JP2010536029A - Apparatus and method for digital magnetic bead analysis - Google Patents

Abstract

  A light transmission assay bead or digital magnetic microbead having a digitally encoded structure that is partially transparent to light and opaque. If hundreds or thousands of LITAB are submerged in the bottom of the microwell in the microplate, the barcode can be accurately and reliably decoded by image processing. Microplates are a common bioassay format; each plate may have 96, 384, or 1536 patient samples. Thus, multiple targets in one sample can be analyzed in a single microwell. The image decoding algorithm includes four main steps: (1) image enhancement; (2) bead segmentation: (3) barcode slit extraction: and (4) barcode decoding. The bead image is obtained from the bottom of an optically transparent microplate and the barcode pattern can be decoded by image software. Thus, the entire bead bioassay experiment can be performed in a microplate without removing the beads.

Description

Related applications

  This application is a continuation-in-part of US patent application Ser. No. 11 / 580,514 filed on Oct. 13, 2006; in US provisional patent application No. 60 / 706,896 filed Aug. 9, 2005. US patent application Ser. No. 11 / 502,606, filed Aug. 9, 2006 claiming priority; US patent application No. 12 / 069,720 filed Feb. 11, 2008 A continuation-in-part of U.S. Provisional Patent Application No. 60 / 964,108, filed August 8, 2007. All other publications and US patent applications disclosed below are also incorporated by reference as if fully set forth herein.

  The present invention relates to the performance of multiplexed bioassays with hundreds or thousands of digital magnetic barcode microbeads for protein, nucleic acid and molecular diagnostics; more particularly, one in a small microwell such as a microplate. Optical image decoding algorithms and methods are developed for the rapid and simultaneous analysis of individual bead barcodes. Digital magnetic beads are not spherical and are not traditional latex microspheres; therefore, high density optical patterns can be accurately imaged and identified.

  Current research in genomics and proteomics requires multiplex data, and techniques are needed that can quickly screen large quantities of targets such as nucleic acids and proteins in very small samples. Microarrays, DNA chips, and protein chips that can screen thousands or millions of targets on a planar platform require large amounts of sample to cover a large surface. The standard surface area is 1 cm × 1 cm or on a microslide. Distributing a small amount of liquid sample over a relatively large chip surface often results in the disadvantage of slow diffusion of molecules and uneven mixing or distribution on the chip surface. These are the reasons why microarray assays require very long reaction times. Moreover, once the microarray chip is printed and assembled, it is impossible to add another test to the multiplex assay.

  Microbead technology overcomes many of the problems of microarray technology and can provide flexibility in library content and bead quantity or bead type in the analysis. With its small volume (within picoliters per bead), thousands of beads can be incubated with very small samples. Many encoding methods have been shown to include particles having metal rods with spectrally distinguishable fluorophores, fluorescent semiconductor quantum dots, and bar-coded color (absorbing) or black and white strips. Both fluorescent and barcode colored strip beads are identified by optical detection in a reflective or emissive configuration. Problems with the reflective configuration are: (1) low light reflection yield, especially when the beads are on the micrometer scale, (2) poor light collection efficiency, and (3) significant fluorescence bands for fluorescent based coded beads It is wide or overlapping, limiting potential code numbers. Another disadvantage of fluorescence-based beads is that most bead-based assays rely on fluorescence readouts and therefore produce more fluorescence spectrum or intensity interference. In the case of multi-metal (Au, Pt, Ni, Ag, etc.) color microrods, the encoding is difficult to manufacture and the number of colors is limited based on different metal materials.

  U.S. Pat. No. 6,057,028, issued August 10, 2004, the entire contents of which are incorporated herein by reference, is a form of bar coding with a diameter of 30-300 nm by a multi-metal rod of 400-4000 nm. Is disclosed. These rods are made by electrodeposition onto an alumina mold; the alumina is then removed leaving these small multilayer objects. This system can have up to 12 coded zones in up to 7 different metals, and the metals have different reflectivities, and therefore brighter or darker in an optical microscope depending on the type of metal Visible, while assay readout is due to fluorescence from the target and probe identification is from the light-dark pattern of the barcode.

  Patent document 2 issued on October 7, 2003, the entire content of which is incorporated herein by reference, discloses semiconductor nanocrystals that act as barcodes, each semiconductor nanocrystal having a distinct emission spectrum. Produce. These characteristic emissions can be observed as colors in the visible region of the spectrum, or they can be decoded to provide information about the specific wavelengths at which discontinuous transitions are observed.

  U.S. Pat. No. 6,057,033, issued May 11, 2004, the entire contents of which are incorporated herein by reference, describes a plurality of labels that combine with a label that includes a marker to produce a wavelength / intensity spectrum in response to excitation energy. An identification system is disclosed that includes an identifiable element and an analyzer that identifies the element from the wavelength / intensity spectrum of the label to be bound.

  In Patent Document 4 issued on February 26, 2002, a bead is placed between a nickel plate and a second plate on which a bar code is electroformed, and the bar code is pressed against the surface of the bead. Disclosed is a method for making a microcarrier by forming microcake-like particles having a barcode.

  U.S. Patent No. 6,057,036, the entire contents of which are incorporated herein by reference, uses color signals based on optical graphics and color information signals, such as barcodes or dot matrix barcodes, to produce affinity reaction probe beads. An identification system is disclosed.

  US Pat. No. 6,057,032, the entire contents of which are incorporated herein by reference, describes a small plane designed to attach a single adherent cell, a cell attachment region designed to protect the cell, and a flow-through. Disclosed are micropallets containing shapes designed to enable or improve the process. The micropalette is preferably sized to accommodate a single cell that is patterned in an easily identifiable manner and is comparable in size.

  Magnetic beads are widely used in high speed, high volume processing automation processes. Magnetic beads are paramagnetic, that is, have magnetic properties when placed in a magnetic field, but do not retain magnetism when removed from the magnetic field. This allows the magnetic collection of the microbeads and resuspension of the beads when the magnetic field is removed. Digital magnetic bead collection and resuspension can be easily and quickly repeated any number of times. In a typical robot automation, a 96-well, 384-well, or 1536-well microplate is placed on a magnetic stand equipped with magnetic pins, and a magnetic field is activated. This allows unbound molecules to be washed from the beads, changing the buffer solution and removing any contaminants in the solution. For example, in the case of a DNA or RNA assay, unbound or non-specific nucleotides can be removed after hybridization. For protein assays, unbound or non-specific antibodies or antigens can be removed after the antibody-antigen reaction. During molecular biology applications, it is often required to perform large scale washings quickly, efficiently, and with minimal difficulty. Magnetic beads are widely used in bioassays, but magnetic beads with high density barcodes are not available.

US Pat. No. 6,773,886 US Pat. No. 6,630,307 US Pat. No. 6,734,420 US Pat. No. 6,350,620 US Patent Application Publication No. 2005/0003556 US Patent Application Publication No. 2005/0244955

  There is a need for digitally encoded magnetic microbeads that provide high optical contrast and high signal-to-noise for reliable decoding and provide magnetic properties for high-speed, high-throughput automated cleaning in microplate formats . The decryption method needs to be simple and robust. Data processing needs to be fast and accurate.

Means for Solving the Invention

  The present invention improves upon prior art systems and parent applications and provides a fast mass processing process for identifying codes shown by simultaneous imaging of multiple digitally encoded beads. In accordance with the present invention, the microbead analysis system obtains an image of a support (eg, well plate or microslide) that supports at least one bead marked with a digital code; An imaging device; and a decoding system that analyzes the images obtained by said imaging device, recognizes and separates the beads from each other and the support in the background, and identifies the digital code indicated by the indicia placed on the beads including. The image processing method of the present invention differs from conventional scanning of barcodes on a pallet in that the pallet is not imaged with a background support, and thus the pallet is not recognized and separated from any background or any other image. . The present invention is particularly useful for efficient analysis of a plurality of microbeads, each having a mark representing a digital code, and the image obtained by the imaging device includes a plurality of beads including a mark on each bead The system is configured to analyze the image to recognize and separate the plurality of beads and identify the digital code indicated by the indicia placed on each bead. By recognizing and separating each bead from each other and from the background in all images of the plurality of beads, the digital code shown on each bead can be identified.

  In some embodiments, the present invention is not kinetics that the beads move with respect to the imaging optics (eg, beads moving in the flow pass through the imaging optics or the beads are scanned by the moving light beam), but the binding of the beads. The present invention relates to an image processing method for decoding digitally encoded beads used in bioassays that perform images in a steady state. Multiple beads may be distributed on a plane (eg, glass microslide) or may be imaged simultaneously in two dimensions by an imaging device (eg, wide field image camera), thereby decoding the multiple beads. As a result, the detection result is improved. Digital encoding may be observed by imaging based on light emission, reflection and / or transmission with respect to the beads.

  In another aspect, the bead imaging aspect of the present invention is such that the beads are moving with respect to the imaging optics as long as the bead imaging system can sharply capture a steady state image of the beads while the beads are moving. Applicable to beads of kinetics (eg, beads moving in a flow pass through binding optics or beads are scanned by a moving light beam).

  In another aspect of the invention, the bead digital encoding comprises a bar coat pattern. A barcode pattern with a series of narrow and wide bands provides a clear signal and 0 and 1 differentiation. The position of the slit on the pallet determines which bit is the least significant bit (LSB) and the most significant bit (MSB). The LSB is located closer to the edge of the pallet and distinguishes it from the other longer end MSB. In some embodiments, the coded beads are a series of alternating light and dark sections having a relative width (eg, a narrow section showing a series of “0” codes and a wide section showing “1” codes, or vice versa). A body having a barcode image (eg, light transmissive and opaque sections). The digital barcode pattern is imaged by an imaging device.

  The image decoding method includes four main sub-processes: (1) image enhancement, (2) bead segmentation, (3) barcode slit extraction, and (4) barcode decoding. In one embodiment, decoding the bead image includes a series of steps taken to analyze the bead image and extract digitally encoded information. Such steps may include: (1) Grayscale transformation to preserve bead image details; (2) Background removal to obtain a uniform background for the bead image; (3) Beads Image edge detection; (4) If the bead images are in contact, an additional filtering step to separate the contacting bead images; (5) Labeling individual bead images (eg using color code or number) ); (6) Display of digital encoding (eg barcode pattern) in each bead image by alignment transformation and intensity (eg intensity plot) of the bead image; and (7) Analysis of intensity to identify the digital code (eg Identification of narrow and wide bands in the barcode pattern).

  The accuracy in decoding the image of the stationary bead can be better than the decoding of the image of the moving bead. Because incorrect identification can lead to misdiagnosis and mistreatment, the accuracy of decoding is very important for clinical diagnosis.

  In the illustrated embodiment, the present invention is digitally encoded as shown by the image providing light contrast assay beads (LITAB) or high contrast and high signal to noise optical detection that facilitates identification of the beads. It relates to the micro palette. The image is partly substantially transparent to light (eg, transparent, translucent, and / or transparent to light) and partly opaque (eg, to light) It is implemented by a physical structure having a pattern (reflecting and / or absorbing). The pattern of transmitted light is identified (eg, by scanning or imaging) and the code indicated by the image on the coded beads can be decoded. In some embodiments, the coded beads are 1D or 2D barcode images (eg, a narrow slit (eg, 5 micrometers wide) showing a series of “0” codes) and a wide slit (eg, 10 micron) showing “1” codes. A body having a series of alternating light transmissive and opaque sections having a relative position, width and spacing similar to metric width), or vice versa). To decode the image, alternating transmissive and opaque sections of the body are scanned and imaged (eg, by a CCD sensor) to identify the code indicated by the image identified from the transmitted light.

  In a further embodiment of the invention, a large number of digital magnetic beads are distributed on a plane and can be detected simultaneously by an image camera. The plane may be on the microslide or the bottom of the microplate. Bar code patterns indicating digital signals such as “0” and “1” for each bead can be identified by image processing.

  In a further aspect of the invention, the digital barcode beads have paramagnetism, i.e., have magnetic properties when placed in a magnetic field, but do not retain magnetism when removed from the magnetic field. Magnetic beads can be washed in a microplate by collecting the beads with an external magnet and resuspending the beads when the magnetic field is removed. Multiple digital magnetic beads allow multiplex assays to be performed in a single well. Microplates are a standard high-speed, high-throughput format in clinical diagnosis where each plate has 96, 384 or 1536 patient samples.

  In a further embodiment of the present invention, the digital magnetic microbead analysis system comprises: (a) a slide or microplate having a plurality of wells; (b) at least one sinking on the surface of the slide or on the bottom of the well of the microplate. One digital magnetic bead; (c) an optical detector located above or below the slide or microplate and imaging at least one magnetic microbead; and (d) processing an image pattern of at least one magnetic microbead Digital processing system performed by image software. In certain embodiments, the number of wells is about 96, 384 or 1536 wells.

  In a further aspect of the invention, both the barcode image and the fluorescence image are taken simultaneously under the microscope and camera. Therefore, all bead experiments can be performed in the microplate without removing the beads. Bar codes are used to identify which molecular probes are immobilized on the beads, while fluorescence is used to detect positive or negative reactions. Multiple targets can be analyzed simultaneously.

  In a further aspect of the invention, the digital magnetic microbead includes a first layer; a second layer; and an intermediate layer between the first layer and the second layer, the intermediate layer having a coding pattern defined thereon. And is partially transparent to light and partially opaque, indicating a code corresponding to each microbead.

  In a further aspect of the invention, the intermediate layer has a series of alternating substantially light transmissive sections and substantially light opaque sections that define a coding pattern. The relative position, width and / or spacing between the transmissive section and / or the opaque section indicates a binary code. The substantially light opaque section includes a light blocking material. The body of each microbead has the longest orthogonal axis of 1 mm or less.

  In a further aspect of the invention, the first and second layers of digital magnetic beads are functionalized with a material selected from the group consisting of proteins, nucleic acids, small molecules, chemicals, and combinations thereof.

  In certain embodiments, the body of the coded bead may be configured to have at least two orthogonal cross sections that differ in relative geometry and / or size. Further, the cross-sectional geometry may be symmetric or asymmetric and / or may be regular or irregular. In certain embodiments, the longest orthogonal axis of the encoded bead is less than 1 mm.

  In a further aspect of the invention, the light transmissive section is defined by a slit through the intermediate layer and the light opaque section is defined by a light reflecting material and / or a light absorbing material. The slit has a first width slit and a second width slit. In the binary code, the first width indicates “0” and the second width indicates “1”. The first width is about 1 to 10 micrometers, the second width is about 1 to 50 micrometers, and the first width is narrower than the second width. The binary code can be decoded by image software.

  For a more complete understanding of the scope and nature of the present invention, as well as the preferred mode of use, reference should be made to the following detailed description taken in conjunction with the accompanying drawings. In the following drawings, like reference numerals designate like or similar parts throughout the drawings.

FIG. 4 illustrates a method for preparing light transmission assay beads (LITAB) for bioassay according to an embodiment of the invention: multiple LITABs in microwells of a microplate Figure 3 illustrates a method for preparing light transmission assay beads (LITAB) for bioassay according to an embodiment of the invention: LITAB for bioassay Figure 2 illustrates a method for preparing light transmission assay beads (LITAB) for bioassay, according to one embodiment of the invention: Schematic showing a photographic image of LITAB Plan view of a LITAB according to an embodiment of the invention Sectional view along line AA in FIG. Shows the transmitted digital signal of a barcode bead showing 0011111001 on an image camera Shows an optical signal pattern showing light transmitted through a slit pattern in LITAB with least significant bit (LSB) or most significant bit (MSB) FIG. 4 illustrates a process for forming beads according to an embodiment of the invention. Shows a metal layer as a layer sandwiched between two polymer layers that can provide the same surface chemistry for molecular fixation Shows microscopic image of barcode microbead Demonstrate the use of edge detection techniques to delineate objects in an image Shows isolated beads separated using Watershed line Shows how to align each microbead to the main axis, and grayscale image is transformed into a matrix containing pixel values Indicates that the bead image is converted to grayscale intensity as a function of pixel value Shows the digital code identified based on the width of the barcode present on each microbead 1 illustrates a LITAB analysis system according to an embodiment of the present invention.

  This description is of the best presently contemplated mode of implementation of the invention. This description is made for the purpose of illustrating the principles of the invention and should not be taken in a limiting sense. The scope of the invention is best determined by reference to the appended claims.

  The detailed description of the method of the invention will be described primarily with respect to the method or process, the symbolic description of the operation, the functionality and features of the invention. The descriptions and explanations of these methods are the methods used by those skilled in the art to most effectively convey the work to others skilled in the art. The method or process performed by the software here is usually a consistent series of steps leading to the desired result. These steps require physical manipulation of physical quantities. Often, but not necessarily, these quantities take the form of electrical or magnetic signals that can be stored, transferred, combined, compared, and otherwise manipulated.

  Devices useful for performing tasks performed by the software of the present invention include general or special purpose digital processing and / or computing devices that may be stand-alone devices or part of a larger system. However, it is not limited to this. The device may be selectively activated or reset by a program, routine and / or sequence of instructions and / or logic stored in the device. That is, use of the methods described and suggested herein is not limited to a particular processing configuration.

  For purposes of illustrating the principles of the present invention, and not by way of limitation, the present invention will now be described by reference to bioanalysis with reference to microbeads in the form of a pallet. However, the present invention is equally applicable to all other geometries without departing from the scope and nature of the present invention and may be applied to other applications requiring identification based on bead identification. Understood. To facilitate the following discussion, the microbeads of the present invention are referred to as LITAB, meaning light transmission assay beads.

  LITAB is digitally encoded as shown by images that provide high contrast and high signal-to-noise light detection to facilitate bead identification. The image is partly substantially transmissive to light (eg, transparent, translucent, and / or transmissive to light) and partly opaque (eg, reflective to light) (And / or absorbable) patterns. The transmitted light pattern is identified (eg, by scanning or imaging) and the code indicated by the image on the coded bead can be decoded. Various barcode patterns such as circles, squares, or other geometric shapes can be designed as long as they indicate "1" or "0" and can be recognized by the decoder. However, LITAB is not spherical; unlike conventional latex-based spherical beads.

  In some embodiments, the coded beads are 1D or 2D barcode images (eg, narrow slits (eg, about 1 to 5 micrometers wide) that show a series of “0” codes to form a binary code) and A series of alternating light transmissive and opaque sections with relative positions, widths and spacings resembling a wide slit indicating a “1” code (eg, about 1 to 10 micrometers wide, or vice versa) Including a main body. FIG. 2 shows LITAB 11 which is a coded bead according to an embodiment of the present invention. The LITAB 11 has a body 25 in the form of a flat pallet or disc. The body of the coded bead may be configured to have at least two orthogonal cross sections that differ in relative geometric properties and / or size. Further, the cross-sectional geometry may be symmetric or asymmetric and / or may be regular or irregular. In this particular embodiment, all three orthogonal axes are of different lengths, and the geometric properties of all three orthogonal cross sections are symmetrical and regular. FIG. 2 (a) shows that the flat geometry resembles a symmetrical stretched or elongated elliptical pallet. FIG. 2B shows a cross section showing the vertical axis (or the longest axis). A series of wide and narrow slits 23 and 24 are provided across the body 25 and may be made of or coated with a substantially light opaque material (eg reflective or absorptive). Wide and narrow slits 23 and 24 indicate logic “1” and “0”, respectively, or vice versa, and together indicate a binary code (each slit represents a bit). In this embodiment, the code is similar to a bar code. The narrow slit 24 may have a width of 5 micrometers and the wide slit 23 may have a width of 10 micrometers. For LITAB having an overall dimension of 100 × 30 × 10 μm to 300 μm × 100 μm × 40 μm, at least about 10 slits may be provided on the LITAB palette, encoding 6 to 12 bits or more, 64 to 4096 In addition, more unique codes are possible. In certain embodiments, the longest orthogonal axis of the encoded bead is less than 1 mm.

  The illustrated embodiment shows narrow and wide spaced pattern of slits, but with narrow and wide spacing between adjacent slits representing 1 and 0 without departing from the scope and nature of the present invention. It is also possible to use a pattern of slits with a certain fixed width. FIG. 2 (c) shows the transmission peak of a single bead detected by a CCD (Electro-Coupled Device) and displayed on a computer screen. When the beads are illuminated with a light beam, the digital barcode 0 or 1 can be identified by the imaging camera and digital signal processor based on the “total intensity” of the transmission peak from the slit or the “bandwidth” of the transmission peak . As shown in the drawing, the barcode pattern can be easily identified based on the peak width. As specifically shown in the embodiment shown in FIG. 2 (c), the beads show a 10-bit barcode indicating 0011111001.

  To decipher the image, alternating transmissive and opaque sections of the body are imaged with light (eg, using a CCD sensor) and the code indicated by the image identified from the transmitted light is identified. For illustrative purposes, FIG. 3 shows a series of signal pulses showing the detection of light transmitted through slits 23 and 24 in LITAB 11 in FIG. 2 (a). The signal corresponds to the contrast of transmitted versus blocked light through the vertical axis of LITAB11. The width of each signal pulse indicates “1” or “0” in the code of LITAB11. In certain illustrated embodiments, a wider width indicates 1 and a narrower width indicates 0. The relative position of the slit on LITAB 11 specifies which bit is the least significant bit (LSB) or the most significant bit (MSB). In one embodiment, the least significant bit is located closer to one edge or end of LITAB 11 and distinguished from the most significant bit at the opposite edge or end. The concept of decoding signal pulses is similar to the decoding of conventional bar codes.

  In another embodiment, the substantially transmissive section does not necessarily have to be a slit throughout the entire thickness of the LITAB body. The slit may be completely or partially filled with a transparent or translucent material in the substantially liquid, yet still provide substantially light transmission compared to the opaque section. For example, the LITAB may have a transparent body that is coated with a light blocking material (eg, a reflective material, or a light reflective or absorbing die) having an opening that defines a slit that exposes the transparent body. The light imaged on this LITAB is transmitted through the body in the section not covered by the blocking material (i.e. the slit) and blocks the light in the coated section. In another embodiment, the encoded LITAB may comprise a reflective thin film or coating to improve the contrast between transmitted light versus blocked / reflected light and the optical efficiency of image recognition and decoding (eg, , Plating or coating the surface of LITAB with a metal film, or providing a sandwich layer in the middle of a paramagnetic metal film, or coating with a light absorbing die).

  LITAB may be made using conventional methods used for thin film formation in clean room microfabrication equipment. The art for conventional photolithography, printing, silk screening, curing, development, etching (eg, chemical etching, ion etching, and / or other removal processes), plating, dicing, and such types of structures and materials A method involving other processing steps well known in U.S. may be used to obtain the structure of LITAB. Referring to FIGS. 4 (a)-(d), in one embodiment of a method for making LITAB, a layer of Ti 52 (eg, 100 nm) is placed on a substrate 50, eg, a clean glass slide (eg, about 1 mm thick). Deposited by e-beam evaporation. Ti acts not only as a surrogate releasing layer but also as a conducting seed layer. The LITAB body 25 may be formed using a layer of polymeric material. For example, a photoresist photosensitive resin (eg, SU-8 as known in the art) may be used to make LITAB. A layer 21 of polymeric material is spin coated on the Ti layer 52 and slits 23 and 24 are formed in such layer using standard photolithography methods. For example, the slits 23 and 24 may be defined by UV light using a photomask (not shown) that defines the desired pattern of wide and narrow slits and the flat shape of the LITAB body 25. An array of LITABs may be formed on a single substrate having different slit patterns, each representing a different code. The photomask may also define the perimeter of the array of LITAB bodies, which are separated from each other at the end of the same photolithography process that defines the slits. Since SU-8 is transparent, a metal layer such as gold (Au, 0.1 μm) top layer 22 (see also FIG. 2B) was supported on the substrate 50 using an e-beam evaporator. Vapor deposition on the Su-8 layer 21. Individual LITAB bodies 25 (shown in FIG. 2 (b)) are finally released from the underlying substrate 50 by dissolving the alternative Ti layer 52 with an etching solution containing hydrofluoric acid (HF). In this way, the gold pattern on the LITAB blocks the light by reflecting it (towards both the side exposed to the Su-8 layer 21 and the adjacent side) and is not covered by the gold layer The slit transmits light. Since the gold layer 22 blocks light, the open slit transmits light, while the LITAB “barcode” provides a high light signal and high light contrast when transmitted light is detected.

  FIG. 5 shows another embodiment of a LITAB 80 that may include a metal layer 81 as an intermediate layer sandwiched between two polymer layers 82. A barcode pattern is formed on the metal 81. For example, slits 84 having different widths and / or intervals are formed in the metal layer 81. In the illustrated embodiment, the polymer (photopolymer: SU-8) layer 82 is a closed layer (ie, has no slits). The method of forming LITAB 80 may include the steps of first forming first photopolymer layer 82, then forming metal layer 81, and etching slit 84 therein. The second photopolymer layer 82 is formed on the metal layer 81 that fills the slits 84 (eg, by spin coating or curing). Alternatively, the slit 84 may be first filled with another transparent material before forming the second photopolymer layer 82. In this embodiment, the surface condition may be the same for both exposed planes of LITAB to provide similar surface coating and fixation conditions. In other embodiments, LITAB is coated with a polymer or functional molecule such as biotin, carboxylate, or streptavidin; therefore, all beads have the same conditions for molecular immobilization.

  A paramagnetic material is embedded in the intermediate layer of the LITAB and is therefore sandwiched between the first and second layers of the polymer film. Paramagnetic materials include magnesium, molybdenum, lithium, aluminum, nickel, and tantalum. It should be noted that the paramagnetic material on LITAB also acts as a light blocking material and therefore no reflective layer is required. The present invention enables decoding based on transmitted light even in the presence of a paramagnetic material. However, for the prior art, magnetic beads and barcode beads exist, but there is no disclosure that magnetic material is incorporated into barcode microbeads. This is because essentially dark brown magnetic materials are incompatible with reflective barcodes that require alternating light and dark lines.

  FIG. 1 shows an embodiment of preparing LITAB for bioassay. As shown in FIG. 1 (a), LITAB11 allows for multiplex allogeneic bioassays in microvolume samples. A mixture of LITAB 11 corresponding to different codes 14 is placed in a small volume of biological sample 12 in a tube or microwell plate 13. LITAB can then be easily and quickly deciphered as needed. In one embodiment, FIG. 1 (b) shows LITAB 11 functionalized with nucleic acid probe 15 for target hybridization 16 and fluorescence detection 17. Several materials are available for bead fixation. In certain embodiments, LITAB may be coated with a covalent DNA binder used in a microarray. The probe beads were then hybridized in solution to a complementary oligo target carrying a fluorophore covalently attached at the 5 'end. FIG.1 (c) is LITAB imaged on CCD with the microscope. Different subjects provide different fields of view.

  Irregularly oriented microbeads can be decoded on a support such as a slide or on the bottom of a microplate by image processing methods. Once the beads have finally settled and distributed to the bottom of the plane in the microplate, a large number of beads can be decoded simultaneously with a wide vision or scanning imaging camera. Microplates are a standard format for high speed, high throughput clinical assays. Each well is used for one sample; each plate holds 96, 384, or 1536 patient samples for 96, 384, and 1536 wells, respectively. Therefore, it is possible to conduct experiments in the microplate without taking the beads out, and the microbeads can be imaged in a steady state with higher accuracy and sensitivity for decoding. Because incorrect identification can lead to misdiagnosis and mistreatment, the accuracy of decoding is very important for clinical diagnosis. Due to the small bead size, hundreds or thousands of beads can be displayed at the bottom of the microwell with minimal overlap. To minimize bead overlap, the total number of beads is limited to a certain number, depending on the area of the microwell. Both the barcode image and the fluorescence image can be constituted by a conventional microscope or an inverter fluorescence microscope.

  In the embodiment illustrated in FIG. 12, the digital magnetic LITAB analysis system 100 includes an image of beads 11 at the bottom of a light source 102 and support (eg, microwell 106 in the illustrated embodiment) for bead pattern illumination. The optical CCD 104 for acquiring The bottom of the microwell 106 is transparent or translucent, allowing sufficient light to pass through the beads. A scanning or conversion mechanism 110 (shown schematically in FIG. 12) moves the microwell 106 relative to the photodetector 104 (eg, a CCD) and a light source to image the desired well. The photodetector 104 can be used for both barcode image and fluorescence detection. The 1M pixel CCD has enough pixels to elucidate the barcode pattern 14 on the bead 11. A lens and optical filter 108 are used to collect and select excitation and fluorescence wavelengths. Control of the system 100 and / or image processing is performed by the control computing unit 112, which has the functionality and features for performing the processing steps according to the present disclosure. The image of the bead may be stored and / or printed / displayed before being processed by the control arithmetic unit 112, so that the signal or data indicative of the image of the bead is processed in real time or near real time. 112 may be sent.

  In another embodiment (not shown), two light sources are used, one light source 102 is below the microwell for barcode illumination in transmissive mode, and one light source 103 is in incident or reflective mode. Above the microwell for fluorescence excitation. The bar code irradiation light source may be white light, but the fluorescence excitation light source needs a wavelength suitable for the absorption of the fluorophore. By measuring the fluorescence intensity, it is possible to identify which beads have a positive biochemical reaction. The fluorescence intensity may be measured by the same or different photodetectors that image the beads. By decoding the digital barcode image, it is possible to identify which biological probe is immobilized on the surface of the microbead. The choice of light source depends on the fluorophore. For example, a mercury light source with an optical filter cube provides UV for visible light excitation. A red diode laser (665 nm) and a small argon laser (488 nm) or green laser (530 nm) are common laser sources for various fluorophores (eg, phycoerythrin (PE), Cy3, and Cy5, etc.). The optical filter set is designed to select specific excitation and fluorescence wavelengths for various fluorophores.

  As soon as the barcode image is obtained from the CCD, the image data is quickly processed by the image software. These are many image decoding algorithms. Depending on the image pattern, different algorithms can vary in decoding speed or accuracy. One image decoding algorithm includes four main steps: (1) image enhancement; (2) bead segmentation; (3) barcode slit extraction; and (4) barcode decoding. Some of these steps are performed using mathematical software such as a toolkit commercially available from “The Mathworks, Inc.” (eg, MATLAB® version 7.4.0.287 (R2007a); 2007 January 29). The function of these steps is described in the following column.

  (1) Image enhancement: The performance of bead decoding is highly dependent on the quality of the image. The accuracy of the decoding process can be improved by image enhancement. This image enhancement using image intensity normalization provides a uniform intensity background. Inhomogeneous background is often due to inhomogeneous irradiation. In order to achieve high image contrast of the beads, background removal and normalization should first yield a homogeneous background.

  (2) Bead segmentation: The purpose of image segmentation is to delineate the beads in the image for further analysis. The basic segmentation routine places boundaries such as lines or curves in the image where the beads can be positioned in the image. According to one embodiment, the Watershed algorithm in MATLAB is applied to separate bead positions. The watershed algorithm in MATLB requires that the image be a black and white image, so that during part of the image processing, the image is converted to a black and white image. The higher density of black pixels (due to opaque areas) corresponds to the edge of the bead, so the watershed transform transforms the found waterlines in the image and treats the surface surrounded by the dark pixels as beads. . Alternatively, a new watershed algorithm in MATLB can be modified to use grayscale images, which results in more accurate bead segmentation and final decoding. LITAB has a certain area, so each bead is separated and filtered from the image based on that area.

  (3) Barcode slit extraction: After bead segmentation, each bead is processed separately to extract the barcode. Extracted from the main image, the bead regions are considered sub-images and processed one by one. The sub-image shows the bead positioning in a random direction with the major and minor axes. Calculate the angle made by the main axis of the bead using the x-axis of the image. The extraction of the slit from the beads is performed after rotating the main axis of the beads about the x axis. After bead rotation, the bead boundaries are removed and the intensity values are averaged along the length of the bar (y-axis). These sub-images are effectively two-dimensional, but carry only transmission intensity information. The intensity plot along the x-axis of the bead shows peaks with two widths (narrow and wide) corresponding to 1 and 0 bits.

  (4) Bit decoding: In order to decode the bar code, the width of the transmission intensity peak is analyzed. Calculate the width of the peak using a half maximum line. In order to extract binary bit information, five pixels are sufficient to draw a narrow slit ('0') in the bead. The width is converted to binary bits using a 10% tolerance. Depending on the rising and falling edges of the intensity peak relative to the peak representing the end edge of the bead, the most significant bit (MSB) or the least significant bit (LSB) of the barcode can be identified. Based on the identification of the two widths, binary digits can be decoded.

According to the illustrated embodiment, the image processing software consists of the following detailed step-by-step procedure for image decoding:
1. Grayscale conversion and storage of bead image details: The image is read and converted to a grayscale image as shown in FIG. Some of the image processing algorithms used in this programming are from MATLB and therefore require that the image be converted to a black and white image. Thus, during the image processing step, the processed image is further converted to a black and white image if necessary.

2. Background extraction to obtain a uniform background: Eliminate uneven illumination of the background. Each image is processed using components and image calculations to obtain a uniform background. The image is extracted from the background and further processed by adjusting the intensity of the image.

3. Bead image edge detection: The detected edges are shown in FIG. This step recognizes and separates the beads from the background or remaining image. A “sobel” filtering technique is used to suppress all edge boundaries of the beads and emphasize the slits within each bead.

4). Labeling individual bead images: Once the beads are separated as individual regions in the image, the regions are labeled using the RGB color code or by numbering each bead (Figure 8). Different parameters such as area, direction angle, pixel list are calculated for each bead using a region pops algorithm.

5. Display conversion of bead image and digital decoding by intensity plot: To place each microbead on the main axis (FIG. 9), the original grayscale image is taken and converted to a matrix containing pixel values. The pixel index list of the microbead is obtained from the region characteristics. The grayscale image is exchanged with the pixel index list for each microbead, leaving the rest of the pixels at zero. Using the previously calculated angle of each microbead, rotate each bead along the main axis.

6). Intensity analysis for digital code identification: Display microbead images by plotting intensity versus number of pixels as shown in FIG. Each intensity plot is analyzed to obtain a barcode in the microbead. Draw a line parallel to the x-axis and take the x value for each intersection of the lines in the plot. Draw a line with the average of the peaks. This means that the sum of all points in the peak is divided by the number of points. Next, the difference between two consecutive x values is obtained. An alternative value is the width of the barcode in each microbead. By calculating the distance between the edge of the barcode and the first slit, the direction of the barcode can be determined. The longer distance corresponds to the forward direction of barcode reading. If the beads appear to be reversed in the image, the barcode bit is reversed. The decrypted result is displayed as shown in FIG.

7). Segmenting of contacting beads: If two beads contact along an edge, the barcodes do not overlap and thus the bead can be decoded. However, the two beads are segmented as a single bead. To disassemble such contacting beads, a “Watershed Transform” using a Distance Transform and a watershed line is used. Distance conversion is a mapping display of a digital image, giving the distance to the boundary pixel closest to each pixel of the image. The image data may be interpreted as a local surface with a gray level peak in the gradient of the image display watershed. In segmentation, widen the slits of these beads and use distance transformation to find the watershed line. Watershed line creation requires the images to be black and white, so each image is converted to black and white based on the limit intensity. The rectangular structure element of the barcode slit is premised on expanding the image. When the watershed line is created, the image in which the edge is detected is overlaid with the watershed line, resulting in separated beads. These separate beads are irregular at the edges, but the barcode is perfect for decoding.

  Although rectangular beads are handled in the present invention, this can be extended to bent rectangular beads. That is, the bead consists of a rectangle and two semicircles at one end. One of the semicircles is larger than the other, indicating MSB and LSB. To decode the bar code, image processing is performed and the beads are gradually lost except for the semicircle. Region based filters can only extract MSB semicircles from beads. This sure semicircle image is converted to an edge image using a “Sobel filter”. The entire image consists of semicircles and straight lines that form a curve. Identify straight lines using the Hough transform in MATLB. Based on the straight line, the length of the barcode portion of the bead is configured to be perpendicular to the straight line but away from the semicircular region. The bar code is extracted by plotting the intensity inside the constructed box.

  Many attempts have been made to improve the algorithm in order to reduce the amount of computation time for image processing. To perform the initial processing for bead segmentation, the image analysis is reduced from 1 μm / pixel to 10 μm / pixel. Once the beads are segmented, the barcode is extracted using the high resolution image. Another attempt is to change the image decoding algorithm from MATLB to C and compile the C program before execution. This significantly improves the execution speed of the image decoding software. Finally, the decryption program is executed by a coprocessor such as NVIDIA and using the CUDA library to link the program with the coprocessor. This system allows 128 coprocessors to execute programs in a balanced manner.

  The method and system of the present invention has been described above in terms of functional modules. Unless indicated to the contrary, one or more functions may be integrated into a software module in a single physical device or software product, or another physical device, without departing from the scope and nature of the present invention. It is understood that functions that are or software modules may be performed. It will be further understood that the boundary between hardware and software is not always clear.

  It will be understood that a detailed discussion of the actual execution of each step making up the method of the present invention is not necessary to allow an understanding of the present invention. The actual implementation is within the functionality and interrelationship of the various software and hardware components in the system, given the normal skill of the programmer and computer engineer, and the disclosure of the system of the present invention. . One skilled in the art can perform the present invention without undue experimentation by applying normal skills.

While the invention has been described with reference to the described embodiments, it will be apparent to those skilled in the art that various modifications and improvements can be made without departing from the scope and nature of the invention. Accordingly, it is to be understood that the invention is not limited by the specific illustrated embodiments, but only by the claims.

Claims (15)

A microbead analysis system,
A support supporting at least one bead marked with a digital code;
An imaging device for obtaining an image of the bead including the support and the indicia; and analyzing the image obtained by the imaging device, recognizing and separating the bead from the rest of the image, and placed on the bead A decryption system configured to identify the digital code indicated by the indicia;
Microbead analysis system equipped with.   The support supports a plurality of microbeads each having a mark indicating a digital code, and an image obtained by the imaging device includes a plurality of beads including the mark on each bead; Configured to analyze an image to recognize and separate the plurality of beads from each other and the support in the background, and to identify a digital code indicated by the indicia placed on each bead The microbead analysis system according to claim 1.   3. The microbead analysis system according to claim 2, wherein the support includes a plate having a plurality of wells, and the plurality of beads are supported in the wells.   The microbead analysis system according to claim 2, further comprising a light source that irradiates the plurality of beads for imaging.   5. The microbead analysis system of claim 4, wherein the light source is positioned to transmit light through the plurality of beads to the imaging device.   The microbead analysis system according to claim 4, wherein the light source is positioned to make light incident on the plurality of beads.   The microbead analysis system according to claim 6, wherein the beads include a material that generates a light signal, and the imaging device includes an optical device that detects the light signal. Each of the beads
First layer;
A second layer: and an intermediate layer between the first layer and the second layer having a coding pattern defined thereon, partly substantially transparent to light and partly An intermediate layer that is substantially opaque and exhibits a digital code corresponding to the bead;
The microbead analysis system according to claim 2, comprising:   9. The microbead analysis system of claim 8, wherein the intermediate layer includes a series of alternating substantially light transmissive sections and substantially light opaque sections that exhibit the coding pattern.   The microbead analysis system according to claim 9, wherein the relative position, width and / or spacing between the permeable section and / or the opaque section indicates a binary code.   9. The microbead analysis of claim 8, wherein the light transmissive section is defined by a slit through the intermediate layer and the light opaque section is defined by a light reflecting material and / or a light absorbing material. system.   The slit includes a slit having a first width and a slit having a second width. In the binary code, the first width indicates “0” and the second width indicates “1”. The microbead analysis system according to claim 11. The decoding system is (1) grayscale transformation that preserves bead image details; (2) background removal to obtain a uniform background for the bead image; (3) bead image edge detection; (4) bead image contact And (5) labeling of individual bead images; (6) display of digital encoding in each bead image by placement and intensity of the bead image; and ( 7) analysis of the intensity to identify the digital code;
3. The microbead analysis system according to claim 2, wherein the microbead analysis system is configured to recognize and separate a specific bead from the rest of the image and to identify a digital code indicated by a mark affixed on the bead. . Supports multiple beads marked with a digital code;
Obtaining an image of the support and the beads containing said indicia;
Analyzing the image to recognize and separate the bead from the rest of the image and to identify the digital code indicated by the indicia placed on the bead;
A method for analyzing microbeads, comprising each step. The step of analyzing comprises
(1) Grayscale conversion to preserve bead image details;
(2) Background removal to obtain a uniform background for the bead image;
(3) Bead image edge detection;
(4) an additional filtering step of separating the contacting bead images when the bead images are in contact;
(5) Labeling of individual bead images;
(6) a representation of the digital encoding in each bead image according to the placement transformation and intensity of the bead image; and (7) an analysis of the intensity to identify the digital code;
The method according to claim 14, comprising the steps of:

Images