Background hotspot reduction for plasmon enhanced fluorescence biosensing
The optical imaging system addresses the challenge of background hotspots in plasmonic chips by distinguishing sample signals from noise through time-based image analysis and particle detection, enhancing the detection of small extracellular vesicles with improved signal-to-noise ratio and accuracy.
Patent Information
- Application Number
- JP2024077569
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-12
- Filing Date
- 2024-05-10
- Publication Date
- 2024-12-03
AI Technical Summary
Plasmonic chips used in fluorescence microscopy for detecting small extracellular vesicles face challenges due to unstable background hotspots that are similar in size and intensity to the signals from the sample, making it difficult to distinguish between noise and signal, especially when using long exposures to improve the signal-to-noise ratio.
An optical imaging system that includes a fluorescence microscope, image acquisition controller, and image processing unit to distinguish between signals from the sample and background hotspots by analyzing changes over time in multiple images, setting optimal exposure times, and applying particle detection algorithms to create and update particle maps, thereby reducing background hotspot noise.
The system effectively reduces background hotspot noise, improving the signal-to-noise ratio and enabling accurate detection of small extracellular vesicles even at low concentrations, minimizing errors in analysis.
Smart Images

Figure 2024167165000001_ABST
Abstract
Description
[Background technology]
[0001] This application claims priority to U.S. patent application Ser. No. 63 / 501,874, filed May 12, 2023, the entire contents of which are incorporated herein by reference. [Technical field]
[0002] The present disclosure relates to plasmon-enhanced biosensing, and more specifically, illustrates plasmon-enhanced fluorescence biosensing in which images are processed to remove background hot spot noise.
[0003] Extracellular vesicles (EVs) in body fluids are considered promising materials for liquid biopsy, because EVs contain a variety of proteins, miRNAs, and mRNAs on their surface or inside the vesicles. EVs are classified into several subtypes, including exosomes, microvesicles, microparticles, ectosomes, oncosomes, and apoptotic bodies, and their sizes vary from about 50 nm to 5 μm. Liquid biopsy of small EVs smaller than 200 nm is actively investigated, but most of the studies use bulk analysis to analyze EV properties because EVs are too small to be observed by common optical microscopes. Fluorescence microscopy is one method that can make single EVs observable, but small EVs have low fluorescence intensity and are difficult to observe. One method to detect small single EVs is plasmon-enhanced fluorescence biosensing, which uses a plasmonic chip instead of the classical glass plate. Many different types of plasmonic tips have been developed, and some types of plasmonic tips can produce unstable background hotspots that are similar in size and intensity to the signal from EVs.
[0004] Sensing using surface plasmons has been discussed, for example, in S. Park, et al., (“Self-Assembly of Nanoparticle-Spiked Pillar Arrays for Plasmonic Biosensing,” Adv. Funct. Mater., 2019, Vol. 29, Issue 43, 1904257), Keiko Tawa, et al., (“Optical microscopic observation of fluorescence enhanced by grating-coupled surface plasmon resonance,” Opt. Express 16, 9781-9790 (2008)), Taylor, AB and Zijlstra, P, (ACS Sens. 2017, 2, 8, 1103-1122), and US Pat. 8,039,267. Surface plasmons are also described in WO 2021 / 211756, which is incorporated herein by reference in its entirety. Summary of the Invention
[0005] According to at least one embodiment of the present disclosure, a system, device, and method are provided that includes an optical imaging system for plasmon-enhanced optical imaging. The optical imaging system includes a fluorescence microscope that images a sample on a plasmonic substrate, an image acquisition controller that acquires a plurality of images in a first channel at different times, and an image processor that differentiates a signal in one or more of the plurality of images as (1) a signal from the sample or (2) a signal from a background hotspot, which may be based on a change over time in one or more of the plurality of images. The image processor differentiates the signal by performing a particle detection analysis on the first image to create a base particle map, performing a particle detection analysis on the at least one second image to create a temporary particle map, and defining a particle in the base particle map as a background hotspot if the particle does not have a corresponding particle in the temporary particle map at a location within a location threshold of the particle, and defining a particle in the base particle map as a signal from the sample if the particle has a corresponding particle in the temporary particle map at a location within a location threshold of the particle. The image processor may perform a number of these processes in an iterative manner to update the base particle map with information from the additional images. The image processor may, for example, display the base particle map or the updated particle map.
[0006] In yet another embodiment of the present disclosure, a system, device, and method including an optical imaging system for plasmonic enhanced optical imaging is provided. The optical imaging system includes a fluorescent microscope for imaging a sample on a plasmonic substrate, an image acquisition controller for acquiring a plurality of images in a first channel at different times, and an image processor for distinguishing a signal in one or more of the plurality of images as (1) a signal from the sample or (2) a signal from a background hotspot. The change may be based on a change in one or more of the plurality of images over time. The image acquisition controller sets an acquisition time for the plurality of images, the acquisition time being set to be longer than a bleaching time of a fluorescent dye in the sample. Specifically, the acquisition time is 100 ms to 5 s or 600 ms to 2 s, and at least 2, 4, 6, 8, 10, or more images are acquired in the first channel at different times.
[0007] These and other objects, features and advantages of the present disclosure will become apparent from the following detailed description of illustrative embodiments of the disclosure, which should be read in conjunction with the accompanying drawings and the appended claims. [Brief description of the drawings]
[0008] Further objects, features and advantages of the present disclosure will become apparent from the following detailed description taken in conjunction with the accompanying drawings illustrating illustrative embodiments of the present disclosure. [Figure 1] FIG. 1 shows a plot of the noise in percentage terms for different exposure times. [Diagram 2] FIG. 2 is a system diagram. [Diagram 3] FIG. 3 is a flow chart of one embodiment of this process. [Figure 4] FIG. 4 is a flow chart for background hot spot noise reduction. [Diagram 5] 5(A) to 5(D) are examples of various particle maps. [Figure 6] FIG. 6 is a flow chart illustrating one embodiment of background hotspot noise reduction. [Figure 7] FIG. 7 is a flow chart illustrating one embodiment of background hot spot noise reduction. [Figure 8] FIG. 8 is a flow chart illustrating one embodiment of background hot spot noise reduction. [Figure 9] FIG. 9 is a graph showing the number of detected background hotspot noise counts. [Figure 10] FIG. 10 is a graph showing colocalization rates based on previous frame data and first frame data.
[0009] Throughout the drawings, the same reference numerals and characters, unless otherwise stated, are used to denote like features, elements, components, or portions of the illustrated embodiments. Also, while the present disclosure will be described in detail with reference to the drawings, it will be explained in conjunction with exemplary embodiments which demonstrate. It is intended that changes and modifications can be made in the exemplary embodiments described without departing from the spirit and scope of the present disclosure, as defined by the appended claims. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0010] When applying a plasmonic chip for highly sensitive virus detection using fluorescent dyes, the analysis is bulk detection rather than single molecule analysis, so tiny background hotspot noise does not have a significant impact on the analysis. In the case of single EV analysis, it is important to detect tiny signals that separate each EV, and the difficulty lies in the fact that the size and intensity of these small signals are similar to the background hotspot noise generated in the plasmonic chip. In addition, if the background hotspot noise stabilizes over time, it can be easily identified by measuring the generation position in advance, but the intensity of the noise that is generated and disappears is further modulated, and the generation position is random. When using long exposure to improve the signal to noise ratio (SNR), the blinking of the background is captured as a bright spot and cannot be distinguished from the signal from the EV. Since images are acquired in places where no bright spots are initially generated while bright spots are generated, the pre-obtained background hotspot position cannot be applied to reduce these.
[0011] Thus, systems, devices, and methods are provided that reduce unwanted background hot spot noise from plasmonic tips for fluorescence microscopy that detect fluorescent signals from small amounts of fluorescent dyes on small biological samples, such as extracellular vesicles.
[0012] EV detection using microscopy Due to their small size, EVs cannot be resolved by optical microscopes. To detect them with a microscope, fluorescent dyes are applied to them and the fluorescent light is detected as bright spots to detect the presence or absence of fluorescent signals. Due to their small size, only a small amount of fluorescent dye is applied to EVs, and the fluorescent signal is weak, so signal amplification is required. One method is to use plasmon resonance.
[0013] Background Hotspot Noise Some plasmonic chips using local plasmonic resonance effects have background hotspot noise, which are tiny bright spots that resemble the signal from the sample. In addition, the location of the noise is random, with strong and weak signals, flashing irregularly, or observed only once at a specific location. A pixel set specified by pixels larger than a minimum size, such as 3×3 or 2×2 pixels, and whose intensity exceeds an intensity threshold, is detected as background hotspot noise. In one example of a plasmonic chip made of gold, there is little hotspot noise from the AF555 wavelength band, and there are many hotspot noises in the AF647 and Cy7 wavelength bands. More specifically, when observing the chip in the AF647 wavelength band, approximately 5% of the total noise appears in every frame (in some embodiments, such as 5 seconds of continuous imaging and 100 ms exposure time per frame), and about 16% of the total noise, including the above-mentioned non-vanishing noise, appears only once. Half of the noise flashes less than four times in a 50-frame image, and 90% flashes less than 10 times. The expected number of noise flashes under the above conditions is about 5. The expected maximum noise output time is about 900 ms, excluding noise that appears constantly.
[0014] A shorter exposure time per frame reduces the amount of detected background hotspot noise but lowers the SNR. On the other hand, a longer exposure time per frame improves the SNR but increases the amount of detected noise. Therefore, we examined how the detected background hotspot noise changes with exposure time and number of frames. We examined the percentage of hotspots that flashed at least once during the entire frame time, and the results are shown in Figure 1. Here, the exposure times were set to 100ms, 500ms, 1s, 2s, and 5s. As the exposure time was longer, the amount of detected noise increased, but the percentage identified as hotspot noise was within a range of plus or minus 2% for the same number of frames. If the noise was to be completely removed, several hundred frames would need to be acquired. When determining the exposure time and number of frames, the total acquisition time, residual noise, and SNR must be taken into consideration. A short exposure time is appropriate to reduce the absolute number of hotspot noise, but this must be balanced against the decrease in SNR quality. A low SNR will result in loss of detection of weak signals from the EV. The allowable percentage of residual noise after noise reduction is one way to determine the preferred acquisition condition, and the optimal number of frames can be determined from Figure 1. In other words, the required number of frames can be determined by determining the exposure time of one frame based on the SNR and the acceptable percentage of residual hot spot noise.
[0015] Because of these characteristics, much of the noise cannot be located in advance, but if multiple images are taken, noise from blinking bright spots can be distinguished from signal intensity decay.
[0016] System Overview The imaging system shown in FIG. 2 is an example of an imaging system having the fluorescence imaging system described herein. The imaging system in FIG. 2 includes a fluorescence microscope 101, a control unit 102, an image processing unit 103, a memory unit 104, and a display unit 105. The fluorescence microscope includes a camera 1011, a lens unit 1012, a filter unit 1013, an illumination unit 1014, an objective lens 1015, and a stage 1016. A sample on a plasmonic chip 1017 is placed on the stage. The control unit 102 controls the entire system processing, including image acquisition under the fluorescence microscope control, image processing, image display, and data recording. The image processing unit 103 performs background hotspot noise reduction using multiple images, and outputs the detected sample position and / or the identified background hotspot noise position. The memory unit 104 temporarily records the acquired images and calculated data. The display unit 105 shows the data of the system, such as information on microscope control, live images, acquired images, and calculation results.
[0017] Generalized Processing The intensity features of noise are more variable than those of fluorescence and fall below the detection level faster than the time it takes for fluorescence bleaching to occur. However, unlike fluorescence bleaching, noise may fall below the detection level and then rise above the detection level again. This difference in intensity change features is used to distinguish whether a detected bright spot is a fluorescent signal or bright spot noise. To this end, multiple image frames are acquired within a time period that does not cause fluorescence bleaching, and if the intensity of the bright spot changes beyond a given intensity threshold, the bright spot is identified as noise.
[0018] The plasmonic tips described herein may be any plasmonic surface that has the effect of enhancing the surface plasmon signal. Examples of plasmonic tips include those described in Park et al., (“Self-Assembly of Nanoparticle-Spiked Pillar Arrays for Plasmonic Biosensing,” Adv. Funct. Mater., 2019, Vol. 29, Issue 43, 1904257) or WO 2021 / 211756, which are incorporated herein by reference in their entirety. These plasmonic tips are nanoplasmonic arrays that amplify one or more specific wavelengths of electromagnetic radiation. These plasmonic tips may be gold nanorods printed on a resist layer and coated with a surface layer such as citrate or PEG. Other example plasmonic tips may be formed from, for example, silver or gold island membranes, gold nanospheres, gold weights, etc.
[0019] Use particle detection results for each frame To identify background hot spot noise, the normal exposure time is split into shorter exposure times and multiple images are acquired to distinguish between unstable noise signals and signals from more stable samples.
[0020] The particle detection process in a single channel will now be described according to an exemplary flow chart shown in FIG.
[0021] First, the wavelength band of the fluorescence microscope is selected based on the wavelength that excites the fluorescent dye, and the imaging parameters (for image acquisition) are set according to the selected wavelength band. Since the bleaching time of the fluorescent dye varies depending on the fluorescent dye and the illumination power, the total exposure time (the sum of the exposure times of each frame) and the illumination power are optimized based on the fluorescent dye.
[0022] There are several approaches to determine the exposure time and number of frames. One method is to use the results of the background noise analysis described above, where the exposure time is set to 900 ms, which is the expected maximum noise duration, and the number of frames is set to 5, which is the expected relight of noise. Another method is based on the signal level of the EV. A short exposure time will result in a loss of detection of weak signals from the EV. The exposure time should be set to the optimal exposure time that allows separation of noise while minimizing undetected EVs. To determine the optimal exposure time, images with different exposure times should be acquired with EV samples. The number of detected EVs and background hot spots will increase with exposure time, but the number of detected EVs will saturate after a given exposure time. Thus, the slope of the graph depicting the number of detected changes versus the number of exposure times at which changes occur can be considered as the time at which all EVs are detectable. Here, we assume that the background noise occurs at random locations and increases monotonically with exposure time. For example, the exposure time was about 600 ms at wavelengths between the AF488 and Cy7 channels in the experiments described here. In this case, the SNR was not very high and some of the EV signals were missed. Considering the SNR and the overall acquisition time, an exposure time of 1 to 2 seconds per frame is optimal in these experimental conditions. If the percentage of residual noise is set to 5%, it is preferable to acquire about 40 frames. Similarly, if it is set to 10% or 2%, it will require about 15 or 80 frames, respectively. Here, the absolute number of residual noise depends on the exposure time, so it is preferable to set it to 1 second.
[0023] Also, in some embodiments, it may be preferable to acquire images at different wavelengths rather than continuously acquiring images at the same wavelength, due to the varying likelihood of noise generating locations.
[0024] The total exposure time under a given lighting condition must be shorter than the bleaching time of the fluorescent dye because once the fluorescent dye bleaches, it becomes indistinguishable from background hotspot noise. The bleaching time of each individual fluorescent module is slightly different, and even when the same fluorescent dye is used, the bleaching time has a distribution. The total exposure time must be short enough to avoid the influence of false detection due to differences in bleaching time. For example, the total exposure time should be shorter than the average of the bleaching time minus three standard deviations of the bleaching time distribution. Once the total exposure time is determined, the number of images to be acquired is determined based on the total exposure time and the exposure time of each frame.
[0025] Regarding the number of frames, the more frames you use, the better the noise reduction, so it is better to use the maximum number of frames that the phosphor applied to the EV does not bleach. However, it can be reduced considering the measurement time and effect. For example, if 5% residual noise is acceptable, the total exposure time should be 5s based on the above example.
[0026] Image acquisition is then performed based on the parameters determined in the previous step, where the wavelengths may be sequentially or alternately acquired, with multiple image acquisitions per wavelength.
[0027] Basic Workflow After acquiring multiple temporal images in the same fluorescence channel, background noise reduction is performed. Figure 4 shows the flow chart of background noise reduction.
[0028] A first frame of image is loaded from the memory unit and processed for particle detection, detecting bright spots in the image as particles. Here, the first frame of image includes not only signals from the sample but also signals from background hot spot noise. A base particle map having values of 0 and 1 is created from the detection results and stored in the memory. FIG. 5(a) shows an example of the base particle map, where the dotted circle is the background hot spot noise. More specifically, the map is created with a rectangular area of a predetermined minimum detection size of the particle centered on the centroid pixel of each detected particle, with a value of 1, and a non-particle area with a value of 0. Alternatively, a rectangular area bounded or inscribed by the detected particle is created with a value of 1, and a non-particle area is created with a value of 0.
[0029] After that, a second frame of images is loaded and a second particle map is created by applying particle detection to the second image in the same way as the first image, also with 0 and 1 values. Figure 5(b) shows an example of the second particle map, where the dotted circles represent the disappeared particle areas and the thin dotted circles represent the newly detected particle areas. A so-called colocalization check is performed by comparing the base particle map and the second particle map to check if particles are detected at the same location. A particle that is in the base particle map and at the same location in the second particle map is identified as a particle (e.g. an EV particle) and also as a signal from the sample and not as hot spot noise. Colocalization of particles depends on whether a particle in the base particle map has a corresponding particle in the temporary particle map within the location threshold of the particle. If a particle is in one of the base particle map and the temporary particle map and does not have a corresponding particle in the other particle map within the location threshold of the particle, it is identified as hot spot noise.
[0030] This process is performed for all particles in the base particle map to identify particles that are not considered background hotspot noise. FIG. 5(c) shows an example of an updated base particle map, and FIG. 5(d) shows an example of a noise map, i.e., a map identified as noise. Now, if the particle areas in the two maps overlap by more than a set location threshold, the particle is determined to be co-located across both images. The location threshold should be greater than half of the overlap area. In some embodiments, the location threshold is greater than 50%, 60%, or 70% of the overlap area. The detected particles may be slightly different in size, and if so, the smaller particle is used as the basis for the decision.
[0031] After acquiring multiple images, background noise reduction processing is performed. The flow chart of background noise reduction is shown in Figure 6, and the processing is performed in a loop.
[0032] As before, the first frame of images is loaded from memory and processed for particle detection to create a base particle map.
[0033] After that, the image of the next frame is loaded and an image registration process is performed between the image of the first frame and the image of the current frame. This process can be omitted if the microscope stage is stable during the acquisition of the series of images. A temporary particle map is created by applying particle detection to the current image as in the first image. A so-called colocalization check is performed to check whether particles are detected at the same location in the base particle map and the temporary particle map, and particles that are in the base particle map but not in the temporary particle map at the same location are removed from the base particle map. This process is performed for all particles in the base particle map to generate an updated base particle map. Figure 5(c) shows an example of an updated base particle map, and Figure 5(d) shows an example of a noise map, i.e., a map identified as noise. This update process is performed with additional images and continues until the image of the final frame. Now, if the particle areas in the two maps overlap by more than a set location threshold, the particle is determined to be co-located across both frames. In some embodiments, the location threshold should be greater than half of the overlap area. In some embodiments, the location threshold is greater than 50%, 60%, or 70% of the overlap region. The detected particles may vary slightly in size, and if so, the smaller particle is used as the basis for the decision. If a colocalization check is performed, the process proceeds with the updated base particle map.
[0034] The updated base particle map is stored in the memory as the particle detection result. In addition, the size information of the particles, the intensity of the detected particles, and some statistical information can be stored as available data. Also, the acquired images are stored if necessary, but to reduce memory usage, it is better to store the summed image or the averaged image of all frames.
[0035] After saving the calculation results such as the updated base particle map, the process is completed. In this embodiment, the base particle map is created from the first image, but the summed or averaged image from all frames can be used to create the base particle map. See Figure 7. In this case, the image with a better signal-to-noise ratio will be applied to create the base particle map.
[0036] Uses the difference in grain intensity between adjacent frames Some background hot spot noise does not disappear completely, but instead shows a larger intensity change than the bleaching speed of the fluorescent dye. This kind of noise cannot be removed by the above method. On the other hand, the fluorescent signal that bleaches and disappears quickly may be removed as noise. It is better to judge whether it is noise or not by the difference in the intensity change of the particle between frames.
[0037] Another embodiment of the background hotspot noise reduction method can be applied to the same system, in which case the overall process follows the flow chart shown in FIG. 3, but the background hotspot noise reduction process is shown in FIG.
[0038] First, the first frame of image is loaded from the memory section and processed for particle detection to detect bright spots in the image as particles, and then a particle map is created.
[0039] The image of the next frame is loaded and image registration is applied to align the image of the current frame to the image of the first frame.
[0040] The intensity of each particle detected based on the particle map is extracted from both the image of the current frame and the image of the previous frame. Here, the maximum or average value of each detected particle can be used to determine the intensity. When using the maximum value, the fluctuation will be large due to the influence of optical shot noise, and when using the average value, the fluctuation will be small. Therefore, it is preferable to set an optimal intensity threshold depending on the value used. The intensity difference is calculated and compared with an intensity threshold determined based on the intensity of the bleaching of the fluorescent dye. In order not to exclude fluorescently labeled particles, the intensity threshold should be set to an intensity change larger than the intensity change due to bleaching. In some embodiments, a single intensity threshold is set for processing. In other embodiments, the intensity threshold may be defined for each image (or for each set of images) based on the image baseline intensity or similar background intensity. If the intensity difference is larger than the intensity threshold, the particle is removed from the particle map. The intensity variation check is repeated for all detected particles in the particle map, and then for all frame images. Finally, an updated particle map is output.
[0041] Both processes of background hotspot noise reduction can be applied simultaneously.
[0042] The one-time method is well suited for detecting a number of EVs, i.e. single channel detection. However, the acquisition and processing times may be longer than desired. In this way, processing may be implemented that reduces the time required for analysis.
[0043] Exemplary embodiments In this embodiment, as shown in FIG. 7, an image for particle detection and an image for identifying background hotspot noise are acquired separately. First, image acquisition parameters are set based on the wavelength used, as described above. Second, an image for particle detection is acquired with a longer exposure time than an image for identifying background hotspot noise. This increases the signal-to-noise ratio and allows weak signals to be detected. Third, an image for identifying background hotspot noise is acquired as described above. Then, particle detection is applied to the image for particle detection to create a particle map(s) showing particle positions. A process for creating a background hotspot noise map similar to the background reduction process described above and shown in FIG. 4 is applied. This process maintains and updates the positions of the identified background hotspot noise to create a background hotspot noise map instead of an updated particle map. Finally, the process ends by removing the identified background noise from the detected particle map, updating the particle map, and saving various data. The background reduction process disclosed in any of the above embodiments is applied to this process of the background hotspot noise map.
[0044] The examples described here were performed using a Zeiss Axio Imager M2 with a Zeiss Colibri-7 LED light source and a Zeiss Plan Apochromat 40x / 0.95 objective. Additionally, Chroma 39002, 39004, and 39007 fluorescence filter sets were used for the AF488, AF555, and AF647 channels. Zeiss multiband filters or Chroma fluorescence filters were used for the Cy7 channel. The plasmonic chip(s) were placed on the microscope stage, and 50 images were acquired with an exposure time of 500 ms. In this experiment, images were acquired sequentially for each channel.
[0045] effect The effect of the background hotspot reduction method disclosed in the present invention is shown in FIG. 9. The background hotspot noise is analyzed using four different wavelength bands corresponding to the wavelength bands of the fluorescent dyes AF488, AF555, AF647, and Cy7. Original BG in FIG. 9 means the number of background hotspot noise in the first frame. Method 1 disclosed as a general method, Method 2 disclosed as an embodiment using particle detection results in each frame, and a mixed method of Methods 1 and 2 were applied. FIG. 9 shows the number of background hotspot noise identified. In these embodiments, noise was rarely detected in the AF488 and AF555 channels, and most noise was in the AF647 and Cy7 channels. Therefore, we focus on the AF647 and Cy7 channels to check the effect of the disclosed method. Overall, Method 1 is effective in noise reduction, but the mixed method can reduce more noise, which means that Method 2 can reduce noise with different properties. In Method 2, the mean value of the detected particle area was used, and those that differed from the previous frame by more than 20% were identified as noise. Using an intensity threshold at or below the maximum value in the region, method 2 may detect noise with higher sensitivity than method 1. Method 2 does not have a lower limit and may not detect noise below the detection level due to signal decay. However, fluorescent signals from EVs that gradually fade may remain. In the AF647 channel, in this experiment, method 1 can reduce noise by 79.5%, and the mixed method can reduce noise by 84.8%. Similarly, in the Cy7 channel, method 1 can reduce noise by about 90.1%, and the mixed method can reduce noise by 92.4%.
[0046] The amount of noise reduction will depend on the exposure time and excitation power. In the above embodiment, it is preferable to set the acquisition time per frame to 600 ms or more based on the detection limit of the desired signal. For more stable signals, it is preferable to set the acquisition time per frame to 900 ms or more. Considering the absolute number of hot spot noises detected, it is preferable to keep the acquisition time per frame shorter than 2 s. The number of shots or images is determined by specifying the percentage of noise that remains as noise without being identified. When the residual noise is set to a certain percentage (e.g., 10%, 5%, 2%), the number of images acquired can be changed based on this (e.g., 10 to 20 frames, 30 to 50 frames, 70 to 90 frames, respectively).
[0047] The absolute number of detected EVs is important for the analysis of single EVs from normal cells and abnormal cells such as cancer. When the plasmonic chip is used for single EV analysis, the background hotspot noise is considered as a signal from single EVs, resulting in inaccurate analysis results that do not remove the hotspots. This error is particularly serious when the EV concentration is low, since the influence of the background is relatively larger than when the EV concentration is high. By applying the background reduction method(s) of the present disclosure, the error is suppressed even when the EV concentration is low.
[0048] In some embodiments, to reduce the computational load, it is desirable to reduce the number of images taken, taking into account the expected maximum noise duration and the expected number of re-lights of the hotspot noise, with images taken at intervals that are at least the product of the expected maximum noise duration and the expected number of re-lights of the hotspot noise, so that the number of noises occurring at overlapping positions can be reduced and effective reduction can be achieved with fewer images.When multiple fluorescent channels are used, it is appropriate to take images of different channels, one frame at a time, and repeat this process to increase the interval between images taken in the same channel.
[0049] Another way to reduce the computational cost in Method 1 is to not perform particle detection analysis for each frame. Using the particle position information in the base particle map and a location threshold for particle detection, we compare the image intensity at each particle position in each frame with the location threshold, and if it is below the location threshold, we predict the particle from the base particle map. This method can achieve the same results as the original method.
[0050] Colocalization rates comparing each embodiment are shown in Figure 10. In this figure, data based on the previous frame and data based on the first frame data are shown for the AF647 channel image at 100ms and 100 frames. The top trace is based on the previous frame data, with about 70% overlap with the previous frame. The bottom trace is based on the first frame data, with about 25% overlap with the first frame after 10 seconds.
[0051] definition Referring to the specification, specific details are set forth above to provide a thorough understanding of the disclosed examples. In other instances, well-known methods, procedures, components, and circuits have not been described in detail so as not to unnecessarily lengthen the present disclosure.
[0052] The term "about" or "approximately" as used herein means, for example, within 10%, within 5%, or less. In some embodiments, the term "about" may mean within measurement error. In this regard, in the specification and claims, all numbers may be read as if preceded by the word "about" or "approximately", even if such term is not expressly present. The phrase "about" or "approximately" may be used to describe a magnitude and / or location that indicates that the stated value and / or location is within a reasonably expected range of values and / or locations. For example, a numerical value may take a value (or range of values) of + / -0.1% of the stated value, + / -1% of the stated value (or range of values), + / -2% of the stated value (or range of values), + / -5% of the stated value (or range of values), + / -10% of the stated value (or range of values), etc. Any numerical ranges described herein are intended to include the endpoints unless specifically stated otherwise, and include all subranges incorporated herein. As used herein, the term "substantially" means allowing for deviations from the descriptor that do not negatively affect the intended purpose. For example, deviations within measurement limits, manufacturing tolerances, or variations of less than 5% can be considered to be within the range of substantially the same. The specified descriptors may be in absolute terms (e.g., substantially spherical, substantially perpendicular, substantially concentric, etc.) or in relative terms (e.g., substantially similar, substantially identical, etc.).
[0053] Unless otherwise noted, and as will be apparent from the disclosure below, discussions throughout this disclosure using terms such as "processing," "computing," "calculating," "determining," "displaying," and the like, are understood to refer to acts and processes in a computer system, similar electronic computing device, or data processing device that manipulates data represented as physical (electronic) quantities in the registers and memory of the computer system to convert it into other data similarly represented as physical quantities in the computer system memory or registers or other information storage, transmission, or display. The computer or electronic operations described herein or recited in the appended claims may generally be performed in any order unless the context dictates otherwise. Also, although various operational flow diagrams are shown in a sequential order, the various operations may be performed in an order other than that shown or claimed, or operations may be performed simultaneously. Examples of such alternative orders may include overlapping, interleaved, interrupted, recorded, incremental, preliminary, supplementary, simultaneous, reversed, or other different orders unless the context dictates otherwise. Moreover, words such as "responsive to," "in response to," "related to," "based on," or other past tense adjectives are not generally intended to exclude such variants unless the context dictates otherwise.
[0054] Terms such as first, second, third, etc. may be used herein to describe various elements, components, regions, parts, and / or sections. It should be understood that these elements, components, regions, parts, and / or sections should not be limited by these terms. These terms are used only to distinguish one element, component, region, part, or section from another region, part, or section. Thus, a first element, component, region, part, or section discussed below could be referred to as a second element, component, region, part, or section without departing from the teachings herein.
[0055] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. As used herein, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly dictates otherwise. The terms "includes" and / or "including," when used herein, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof not expressly stated.
[0056] In describing the exemplary embodiments illustrated in the drawings, specific terminology has been used for the sake of clarity, however, the disclosure of this patent specification is not intended to be limited to the specific terminology so selected, and each specific element should be understood to include all technical equivalents which function in a similar manner.
[0057] Although the present disclosure has been described with reference to exemplary embodiments, it is to be understood that the present disclosure is not limited to the disclosed exemplary embodiments, and the scope of the following claims is to be accorded the broadest interpretation encompassing all such modifications and equivalent structures and functions.
Claims
1. a fluorescence microscope for imaging the sample on the plasmonic substrate; an image acquisition controller configured to acquire a first image at a first time and at least one second image at a second time; an image processor for distinguishing signals in the at least one first image into (1) signals from the sample or (2) signals from background hotspots; The signal distinction is as follows: performing a particle detection analysis on the first image to generate a base particle map; performing a particle detection analysis on the at least one second image to generate a temporary particle map; defining a particle in the base particle map as a background hotspot if the particle does not have a corresponding particle in the temporary particle map at a location within a location threshold of the particle; defining a particle in the base particle map as a signal from the sample if the particle has a corresponding particle location in the temporary particle map that is within a location threshold of the particle. Optical imaging system for plasmon-enhanced optical imaging.
2. performing particle detection analysis on the at least one second image only at locations on the at least one second image that correspond to locations on the base particle map having detected particles; The optical imaging system of claim 1 .
3. the background hotspots are subtracted from the base particle map to generate an updated particle map; The optical imaging system of claim 1 .
4. The updated particle map is further updated by an iterative process; The iterative process includes: performing particle detection analysis on another image of the at least one second image to generate a second (or subsequent) temporary particle map; defining, for a particle in the updated base particle map, a particle as a background hotspot if there is no particle in the second (or subsequent) temporary particle map that corresponds to a location within a location threshold of the particle; The optical imaging system of claim 3 .
5. the location threshold requires that the particle's area overlaps more than half of the particle maps; The optical imaging system of claim 1 .
6. The particles detected in the base particle map and defined as signals from the sample are extracellular vesicles (EVs) or EVs with one or more fluorescent dyes attached. The optical imaging system of claim 1 .
7. the image acquisition control unit sets an acquisition time for the first image and the second image, the acquisition time being set shorter than a bleaching time of the fluorescent dye in the sample; The optical imaging system of claim 1 .
8. the acquisition time is between 100 ms and 5 s, and at least two images are acquired at different times in the first channel; The optical imaging system of claim 1 .
9. The signal distinction is as follows: registering the at least one second image with the first image; checking colocalization between said base particle map and said temporary particle map; updating the base particle map; Further comprising: The optical imaging system of claim 1 .
10. The image processing unit further includes: subtracting signals from the background hotspots from the at least one second image; combining the at least one second image to form a combined image; sending said combined image to a display; The optical imaging system of claim 1 .
11. the fluorescence microscope is a multi-channel fluorescence microscope; the image acquisition control unit acquires a plurality of images at different times in a first channel and a second channel, The image processor distinguishes signals in one or more images among the plurality of images in each of the first channel and the second channel. The optical imaging system of claim 1 .
12. The image acquisition control unit further acquires an image having a high signal-to-noise ratio (SNR) for data analysis; The at least one second image acquired at a different time is acquired separately for background identification. The optical imaging system of claim 1 .
13. the high SNR image is acquired prior to the at least one second image for background identification; The acquisition time of the at least one second image for background identification is 10 ms to 1 s. The optical imaging system of claim 12.
14. acquiring a plurality of images of a sample on a plasmonic substrate using a fluorescence microscope, the plurality of images being acquired at different times in a first channel; distinguishing signals in one or more of the plurality of images into (1) signals from the sample or (2) signals from background hotspots; The signal distinction is as follows: performing a particle detection analysis on a first image of the plurality of images to generate a base particle map; performing a particle detection analysis on a second image of the plurality of images to generate a temporary particle map; defining a particle in the base particle map as a signal from the sample if the particle has a corresponding particle location in the temporary particle map that is within a location threshold of the particle. Analysis method.
15. and setting an acquisition time of the plurality of images shorter than a bleaching time of the fluorescent dye in the sample. The analytical method according to claim 14.
16. the acquisition time is between 100 ms and 900 ms, and at least two images are acquired at different times in the first channel; The analytical method according to claim 14.
17. Distinguishing signals within one of the plurality of images registering the second image from the plurality of images with the base particle map; checking colocalization between said base particle map and said temporary particle map; updating the base particle map; Further comprising: The analytical method according to claim 14.
18. the background hotspots are subtracted from the base particle map to generate an updated particle map; The analytical method according to claim 16.
19. displaying the updated particle map. The analytical method according to claim 18.
20. calculating a number of detected particles on the updated particle map. The analytical method according to claim 18.
21. acquiring a plurality of images from each of at least two channels; distinguishing signals in one or more images of the plurality of images from each of the at least two channels; Further comprising: The analytical method according to claim 16.
22. performing particle detection analysis on the second image of the plurality of images only at locations on the second image that correspond to locations on the base particle map at which particles are detected; The analytical method according to claim 16.
23. The updated particle map is further updated by an iterative process; The iterative process includes: performing particle detection analysis on another image of the plurality of images to generate an additional temporary particle map; defining a particle in the updated base particle map as a background hotspot if there is no particle in the additional temporary particle map that corresponds to a location within a location threshold of the particle; The analytical method according to claim 18.
