separation is repeated for each fluorophore pair. The present disclosure has four major advantages over existing multimolecular fluorescence imaging techniques.

First, in the present disclosure, the image acquisition speed is fast because the number of fluorescent materials and the number of images to be acquired are equal. Second, the present disclosure does not require a spectral detector. Third, in the present disclosure, since signal separation occurs independently for each fluorophore pair, signal separation phase errors do not propagate to other fluorescence pairs. Finally, since the present disclosure infers only one element from millions of pixels, its accuracy is much higher than that of conventional blind separation techniques.

The present disclosure is applicable to a variety of samples such as cultured cells, mouse tissue sections, and clinical samples, and uses preformed antibody complexes that allow multiple primary antibodies produced in the same host to be used without a spectral detector. It is possible to observe 10 proteins using only antibodies. In addition, it is possible to improve the existing repeat-label multi-molecular imaging technique by increasing the number of fluorescent substances that can be used in one labeling and imaging through combination with repeat-label multi-molecular imaging techniques.

7 is an example of multiple molecular imaging for each laser according to an embodiment, verification thereof, and a result combined with signal amplification technology.

Referring to FIG. 7 , (a-d) is a fluorescent material emission spectrum (solid line) and detection wavelength band (dashed square) for each excitation laser used. (a) CF488A (solid red line), ATTO 514 (green solid line) and ATTO 490LS (large stoke's shift, blue solid line) are excited with a 488-nm laser. Images are successively acquired in three detection wavelength bands: 489-505 nm (red dotted square), 506-530 nm (green dotted square), and 630-680 nm (blue dotted square). (b) Excitation of CF405S (solid red line), ATTO 390 (green line) and CF405L (large stoke's shift, blue solid line) using a 405-nm excitation laser. Each detection wavelength band is 406 to 430 nm (red dotted square), 431 to 460 nm (green dotted square), and 580 to 620 nm (blue dotted square). (c) Two fluorescent materials, CF568 (solid red line) and ATTO Rho101 (solid green line), were excited with a 557-nm laser. The detection wavelength bands are 558 to 575 nm (red dotted square) and 576 to 620 nm (green dotted square), respectively. (d) Two fluorescent materials, CF633 (solid red line) and CF660R (solid green line), are excited with a 640-nm laser. The detection wavelength bands are 641 to 650 nm (red dotted rectangle) and 651 to 690 nm (green dotted rectangle).

(e-h) Images of two or three multimolecular immunofluorescence imaging in HeLa cells using the fluorescent material described in a-d and the corresponding detection wavelength band, and signal separation by minimizing the amount of mutual information. The target proteins are lamin A/C (red), GM130 (green) and vimentin (blue). (e) The fluorescent material shown in a, the detection wavelength band, and the excitation laser were used. (f) The conditions shown in b were used. (g) The conditions shown in c were used. (h) The conditions shown in d were used.

(i-1) The image shown in e is an image for each channel before and after signal separation by minimizing the amount of mutual information. (i) An image obtained in the first detection wavelength band (red dotted line) in a. Only Lamin A/C is observed. (j) Images obtained in the second detection wavelength band (green dotted line) in a. Both Lamin A/C and GM130 are observed. (k) Images after signal separation. After multiplying the image shown in i by the α value estimated by minimizing the amount of mutual information, it was subtracted from the image shown in j. Only GM130 is shown. (l) An image obtained in the third detection wavelength band (a blue dotted line), and the images shown in i, k, and l were combined to generate the image shown in e.

(m-p) This is the verification result of α estimation by minimizing the amount of mutual information. (m) Immunofluorescence image for PV protein, and (n) Immunofluorescence image for Calbindin protein. (o) An image of artificially merging m and n so that the α value becomes 2. Both PV and calbindin are visible in this image. (p) This is the result of separating only the calbindin signal from the o image by applying signal separation by minimizing the amount of mutual information to the image of m and the artificially created image of o. (q-t) A diagram illustrating selective signal amplification using an anti-fluorophore antibody against a fluorescent substance. (q) A schematic diagram showing a mechanism for selectively amplifying a signal. In cultured cells, GM130 was labeled using Alexa 488 and Lamin A/C using ATTO 514. The GM130 signal was selectively amplified using an anti-Alexa488 antibody.

(r-s) Fluorescence images obtained in the first detection wavelength band before (r) and after (s) signal amplification. The same imaging and visualization conditions were used for r and s, and the signal at s was adjusted to prevent saturation of the signal brightness. Green: GM130 (Alexa 488); Blue: Indicates DAPI. Lamin A/C signals were not visualized. (t) Changes in signal brightness (n = 10, error bars: standard deviation) of GM130 (Alexa 488) before and after signal amplification.

First, we verified the validity of signal separation by minimizing the amount of mutual information using four standard excitation lasers. As shown in (a-d), the detection wavelength bands for each of the four excitation lasers (405, 488, 560, 640 nm) were selected. To verify the signal separation, two proteins known to be spatially separated inside the cell, Lamin A/C and GM130, were used as target proteins. After labeling Lamin A/C and GM130 with a fluorescent material overlapping each of the emission wavelength bands by indirect immunostaining, two images were obtained in the detection wavelength band shown in (a-d). After that, by minimizing the amount of mutual information, the two images were separated into an image containing only the Lamin A/C signal and an image containing only the GM130 signal. For 405 nm and 488 nm lasers, vimentin was stained with a large stoke's shift fluorescence substance, and images were obtained by additionally using a detection wavelength band that can selectively detect only this fluorescence substance.

As shown in (e-h), it can be seen that Lamin A/C and GM130 were successfully separated and imaged in all four excitation lasers.

(i-1) shows this separation process. Comparing the image before separation (j) and the image after separation (k), it can be seen that the nuclear membrane image was successfully separated and removed from the image before separation.

The images in (e-h) were also successfully separated by a separation method using Gram-Schmidt orthogonalization (see Fig. 20). In addition, it was shown that the above separation method using the minimization of the amount of mutual information works even in the separation of two proteins labeled by direct staining.

In order to verify that the mutual information minimization accurately infers the α value, two images obtained independently from one place were merged using an artificial α value, then separated again by the mutual information minimization, and the original image and the separated image were compared (mp ). Comparing the original image (n) with the image after separation (p), it can be seen that there is no difference between the two images. The α value used for merging the images and the α value inferred by minimizing the amount of mutual information showed only a difference of 0.5%. Additionally, we tested the accuracy of α estimation by generating composite images with α values ranging from 0 to 3 in increments of 0.001. The mean error of α estimation was 0.81%. Next, the basic assumption of the present disclosure that the signal of the second fluorescent material is not detected in the first detection wavelength band was verified. First, the fluorescence signal brightness ratio between the two fluorescent materials in the first detection wavelength band was calculated from the emission spectrum of (a-d). The ratios were 178.7 for the 405-nm laser, 26.4 for the 488-nm laser, 8.5 for the 557-nm laser, and 17.7 for the 640-nm laser. In the 557-nm laser, the signal brightness ratio of the fluorescent material was the lowest, but in fact, when an image was obtained in the first detection wavelength band by immunostaining with these two fluorescent materials, the signal of the second fluorescent material was not seen. The reason is thought to be that among the two fluorescent materials CF568 and ATTO Rho101 used in the 557 nm laser, the absorption coefficient at 557 nm of ATTO Rho101, the second fluorescent material, is lower than that of CF568. To confirm that this assumption is valid for other microscopes, the fluorescence signal brightness of the above two fluorescence material solutions was measured in different microscope systems. As a result, the signal brightness ratio of the two fluorescent materials in the first detection wavelength band was 28.4. This signal brightness ratio can be further increased through signal amplification. Our research team showed that the signal intensity ratio in the first detection wavelength band could be improved to more than 100:1 by enhancing the signal of only the first fluorescent material among the two fluorescent materials with overlapping emission spectra (q-t).

8 is a result using the antibody produced in the same host according to an embodiment.

Referring to FIG. 8 , it is a signal separation result by multimolecular fluorescence imaging using a primary antibody produced in the same host and minimizing the amount of mutual information.

(a-d) is the verification of the immunostaining using the previously formed antibody complex. (a) Schematic diagram of the validation experiment. A mouse brain section including the hippocampus was stained with a preformed MAP2-labeled rabbit antibody complex and MAP2-labeled chicken antibody.

(b-d) Confocal microscopy images of brain slice samples. (b) the signal of the rabbit antibody complex formed in advance for labeling MAP2, (c) the signal of the chicken antibody labeling MAP2, and (d) the combined image of b and c.

(e-h) Validation of immunostaining using several preformed antibody complexes. (e) Schematic diagram of the validation experiment. Thy1-YFP mouse brain sections were stained with two preformed rabbit antibody complexes against YFP and calbindin. Only the preformed antibody complex that labels Calbindin has a fluorescent substance. (f-h) is a confocal microscope image of the sample, (f) the YFP signal of the sample, (g) the signal of the preformed antibody complex labeling Calbindin, and (h) f and g images are combined images.

(i-1) Signal separation results for multiplex immunostaining by antibody complexes with fluorescent substances with overlapping emission spectra. BS-C-1 cells were stained with two preformed rabbit antibody complexes against Lamin B1 and GM130. The two antibody complexes have only one fluorescent substance, either CF633 or Alexa 680. Two images were obtained in two detection wavelength bands under a 640-nm laser and separated by minimizing the amount of mutual information. (i) An image obtained from the first detection wavelength band, and only Lamin B1 is observed. (j) Both Lamin B1 and GM130 images obtained from the second detection wavelength band are observed, (k) the result of signal separation by minimizing the amount of mutual information, and (l) the combined image of i and k.

(m-n) Simultaneous imaging of mRNA and protein with one excitation laser using the present disclosure. GAPDH mRNA was labeled with Alexa 488 and vimentin protein was labeled with ATTO 514. (m) before image separation (red: vimentin; yellow: GAPDH mRNA), and (n) after image separation (red: vimentin; green: GAPDH mRNA).

(o-p) Four-color multi-molecular fluorescence imaging results of a rat brain section to which the SHIELD technique, a brain clearing technique, was applied with two excitation lasers. (o) before image separation (blue: NeuN and NF-H; red: GFAP and Calretinin) and (p) after image separation (blue: NeuN; green: GFAP; red: calretinin; purple: NF-H).

Next, by combining the present disclosure and antibody complex technology, it was confirmed whether it is possible to simultaneously observe several proteins with an antibody produced in the same host. In the conventional indirect immunostaining method, each protein had to be labeled with a primary antibody produced in a different host. However, since most antibodies are produced in 3 or 4 hosts including rabbits, it is necessary to use antibodies produced in the same host to simultaneously observe 10 or more proteins. Therefore, the present research team combined the present disclosure with the previously developed antibody complex technology. This technology enables the use of multiple primary antibodies produced in the same host by combining different primary antibodies with each Fab fragment secondary antibody attached to each other in a tube. As can be seen in (a-d), the rabbit antibody complex showed the same protein expression pattern as that of normal direct immunostaining using chicken antibody. In addition, as shown in (f-h), crosstalk between the antibody complexes did not appear when multiple proteins were imaged simultaneously using the antibody complex technique. In addition, it was confirmed that the antibody complex technology works well even in staining with chicken antibody complexes rather than rabbits. We stained 46 proteins using rabbit antibody complex technology, and confirmed that their expression patterns did not differ from known patterns. In addition, it was confirmed that, despite the size of the antibody complex being larger than that of the primary and secondary antibodies, it was possible to stain tissues with a similar thickness compared to the indirect staining method.

As shown in (i-1), by combining the antibody complex technology with the present disclosure, it was possible to image and separate the two proteins inside the cell into two antibodies produced in rabbits and two fluorescent substances with overlapping emission spectra. And the images thus obtained showed that the signals obtained by indirect immunostaining were consistent. The present disclosure relates to a technology for a fluorescent material, and is applicable to any technology using a fluorescent material. It was shown that this technique can separate proteins and mRNAs after labeling them with two fluorescent substances with overlapping emission spectra (m,n). In addition, it was shown that the multi-molecular imaging capability of the two techniques can be improved by applying the present disclosure to the tissue clearing technique (o,p) and the ultra-high-resolution imaging technique through tissue expansion.

9 is a result of simultaneous imaging of 10 proteins according to an embodiment.

Referring to FIG. 9 , it is a result of ultra-multi-molecular fluorescence imaging of a rat brain according to the present disclosure.

(a) multi-molecular fluorescence imaging of 10 proteins, (b) multi-molecular imaging of 8 proteins in the dentate gyrus of the mouse hippocampus, (c) b is an enlarged view of the area surrounded by dotted lines at the bottom, (d) ) is the structure of a general blood-brain barrier (BBB).

(e-h) An enlarged view of the dotted box area at the top of b to clearly show the cellular structure of the blood-brain barrier observed in (b).

(i-n) is an enlarged view of each protein in the region enclosed by the dotted line in c.

(o) 10-color multimolecular fluorescence imaging in the rat hippocampus.

(p-t) is an enlarged view of the area surrounded by the dotted rectangle of o. (p) A diagram showing 10 types of proteins at once, (qt) Multi-molecular fluorescence images of 3 or 2 colors after signal separation for each single excitation laser, (u) 10 types of proteins obtained from each excitation laser A list of target proteins in different colors.

Multiple proteins inside the tissue were simultaneously imaged using the present disclosure. First, imaging of 8 proteins was attempted, and rat brain sections were simultaneously stained with 8 rabbit antibody complexes. Stained rat brain sections were imaged at 8 detection wavelength bands while each excitation with 4 excitation lasers (a). Target proteins are PV, calbindin, NeuN, GFAP, GluT, ZNF3, laminin and calretinin (b). Through this, we confirmed the structure of the previously known blood-brain barrier (BBB; upper white box in b) in the mouse brain hippocampus, and the previously unknown protein expression diversity in cells (lower white box in b, enlarged is c) was able to confirm As can be seen in (d), the BBB generally consists of endothelial cells, basal lamina, and ends of astrocytes from the inside. As can be seen in (e-h), 8-color imaging clearly showed the cellular structure of these BBBs. In addition, as shown in (c), 8-color imaging showed a diversity of intercellular protein expression in the thalamus, which was not previously known. First, our research team compared the expression patterns of five proteins (calbindin, calretinin, ZNF3, NeuN, and PV) strongly observed in the thalamus with the literature on the expression patterns of each protein, and confirmed that they were consistent. . And how the combination of expression of these five proteins differs for each cell was studied. Comparing the protein expression of the two cells in the white box in (c), the cell on the left showed high expression of calbindin, calretinin and ZNF3, whereas the cell on the right showed high expression of NeuN and laminin (i-n). As such, the present disclosure shows the expression patterns of several previously known proteins in one cell, thereby making it possible to study the diversity of protein expression for each cell.

Further, 10 proteins can be imaged simultaneously. The rat brain sections were labeled with 10 rabbit antibody complexes each with 10 fluorescent substances attached thereto. After that, images were obtained in 10 different detection wavelength bands using 4 excitation lasers, and separated into 10 images containing only one protein signal by minimizing the amount of mutual information. In this experiment, NeuN, CNP1, calbindin, calretinin, GFAP, MAP2, NF-H, GluT, NF- L and PV were observed. As a result, 10 cell type markers and various biomolecules could be observed at once over several millimeters (o-p). All 10 proteins showed the same pattern as the previously known expression patterns of each protein. In particular, the expression levels of calbindin and calretinin were significantly changed at the boundary between the thalamus and midbrain (o dotted circle). These changes were consistent with those reported in various databases and literature. Calbindin is a calcium-binding protein and is highly expressed in the human brain. The observed expression pattern of Calbindin was high in the visualization of mossy fiber projection in the dentate gyrus (DG) and low in the thalamus (TH), which was consistent with the known literature.

Calretinin is a calcium-binding protein and is highly expressed in the human brain. The observed expression of Calretinin was in good agreement with the literature and was highly expressed in the molecular layer (DG-mo) and granule cell layer (DG-sg) of DG and in the thalamus (TH). ZNF3 is a transcription factor and is highly expressed in the human brain. ZNF3 in b was also highly expressed in DG-sg, polymorph layer (DG-po) and TH of DG. NeuN is a neuronal marker and is highly expressed only in the brain. As shown in (b), NeuN was expressed in DG-sg, DG-po, and TH, but the expression was higher in DG than in TH, which was consistent with the known literature. Parvalbumin (PV) is a calcium-binding albumin protein and is highly expressed in the human brain. The observed PV was widely expressed in DG-po, DG-sg, DG-mo and TH of DG, which is consistent with the known literature.

In addition, three-dimensional multi-molecular imaging was attempted. Thy1-YFP mouse brain sections were stained with DAPI and 5 rabbit antibody complexes, and then z-stack images were obtained at 7 detection wavelength bands and separated by minimizing the amount of mutual information. Through this, we succeeded in three-dimensional imaging of the distribution of 7 biomolecules (nucleus stained with DAPI, YFP that the brain section has by itself, and 5 proteins stained with antibody). In this experiment, the YFP protein and the CF488A fluorescent material were simultaneously excited with a 488-nm excitation laser, and the signals were separated. In this experiment, it was shown that it is possible to separate the signals of fluorescent proteins and fluorescent substances with overlapping emission spectra.

10 is a multi-molecular imaging result of a pathological sample according to an exemplary embodiment.

Referring to FIG. 10 , multimolecular fluorescence imaging in FFPE clinical samples according to the present disclosure.

(a-1) Two-color multimolecular fluorescence imaging of 12 types of human tissues using a single excitation laser (red: KRT 19 (CF488A); green: histone H3 (ATTO 514)). (a) breast, (b) salivary glands, (c) stomach; trunk, (d) small intestine; jejunum, (e) large intestine, (f) kidney; cortex, (g) kidney; medulla, (h) bladder, (i) prostate, (j) endometrium; proliferative phase, (k) endometrium; secretory, (l) thymus.

(mp) Four-color multimolecular fluorescence imaging using two excitation lasers on the same human tissue type (blue: KRT 19 (CF488A); green: histone H3 (ATTO 514); red: COX IV (CF568); yellow: vimentin) (ATTO Rho 101)). (m) salivary glands, (n) stomach; trunk, (o) small intestine; jejunum, (p) kidney; is water quality

It has been shown that the present disclosure is also applicable to pathological samples. A tissue microarray (TMA) containing Formalin-fixed paraffin-embedded (FFPE) samples from 12 organs was tested for operation in clinical samples using the present disclosure. First, keratin 19 and histone H3 were used as target proteins. Both of these proteins are highly expressed in almost all human organs, and their expression level changes or modifications are used as cancer markers. Since Keratin 19 and histone H3 are expressed at different locations within the cell (cytoplasm and nuclear protoplasm), if the signals of the two proteins are successfully separated, the two expression patterns should be spatially separated. As shown in (a-l), it can be seen that keratin 19 and histone H3 are clearly spatially separated in all 12 organs. These results show that two proteins can be simultaneously imaged with one excitation laser in a clinical sample using the present disclosure. Some nuclei appear yellow (i, j, l), which seems to indicate that histone H3 and keratin 19 under histone H3 were simultaneously detected due to the limitation of the axial resolution of the microscope.

Furthermore, expression of 4 proteins (keratin 19, histone H3, vimentin, cytochrome c oxidase subunit 4; COX IV) was imaged in FFPE samples from 4 organs. Like Keratin 19 and histone H3, vimentin and COX IV are also closely related to cancer and are used as cancer markers. Also, since the ratio of vimentin and keratin is related to the status of epithelial-mesenchymal transition (EMT), it is clinically important to image vimentin and keratin at the same time in a single sample. As shown in (m-p), four proteins could be observed simultaneously using two pairs of fluorescent materials and two excitation lasers with overlapping emission spectra in a pathological sample. The above experimental results show that the present disclosure works successfully even in pathological samples.

11 is a multi-molecular imaging result using a bandpass filter without a spectral detector according to an exemplary embodiment.

Referring to FIG. 11 , it is a 10-color multi-molecular fluorescence imaging result in a rat brain using a band-pass filter-based microscope without a spectral detector.

(a) 10-color multimolecular fluorescence imaging in the CA1 region of the rat hippocampus using the present disclosure. The above images were obtained using spinning-disk confocal microscopy equipped with 8 bandpass filters and 4 excitation lasers.

(be) The image shown in a is an image of three channels (bc) or two channels (de) obtained using each single laser among the four excitation lasers, and the wavelengths of the excitation lasers are 405 nm (b) and 488 nm (c). ), 561 nm (d) and 637 nm (e).

(f-h) 6 channels (f) or 5 channels (g-h) selected from the image shown in a (g-h) are used to clearly observe biologically relevant channels. Different colors were used for the proteins in each image to maximize visibility. All scale bars are 50 μm.

It has been shown that 10 proteins can be observed simultaneously with only a band-pass filter without a spectral detector. Among the many commercially available band filters, an optimal band filter that matches well with the detection wavelength band used was selected. From the spectral view, the brightness ratio between the two fluorescent materials in the first band filter in the 561 nm and 637 nm excitation lasers was not high enough. However, when this ratio was actually measured using a fluorescent material solution, the brightness ratio of the two fluorescent materials was 32.8 for the 561-nm laser and 20.3 for the 640-nm laser. This ratio can be improved by using a custom band-pass filter. It can also be further improved by using the signal amplification process shown in Fig. 7(q-t). In order to use 10 fluorescent substances at the same time, images were obtained in 10 detection wavelength bands, but a maximum of 8 band filters can be installed in a general microscope. Therefore, it was possible to image 10 proteins with 8 bandpass filters by repeatedly using one filter on two excitation lasers.

As shown in (a-e), using this method, we succeeded in imaging 10 proteins simultaneously using only rabbit antibodies in a mouse brain section. In order to more clearly indicate the biological meaning seen in the above results, 5-6 proteins out of 10 proteins were selected and displayed (f-h). In the white box (i) of (f), the arrangement of astrocytes (GFAP) and epithelial cells (GluT) in the BBB can be seen. Comparing the protein expression of two PV-positive neurons in boxes (ii) and (iii) in (f,g), neurons in (ii) express high levels of ZNF3 but low levels of MAP2. do. In contrast, neurons in (iii) express low ZNF3 but high MAP2 expression. (h) The white box (iv) can observe the location of the synapse through SV2A expression.

As such, in the present disclosure, it is possible to simultaneously observe the expression distribution of 10 proteins without a spectral detector using 8 bandpass filters. In addition, it was confirmed that these images were well separated even by a fluorescence measurement-based separation method (see Fig. 18).

The present disclosure can be combined with repeat staining multimolecular imaging technology (see FIG. 19 ). It was tested whether t-CyCIF, one of the repeat staining multimolecular imaging techniques, could be combined with the present disclosure to improve the multimolecular imaging ability of t-CyCIF. After imaging and image separation by labeling two proteins in a rat brain section with two fluorescent substances with overlapping emission spectra, the fluorescent substances were chemically inactivated, and the other two proteins were labeled with the same two fluorescent substances for imaging and image separation. The patterns of the four proteins obtained by the above procedure all matched the expected staining patterns. These results show that by combining the present disclosure with the repeat staining multimolecular imaging technique, 4 or more and up to 10 proteins can be observed simultaneously in one staining and labeling round.

As described above, since 10 types of fluorescent materials are applied to the sample at once, the complexity and time required for the experimental procedure are greatly reduced compared to other multimolecular imaging techniques. In addition, since images of 10 types of biomolecules can be obtained with one imaging without solution exchange or chemical treatment, there is no need for image matching and it is possible to obtain three-dimensional images. And since it can be implemented only with a band filter without using a spectral detector, the equipment requirements are greatly reduced. Therefore, it appears that the present disclosure can be very usefully used in biological research and diagnosis.

In the present disclosure, in addition to the four visible ray band lasers (400 - 650 nm) used above, a near-infrared (NIR) laser and two NIR fluorescent molecules (eg, ATTO740 and Cy7.5) with overlapping emission spectra are used. Thus, 12 biomolecules can be imaged with only one round of imaging. In addition, if a large stoke's shift fluorescent material excited at 560 nm, 640 nm, and near-infrared light is additionally used, 15 biomolecules can be imaged simultaneously. If a custom multi-band pass filter is used, 15 biomolecules can be simultaneously imaged with only 8 band filters. It can improve signal brightness and signal to noise ratio (SNR). In addition, the present disclosure can be implemented using a simple optical microscope equipped with a low-cost light source (LED or lamp) and a band filter. A custom bandpass filter can be used to achieve high signal brightness and isolation with simple equipment. In combination with various imaging techniques such as tissue clearing, expansion microscopy, mRNA or protein multimolecular fluorescence imaging, bioanalysis, and cell tracking, the imaging capabilities of these techniques can be improved. It can also be combined with fluorescent barcode technology to increase the amount of information that one barcode can encode. The present disclosure is expected to be usefully used in various fields requiring biomolecular spatial information. Recently, as a multi-scale three-dimensional atlas of biomolecules is constructed, the method of understanding biology is changing. In this atlas construction, fluorescence imaging is being used as one of the main tools for mapping biomolecules. The present disclosure can visualize more biomolecules at the same time, and provide information about colocation information between protein or mRNA molecules or correlation of their expression. In addition, visualization of the molecular structure of pathological samples will be helpful in the diagnosis and prognosis of cancer.

Cell culture and fixation were as follows.

BS-C-1, HeLa, and NIH-3T3 cells were purchased from the Korean Cell Line Bank. Cells were cultured in Nunc Lab-Tek II chamber # 1.5 coverslips. BS-C-1 cells were cultured in Minimum Essential Medium (MEM) supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and 1% sodium pyruvate. HeLa cells were cultured in Dulbecco's modified Eagles' medium (DMEM) supplemented with 10% FBS and 1% penicillin-streptomycin. NIH-3T3 cells were cultured in DMEM supplemented with 10% bovine calf serum (BCS) and 1% penicillin-streptomycin. All cells were cultured under 37 °C, 5% CO2 conditions. For fixation, the cells were washed 3 times with 1x PBS, fixed in 1x PBS with 4% paraformaldehyde (PFA) for 10 minutes, and washed 3 times with 1x PBS. Then, the cells were immersed in 1Х PBS with 0.1M glycine for 10 minutes and washed 3 times with 1ХPBS.

Mice perfusion and slicing are as follows.

가 유지되는 환경에서 4% PFA가 첨가된 1x PBS에 2시간동안 보관한다. 그런 뇌를 절편기(vibratome, Leica VT1000S)를 사용하여 150μm 두께로 잘랐다. 잘려진 뇌 절편은 사용하기전까지 4

에서 0.1M 글리신이 첨가된 1x PBS에 보관하였다.All of the following procedures involving animals were approved by the Korea Advanced Institute of Science and Technology (KAIST-IACUC). C57BL/6J mice aged 8-14 weeks were used. Mice are anesthetized with isoflurane, and the heart is fixed by perfusion of 1x PBS with 4% PFA. Brains were harvested from rats fixed by perfusion, 4

Store in 1x PBS with 4% PFA for 2 hours in an environment where The brain was cut into 150 μm thick using a vibratome (Leica VT1000S). The cut brain sections are stored until use.

was stored in 1x PBS supplemented with 0.1M glycine.

The binding between the fluorescent material and the antibody is as follows.

For binding Alexa and CF fluorescent material, add 9 times the number of moles of antibody solution to the antibody solution with 10 μL of 1 M sodium bicarbonate (pH 8.3) and succinimidyl ester-fluorophore stock dissolved in DMSO. This solution is reacted at room temperature and dark conditions for 1 hour. For ATTO fluorescence binding, the buffer of the antibody solution was exchanged with the binding buffer (0.01 M sodium bicarbonate in 1x PBS, pH 8.3) using Zeba spin desalting columns. After that, succinimidyl ester-fluorophore stock (10 mg/ml) dissolved in DMSO in the buffer-exchanged antibody solution is added to the antibody solution by 3 times the number of moles of antibody. This solution is reacted at room temperature and dark conditions for 1 hour. A NAP-5 gel filtration column was used to purify the bound antibody. The column was equilibrated with 10 mL of 1x PBS. Put 100 μL of the reacted solution into the columns. When 400 μL of 1x PBS is added to the column, a separation layer of unbound fluorescent material and bound antibody is observed. When a collection tube is placed under the column and 500 μL 1x PBS is added, the eluate is collected and concentrated using a Vivaspin column (MWCO: 5,000).

The formation of the antibody complex is as follows.

1x PBS, a solution containing the Fab fragment to which a fluorescent material is bound, and a solution containing the primary antibody were mixed in a volume ratio of 10:2:1 and reacted at room temperature for 10 minutes under dark conditions. After that, a 5-fold excess of blocking buffer (10% normal rabbit serum, 0.2% Triton X-100, and 1x PBS) is added to the solution, mixed gently under dark conditions, and reacted for 10 minutes at room temperature. Twenty-six kinds of fluorescent substances were tested on the preformed antibody complex.

Cells and mouse brain sections were stained using a preformed antibody complex as follows.

All subsequent steps were carried out at room temperature. For permeabilization and blocking, cells are soaked in blocking buffer (10% normal rabbit serum, 0.2% Triton X 100,1x PBS) for 30 min. Cells are then stained for 30 min in the preformed antibody complex mixture and washed 3 times with blocking buffer. For staining the rat brain sections, the blocking, staining, and washing steps are the same as in the cultured cell protocol, except for the incubation time (blocking: 3 hours, washing: 30 minutes, staining: overnight).

Sample preparation and pathology sample staining were as follows.

오븐에서 1시간 건조시켰다. 슬라이드를 xylene에 2회 각 5분간 탈파라핀했다. 수화를 위해 슬라이드를 실온에서 각 3분씩 100 % 에탄올 2회, 95 % 에탄올, 80 % 에탄올, 탈이온수에 순서대로 넣었다. 이 연구에 사용된 슬라이드는 염색 전에 heat induced epitope retrieval(HIER) 처리되었다. 슬라이드를 10 mM sodium citrate solution(10 mM Sodium Citrate, 0.05% Tween 20, pH 6.0)에 95 ~ 100

에서 30 분간 두고 1x PBS에서 20 분간 냉각했다. 슬라이드를 블로킹 버퍼(10 % normal rabbit serum, 0.2 % Triton X-100,1x PBS)로 RT에서 3시간 블로킹하고, 상온에서 밤새 항체 복합체와 반응시킨 후, 상온에서 블로킹 버퍼로 3회 각 30분씩 세척하였다.All experiments using human samples were approved by the Bioethics Review Committee of the Korea Advanced Institute of Science and Technology (KAIST IRB). Human normal multi-organ tissue microarray (TMA) used in FIG. 9 was purchased from Novus Biologicals (NBP2-30189). For formalin-fixed, paraffin-embedded pathology samples, slide the sample slides to 60

It was dried in an oven for 1 hour. The slides were deparaffinized twice in xylene for 5 minutes each. For hydration, slides were placed in 100% ethanol twice, 95% ethanol, 80% ethanol, and deionized water for 3 minutes each at room temperature in that order. The slides used in this study were subjected to heat induced epitope retrieval (HIER) prior to staining. Slide 95-100 in 10 mM sodium citrate solution (10 mM Sodium Citrate, 0.05% Tween 20, pH 6.0)

incubated for 30 min and cooled in 1x PBS for 20 min. The slides were blocked with blocking buffer (10% normal rabbit serum, 0.2% Triton X-100,1x PBS) at RT for 3 hours, reacted with the antibody complex overnight at room temperature, and then washed 3 times with blocking buffer at room temperature for 30 minutes each. did

RNA-FISH is as follows.

Advanced Cell Diagnostics (RNAscope) was performed with NIH-3T3 cells according to the manufacturer's instructions. Briefly, fixed cells were dehydrated with ethanol and rehydrated with ethanol and 1x PBS. Rehydrated cells are digested by diluting Protease III (RNAscope) at a ratio of 1:15 with 1x PBS, and then rinsed with 1x PBS. Before processing the RNAscope amplifier, the cells are reacted with the RNAscope probe and rinsed with 1x RNAscope wash buffer. After amplification, the cells are labeled with DAPI and rinsed in 1x PBS. The cells were then stained with preformed antibody complexes.

Inactivation of the fluorescent material (t-CyCIF) is as follows.

In the first staining round, mouse brain sections were stained using an antibody complex labeled with PV (labeled CF488A) and an antibody complex labeled with NeuN (labeled ATTO514). DAPI was used as a reference marker for the sample. After the first imaging round, the fluorescent material is inactivated by immersing the fluorescent material-stained sample in an inactivation buffer (4.5% H2O2 and 24 mM NaOH in 1x PBS) at room temperature for 2 hours under white light. Then, the samples were washed 4 times each 1 hour with 1x PBS, and a second staining round was performed. In the second staining, two types of antibody complexes labeling calretinin (labeled as CF488A) and GluT were used in the same sample. DAPI was also used again.

Signal amplification using an anti-fluorophore antibody is as follows.

All antibodies used in this section were diluted 1:500 with blocking buffer (5% normal donkey serum, 0.2% Triton X-100,1x PBS). BS-C-1 cells are soaked in blocking buffer for 1.5 h for permeabilization and blocking. Subsequently, a solution containing a rabbit primary antibody (rabbit anti-GM130) for labeling GM130 was added and mixed, and each was washed three times with blocking buffer every 5 minutes. Then, the cells were mixed in a solution with a donkey anti-rabbit conjugated with alexa488 conjugated to alexa488, a fluorescent substance, and washed 3 times every 5 minutes with a blocking buffer. Cells were immersed in a solution containing rabbit anti-alexa488 antibody, and then washed three times with blocking buffer for 5 minutes each. Finally, the cells were immersed in a solution with the above secondary antibody (donkey anti-rabbit conjugated with alexa488), and then washed 3 times for 5 minutes each with blocking buffer. After signal amplification for the rabbit primary antibody, the cells were mixed in a solution with a mouse anti-laminA/C labeling lamin A/C and washed 3 times with blocking buffer for 5 minutes each. Finally, the cells were immersed in a solution containing ATTO514-conjugated donkey anti-mouse conjugated with ATTO514, and then washed three times with blocking buffer for 5 minutes each.

Fluorescent material brightness measurement is as follows.

Fluorescence brightness measurements were performed on an Andor rotating disk confocal microscope. For CF405S, ATTO390, and CF405L, 100 μM fluorescent material solution was used. For CF488A, ATTO514, ATTO490LS, CF568, ATTO Rho101, CF633, and CF660R, a 10 μM fluorescent material solution was used. Fluorescent material solutions were imaged using one of four excitation lasers (405-nm, 488-nm, 561-nm, 637-nm) at a 10× objective.

Protein-retention expansion microscopy is as follows.

에서 반응시킨다. 시료를 젤 버퍼(monomer solution, 0.2%(w/w) APS, 0.2%(w/w) TEMED, 0.01%(w/w) H-TEMPO)가 채워진 2개의 커버 글라스 사이에 놓고, 37

에서 1.5시간 반응시킨다. 만들어진 젤을 digestion buffer(25 mM EDTA, 50 mM Tris, 0.5% Triton X-100, 1M NaCl)에 1:100으로 희석한 proteinase K를 처리해주고 밤새 37

에서 부드럽게 섞어준다. Digestion 후 젤을 탈이온수에 넣고 부드럽게 섞어준다.The stained sample was reacted with acryloyl-X, SE (AcX) diluted in 1x PBS overnight at room temperature, and then washed 3 times with 1x PBS for 30 minutes each. Then, the sample was soaked in monomer solution (7.5% (w/w) sodium acrylate, 2.5% (w/w) acrylamide, 0.15% (w/w) bis-acrylamide, 1x PBS, 2 M NaCl) twice for 30 minutes each. 4

react in The sample was placed between two cover glasses filled with gel buffer (monomer solution, 0.2% (w/w) APS, 0.2% (w/w) TEMED, 0.01% (w/w) H-TEMPO), 37

reacted for 1.5 hours. The prepared gel was treated with proteinase K diluted 1:100 in digestion buffer (25 mM EDTA, 50 mM Tris, 0.5% Triton X-100, 1M NaCl) and treated with proteinase K overnight.

mix gently. After digestion, put the gel in deionized water and mix gently.

SHIELD is as follows.

에서 하루 동안 부드럽게 섞어준다. 시료을 SHIELD-ON Buffer와 SHIELD-Epoxy Solution 7 : 1 비율의 혼합액에 옮기고 37

에서 6 시간 동안 부드럽게 섞어준다. 시료는 SHIELD-ON 버퍼에 담가 37 °C에서 하룻밤동안 부드럽게 섞어준다. 조직 투명화를 위해, 시료를 투명화 용액(300 mM sodium dodecyl sulfate, 10 mM boric acid, 100 mM sodium sulfite, pH 9.0)에 담가 37

에서 1 일간 부드럽게 섞어준다. 이미징은 Nikon C2 plus 공초점 현미경 with DUVB 디텍터, Leica TCS SP8, Andor Dragonly회전 디스크 공초점 현미경에서 40x 1.15 NA water immersion를 사용하여 얻었다.Immerse a 500 μm thick rat brain section in SHIELD-OFF solution (1 : 1 : 2 ratio of DI water, SHIELD buffer, SHIELD-Epoxy solution) 4

Mix gently throughout the day. Transfer the sample to a mixture of SHIELD-ON Buffer and SHIELD-Epoxy Solution 7: 1

Gently mix for 6 hours. Samples are immersed in SHIELD-ON buffer and gently mixed overnight at 37 °C. For tissue clearing, the sample was immersed in a clearing solution (300 mM sodium dodecyl sulfate, 10 mM boric acid, 100 mM sodium sulfite, pH 9.0) 37

Mix gently for 1 day. Imaging was obtained on a Nikon C2 plus confocal microscope with DUVB detector, Leica TCS SP8, Andor Dragonly rotating disk confocal microscope using 40x 1.15 NA water immersion.

Image processing is as follows.

The images obtained from the spectral detector were directly applied with the spectral separation algorithm, and the images obtained with a confocal microscope equipped with a bandpass filter were pre-processed before signal separation. For pre-processing, first, vignetting of the image was corrected using the image of a fluorescent slide. Second, the ring-shaped artifact on the image due to the interference of the excitation laser was removed by subtracting the image taken without placing any object. Third, pixel shift due to chromatic aberration was resolved through image registration. For signal separation by minimizing the amount of mutual information, we started with the initial estimation of α and measured the amount of mutual information between IMG1 and (IMG2-αxIMG1). Here, IMG1 and IMG2 are the first and second images, respectively. When measuring the amount of mutual information, both images were quantized to 3 bits to improve the joint histogram calculation speed. Since we use millions of pixels for one parameter estimation, quantization does not affect the α estimation accuracy. Using gradient descent, we found the optimal value of α with the minimum amount of mutual information. After finding the optimal α _opt , (IMG2-αopt x IMG1) is calculated to obtain the image of the second fluorescent material. For color separation through Gram-Schmidt orthogonalization, α _GS is estimated as shown in Equation 12 below.

와

는 IMG1과 IMG2를 각각 벡터로 재정리한 것이다. 그리고 위와 동일하게 두 번째 형광 물질의 이미지를 얻기 위해 (IMG2-α_GSxIMG1)를 계산해준다. 위의 모든 이미지 프로세싱은 자체 제작한 MATLAB 코드를 사용하여 실행했다.where <,> is the vector dot product,

Wow

is the rearrangement of IMG1 and IMG2 into vectors, respectively. And in the same way as above, (IMG2-α _GS xIMG1) is calculated to obtain the image of the second fluorescent material. All of the above image processing was performed using our own MATLAB code.

12 is a validation result of antibody complex staining.

Referring to FIG. 12 , in order to verify the affinity and specificity of the antibody complex, the existing antibody and antibody complex are stained for one target protein at the same time, and the fluorescence signal is observed. From left to right, the conventional antibody staining, the antibody complex staining, and the two images combined are shown.

(a) Confocal microscopy images of Thy1-YFP rat brain sections labeled with GFP antibody, (b) Confocal microscopy images of GFAP in rat brain sections, (c) Confocal microscopy images of MBPs in the hippocampal region of rat brain sections , (d) Confocal images of vimentin in BS-C-1 cells. Scale bars: (a-c) 50 μm, (d) 20 μm.

13 is a result of verifying the presence or absence of crosstalk between antibody complexes.

Referring to FIG. 13 , it is an additional experimental result image of the crosstalk verification image shown in FIG. 8 (e-h). Thy1-YFP mouse brain sections are stained with YFP antibody complex and neurofilament heavy chain (NF-H) antibody complex.

(a) YFP fluorescence signal, (b) NF-H stained with antibody complex (CF568, red), (c) a and b combined images. When it can be seen that there is no crosstalk between a and b, it can be seen that normal rabbit serum (NRS) effectively blocks the unbound Fab fragments antibody. Scale bar: 50 μm.

14 is a validation result of signal separation by minimizing the amount of mutual information in a sample stained using an antibody complex.

Referring to FIG. 14, a mouse brain section was stained with a calretinin (rabbit, CF568 conjugated) antibody complex and GFAP (rabbit, ATTO Rho101 conjugated) antibody complex, and indirectly stained with an GFAP (chicken, visualized by CF633) antibody.

(a) Schematic of the signal separation effectiveness experiment, (b) when excited at 557-nm in the first detection wavelength band, only the calretinin signal is seen. (c) When excited at 557-nm in the second detection wavelength band, both the calretinin signal and the GFAP (rabbit) signal are visible. (d) Signal-separated GFAP (rabbit) image, (e) GFAP (chicken) image obtained by general indirect immunostaining method, and (f) b, d, e are combined images.

15 is a result of verifying compatibility between the present disclosure and an expansion microscopy (ExM).

15 , it is a signal separation image of a rat brain section to which an expansion microscope is applied. (a) before signal separation (red: NeuN and GFAP), (b) after signal separation (red: NeuN, green: GFAP), (c) before signal separation (red: Neurofilament heavy chain (NF-H) and GluT) , (d) after signal separation (red: NF-H, green: GluT). Scale bar: 50 μm.

16 is a linearity graph of signal brightness according to laser exposure time, and FIG. 17 is a linearity graph of signal brightness according to laser intensity.

Referring to FIG. 16 , the brightness of 10 fluorescent materials according to the laser exposure time was measured in an Andor rotating disk confocal microscope using an optimized optical filter.

(a-c) CF405S (red), ATTO390 (green), and CF405L (blue) measurement results.

(d-f) CF488A (red), ATTO514 (green), ATTO490LS (blue) measurement results.

(g-h) CF568 (red) and ATTO Rho101 (green) measurement results.

(i-k) CF633 (red) and CF660R (green) measurement results.

Referring to FIG. 17 , brightness according to laser intensity of ten fluorescent materials used in the present disclosure is measured in an Andor rotating disk confocal microscope using an optimized optical filter.

(a-c) CF405S (red), ATTO390 (green), and CF405L (blue) measurement results.

(d-f) CF488A (red), ATTO514 (green), ATTO490LS (blue) measurement results.

(g-h) CF568 (red) and ATTO Rho101 (green) measurement results.

(i-k) CF633 (red) and CF660R (green) measurement results.

18 is a diagram for explaining signal separation based on measurement of a fluorescent substance.

Referring to FIG. 18 , (a-d) In signal separation based on measurement of fluorescence, the measured α value of CF488A ( FIGS. 16 and 17 ) is corrected based on the laser intensity and exposure time set during imaging to separate signals. (a) SV2A, (b) before signal separation (SV2A and GluT), (c) after signal separation (GluT), (d) a and c combined images. Scale bar: 50 μm.

19 is a diagram for explaining application of the present disclosure to another multimolecular imaging technique (t-CyCIF).

Referring to FIG. 19 , (a) is a conceptual schematic diagram of applying the fluorescent material inactivation technology (t-CyCIF) and the present disclosure together.

(b-c) Confocal microscopy images of the hippocampus of a rat brain slice. (b) In the first imaging round, PV (CF488A, red), NeuN (ATTO514, green), and DAPI (blue) were labeled, and signal separation was performed by minimizing the amount of mutual information. (c) After inactivation of the fluorescent material, calretinin (CF488A, red), GluT (ATTO514, green) and DAPI (blue) were labeled in the second imaging round, and signal separation was performed by minimizing the amount of mutual information. Scale bar: 50 μm.

20 is a signal separation result through Gram-Schmidt orthogonalization of the present disclosure.

Referring to FIG. 20 , it is a multimolecular fluorescence image obtained by signal separation through Gram-Schmidt orthogonalization of the original images before separation of FIG. 7 (e-h). After staining by targeting Lamin A/C (image obtained in the first detection wavelength band), GM130 (image obtained after separating the image in the first detection wavelength band from the image obtained in the second detection wavelength band), and vimentin (image obtained with large stoke's shift fluorescent material) signal was separated. Scale bar: 30 μm.

Table 2 below is a list of primary antibodies that have been validated that the present disclosure works.

# Antibody Vendor Catalog# Host Clonality mouse brain marker One ABAT ATLAS HPA041690 Rb Poly 2 α-internexin/NF66 EnCor RPCA-α-Int Rb Poly 3

-tubulin

-tubulin Abcam ab6046 Rb Poly 4 ARFGEF1 ATLAS HPA023822 Rb Poly 5 Calbindin Abcam ab11426 Rb Poly 6 Calretinin Abcam ab702 Rb Poly 7 CAMK2B ATLAS HPA026307 Rb Poly 8 DDX3X ATLAS HPA001648 Rb Poly 9 E1F1AX ATLAS HPA002561 Rb Poly 10 FGF3 ATLAS HPA012692 Rb Poly 11 GABA-A receptor α1 SYSY 224 203 Rb Poly 12 GAD 65/67 Millipore AB1511 Rb Poly 13 GFAP HPA HPA056030 Rb Poly 14 GFP Abcam ab290 Rb Poly 15 GluR1 Abcam ab31232 Rb Poly 16 Iba1 SYSY 234 003 Rb Poly 17 INA ATLAS HPA008057 Rb Poly 18 Lamin B1 Abcam ab16048 Rb Poly 19 Laminin Abcam ab11575 Rb Poly 20 Laminin EnCor RPCA-Laminin Rb Poly 21 LHX2 ATLAS HPA000838 Rb Poly 22 MAP2 Abcam ab32454 Rb Poly 23 MBP Abcam ab40390 Rb Poly 24 MBP HPA HPA049222 Rb Poly 25 MBP Aves MBP Chk Poly 26 NECAB2 ATLAS HPA013998 Rb Poly 27 NeuN Millipore ABN78 Rb Poly 28 Neurofilament 200 Sigma N4142 Rb Poly 29 Neurofilament-M EnCor RPCA-NF-M Rb Poly 30 Neuropeptide Y Immunostar 22940 Rb Poly 31 Parvalbumin Novus NB120-11427 Rb Poly 32 PCP4 ATLAS HPA005792 Rb Poly 33 RAP1GAP ATLAS HPA001922 Rb Poly 34 SLC2A1 ATLAS HPA031345 Rb Poly 35 Sox2 SYSY 347 003 Rb Poly 36 Somatostatin HPA HPA019472 Rb Poly 37 SV2A Abcam ab32942 Rb Poly 38 TBR1 Abcam ab31940 Rb Poly 39 TH Abcam ab112 Rb Poly 40 VAMP2 abcam ab3347 Rb Poly 41 ZNF3 ATLAS HPA003719 Rb Poly cell marker 42

-tubulin Abcam ab6046 Rb Poly 43 Clathrin heavy chain Abcam ab21679 Rb Poly 44 Collagen IV Abcam ab6586 Rb Poly 45 Fibronectin Abcam ab2413 Rb Poly 46 GM130 Abcam ab52649 Rb Mono 47 LaminB1 Abcam ab16048 Rb Poly 48 Pericentrin Abcam ab4448 Rb Poly human tissue markers 49 KRT19 ATLAS HPA002465 Rb Poly 50 COX IV Abcam ab16056 Rb Poly 51 Histone H3 Abcam ab1791 Rb Poly 52 Vimentin Abcam ab45939 Rb Poly 53 LaminB1 Abcam ab16048 Rb Poly 54 Fibronectin Abcam ab2413 Rb Poly 55 Collagen IV Abcam ab6586 Rb Poly

[Table 3] is a list of fluorescent materials tested by the present disclosure.

# Dye Vendor Catalog# One ATTO390 ATTO-TEC AD 390-31 2 CF405S Biotium #92110 3 CF405M Biotium #92111 4 CF405L Biotium #92112 5 Dylight405 Life technologies 46400 6 ATTO430LS ATTO-TEC AD 430LS-31 7 CF488A Biotium #92120 8 ATTO488 ATTO-TEC AD 488-31 9 Alexa488 Life technologies A20000 10 ATTO490LS ATTO-TEC AD 490LS-31 11 ATTO514 ATTO-TEC AD 514-31 12 CF514 Biotium #92103 13 ATTO532 ATTO-TEC AD 532-31 14 Alexa546 Life technologies A20002 15 ATTO565 ATTO-TEC AD 565-31 16 CF568 Biotium #92131 17 ATTO Rho101 ATTO-TEC AD Rho101-31 18 Alexa594* Jackson
ImmunoResearch 111-587-008 19 ATTO594 ATTO-TEC AD 594-31 20 ATTO633 ATTO-TEC AD 633-31 21 CF633 Biotium #92133 22 Alexa647* Jackson
ImmunoResearch 111-607-008 23 ATTO647N ATTO-TEC AD 647N-31 24 Alexa680* Jackson
ImmunoResearch 111-627-008 25 CF660R Biotium #92134 26 CF680R Biotium #92107

The method of operating the electronic device 200 according to various embodiments of the present disclosure includes acquiring at least one image for a plurality of fluorescent materials each labeled with different biomolecules ( 310 ), and acquiring It may include a step 320 of separating the obtained image into images for each of the biomolecules.

According to various embodiments, at least any two of the fluorescent materials may respectively exhibit emission spectra of a similar wavelength band.

According to various embodiments, the wavelength band includes a first detection wavelength band corresponding to a portion of any one of the emission spectra, and a second wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other of the emission spectra. It may include two detection wavelength bands. According to various embodiments, acquiring an image 310 may include acquiring a first image in a first detection wavelength band, and acquiring a second image in a second detection wavelength band.

According to various embodiments, the step of separating the obtained image 320 may include obtaining a third image from a second image based on the first image.

According to various embodiments, the first image may be detected as one image among biomolecules, and the third image may be detected as another image of biomolecules.

According to various embodiments, the acquiring of the third image may include acquiring the third image by minimizing information shared between the first image and the second image from the second image.

According to various embodiments, the acquiring of the third image may include predicting a ratio of the brightness of any one of the fluorescent materials in the first detection wavelength band to the brightness of any one of the fluorescent materials in the second detection wavelength band. and obtaining a third image from the second image based on the ratio.

According to various embodiments, the obtaining of the third image based on the ratio may include obtaining the third image by subtracting a result of multiplying the ratio by the first image from the second image.

According to various embodiments, the step of acquiring an image 310 may further include generating an image of emission spectra of at least any two of the fluorescent materials as light in a wavelength band is irradiated to the fluorescent materials. and the first image and the second image may be respectively obtained from the generated images.

According to an embodiment, the first image and the second image may be respectively obtained by a bandpass filter.

According to another embodiment, the first image and the second image may be respectively acquired by a spectral detector.

The electronic device 200 according to various embodiments of the present disclosure includes a memory 240 and a processor 250 connected to the memory 240 and configured to execute at least one instruction stored in the memory 240 . and the processor 250 may be configured to acquire at least one image for a plurality of fluorescent materials each labeled with different biomolecules, and to separate the obtained image into images for each of the biomolecules. can

According to various embodiments, at least any two of the fluorescent materials may respectively exhibit emission spectra of a similar wavelength band.

According to various embodiments, the wavelength band includes a first detection wavelength band corresponding to a portion of any one of the emission spectra, and a second wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other of the emission spectra. It may include two detection wavelength bands.

According to various embodiments, the processor 250 may be configured to acquire a first image in a first detection wavelength band and to acquire a second image in a second detection wavelength band.

According to various embodiments, the processor 250 obtains a third image from a second image based on the first and second images, detects the first image as any one image among biomolecules, and 3 images may be configured to detect an image of the other of the biomolecules.

According to various embodiments, the processor 250 may be configured to obtain the third image by minimizing information shared between the first image and the second image from the second image.

According to various embodiments, the processor 250 predicts a ratio of the brightness of any one of the fluorescent materials in the first detection wavelength band to the brightness of any one of the fluorescent materials in the second detection wavelength band, and the ratio may be configured to obtain a third image from the second image.

According to various embodiments, the processor 250 may be configured to obtain the third image by subtracting a result of multiplying the ratio by the first image from the second image.

According to various embodiments, the processor 250 generates an image of emission spectra of at least two of the fluorescent materials as light in a wavelength band is irradiated to the fluorescent materials, and a first image from the generated image and a second image, respectively.

According to an embodiment, the first image and the second image may be respectively obtained by a bandpass filter.

According to another embodiment, the first image and the second image may be respectively acquired by a spectral detector.

The embodiment of the present disclosure described above is not implemented only through the apparatus and method, and may be implemented through a program for realizing a function corresponding to the configuration of the embodiment of the present disclosure or a recording medium in which the program is recorded.

Although the embodiments of the present disclosure have been described in detail above, the scope of the present disclosure is not limited thereto, and various modifications and improvements using the basic concept of the present disclosure defined in the following claims are also provided. will belong to

Claims (20)

A method of operating an electronic device, comprising:
acquiring at least one image for a plurality of fluorescent materials each labeled with different biomolecules; and
Separating the acquired image into images for each of the biomolecules,
At least any two of the fluorescent materials,
each of the emission spectra of a similar wavelength band,
The wavelength band,
a first detection wavelength band corresponding to a portion of any one of the emission spectra, and
a second detection wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other one of the emission spectra;
Acquiring the image includes:
acquiring a first image in the first detection wavelength band; and
acquiring a second image of the second detection wavelength band;
Separating the acquired image comprises:
based on the first image, obtaining a third image from the second image;
The first image is detected as any one image of the biomolecules,
the third image is detected as an image of the other one of the biomolecules;
The step of obtaining the third image is
A method of obtaining the third image by orthogonalizing the first image and the second image.
delete delete delete delete delete A method of operating an electronic device, comprising:
acquiring at least one image for a plurality of fluorescent materials each labeled with different biomolecules; and
Separating the acquired image into images for each of the biomolecules,
At least any two of the fluorescent materials,
each of the emission spectra of a similar wavelength band,
The wavelength band,
a first detection wavelength band corresponding to a portion of any one of the emission spectra, and
a second detection wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other one of the emission spectra;
Acquiring the image includes:
acquiring a first image in the first detection wavelength band; and
acquiring a second image of the second detection wavelength band;
Separating the acquired image comprises:
based on the first image, obtaining a third image from the second image;
The first image is detected as any one image of the biomolecules,
the third image is detected as an image of the other one of the biomolecules;
The step of obtaining the third image is
predicting a ratio of the brightness of any one of the fluorescent materials in the first detection wavelength band to the brightness of any one of the fluorescent materials in the second detection wavelength band; and
based on the ratio, obtaining the third image from the second image.
8. The method of claim 7, wherein, based on the ratio, obtaining the third image comprises:
and obtaining the third image by subtracting the result of multiplying the ratio by the first image from the second image.
The method of claim 8, wherein acquiring the image comprises:
The method further comprises generating an image of emission spectra of at least any two of the fluorescent materials as the light in the wavelength band is irradiated to the fluorescent materials,
wherein the first image and the second image are respectively obtained from the generated image.
10. The method of claim 9, wherein the first image and the second image,
A method obtained by a bandpass filter or a spectral detector, respectively.
In an electronic device,
Memory; and
a processor coupled to the memory and configured to execute at least one instruction stored in the memory;
The processor is
For a plurality of fluorescent materials each labeled with different biomolecules, at least one image is acquired,
configured to separate the acquired image into images for each of the biomolecules,
At least any two of the fluorescent materials,
each of the emission spectra of a similar wavelength band,
The wavelength band,
a first detection wavelength band corresponding to a portion of any one of the emission spectra, and
a second detection wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other one of the emission spectra;
The processor is
acquiring a first image of the first detection wavelength band;
acquiring a second image of the second detection wavelength band;
Based on the first image, obtaining a third image from the second image,
detecting the first image as an image of any one of the biomolecules;
configured to detect the third image as an image of the other one of the biomolecules;
The processor is
and obtain the third image by orthogonalizing the first image and the second image.
delete delete delete delete delete In an electronic device,
Memory; and
a processor coupled to the memory and configured to execute at least one instruction stored in the memory;
The processor is
For a plurality of fluorescent materials each labeled with different biomolecules, at least one image is acquired,
configured to separate the acquired image into images for each of the biomolecules,
At least any two of the fluorescent materials,
each of the emission spectra of a similar wavelength band,
The wavelength band,
a first detection wavelength band corresponding to a portion of any one of the emission spectra, and
a second detection wavelength band in which at least a portion of any one of the emission spectra overlaps with at least a portion of the other one of the emission spectra;
The processor is
acquiring a first image of the first detection wavelength band;
acquiring a second image of the second detection wavelength band;
Based on the first image, obtaining a third image from the second image,
detecting the first image as an image of any one of the biomolecules;
configured to detect the third image as an image of the other one of the biomolecules;
The processor is
predicting a ratio of the brightness of any one of the fluorescent materials in the first detection wavelength band to the brightness of any one of the fluorescent materials in the second detection wavelength band;
and obtain the third image from the second image based on the ratio.
The method of claim 17, wherein the processor comprises:
and obtain the third image by subtracting the result of multiplying the ratio by the first image from the second image.
The method of claim 18, wherein the processor comprises:
As the light of the wavelength band is irradiated to the fluorescent materials, an image of at least any two emission spectra of the fluorescent materials is generated,
an apparatus configured to obtain the first image and the second image from the generated image, respectively.
The method of claim 19, wherein the first image and the second image,
A device obtained by a bandpass filter or a spectral detector, respectively.

Images