Device and method for diagnostic testing of COVID-19, viruses, antibodies and markers

[0106] The apparatus of the embodiments of the embodiment includes a central unit comprising an electronic component, a camera, an optical element, a digital data collection element, and a communication element communicating with the Internet and cloud-based expert diagnostic servers, and provides COVID-19 and other viruses or Field portable diagnostic detection of bacterial infections. The same centering unit supports at least three different microfluidic discs 68 (CD) for diagnostic measurement or detection, ie, using a surface acoustic wave (SAW) detector for viral detection, microarray for antibodies such as IgG and IgM Serological detectors, and RT-PCR assays using a fluorescent detector for RT-PCR, these detectors are labeled as A10, A20, and A30 CDs, respectively. Detector. The unit and its corresponding CD are measured or assay devices, not performing high level diagnostic analysis, but provides data required for diagnostic analysis to fully develop diagnostic databases and experts in the cloud communication with the internet communication with the central unit. .

[0107] Central unit

[0108] figure 1 The hidden unit 10 shown is a desktop rectangular chassis having a color touch screen 12 on its top surface, and has a closed cover 14, in the cover 14, the microfluidic disc 68 (CD) is placed figure 2 The main shaft 16 shown is operated. The corresponding microfluid disk 68 will be described in detail below. Such as image 3 As shown, one end of the unit 10 is provided with a plurality of data and power connector, such as an external AC power outlet 24, a power switch 22, an external USB port 20, and an external Ethernet port 18. In the interior layout of the unit 10 Figure 4 The main circuit, digital circuit, photoelectric element, and electromechanical components are shown. In the illustration element including the CD motor 26 (which makes the CD rotation), the CD pointer 28, the camera illuminator 30, the camera 32, the power source 34 with cooling fan 38, the motor controller 36, and the control panel 40. The side of the motor 26 is a laser 48 used in the CD operation, for example, for opening the selected valve selected in the CD. Also included, CPU or Raspberry PI 42, electric fuses 44, and second cooling fan 46. On a long side of the unit 10, a fast response (QR) scanner (not shown) is also provided, wherein the patient information is integrated into the data output.

[0109] Now, the support circuit and photoelectric elements are shown by reference. Figure 5Block diagram to better understand the operation of unit 10. The photoelectric control panel 40 supports and coupled to the CPU plate 42 of the power source 34, providing a plurality of DC voltages (e.g., 5 and 24V DC) and ground connections. The CPU board 42 has a RaspBerry PI 43CPU that is the primary control circuit of the unit 10, and processes all advanced programming commands, communication, and data processing. The CPU 41 on the optical plate 40 is a state machine and supplies the required drive and command signal to the motor 26 and the respective LEDs 56 and the laser 48. The CPU 41 controls the speed and rotation supplied by the motor 26 by transmitting a direction-disconnect motor command transmitted to a brushless motor driver 52. The driver 52 also transmits the speedometer signal TACHO to the CPU 41, and receives the reference signal Vref from the OPAMP 53, and the OPAMP 53 communicates with the CPU 41 via the airborne digital-analog converter.

[0110] The CPU 43 is an ARM-based (advanced RISC machine) processor with a Linux operating system. The CPU 43 is coupled to and drives the camera 32 and provides raw image processing via a USB connection to generate a transmitable digital data image, which is ultimately transmitted to the cloud 134 via the wireless module. CPU 43 and the fan 55, clock 35, RAM memory 37 and the eMMC (embedded multimedia controller) 39 Flash, Micro Secure Digital memory card (SD) 61, an earpiece speaker 63 with an audio amplifier 65, the power management circuitry and a power source 71 connected The device 69 is associated. The memory card 61 is used to capture a copy of the detection result of additional transmission on the cloud 134. The audio amplifier 65 will be used with the speaker 63, and the speaker 63 transmits the health status or state of the device to the user (detection state, error, etc.). The CPU 43 is coupled to the display 12 via HDMI and USB connections. Display 12 optionally drives a pair of stereo speaker 13 to communicate with the user. Optionally, CPU 43 via jack 91 and USB port 7 6 DOF inertial measurement unit (IMU) 93, a microphone 95, with a GNSS antenna (GNSS) 97, mouse / keyboard 99, barcode reader 89 Communication, the barcode reader is used to perform position tracking, processing history, and user interaction and developer programming at the site.

[0111] The microcontroller having the CPU 41 (which the memory 43 and the external oscillator / clock 43 in the photop plate 40) is coupled to the CPU 42, and according to Figure 7 The protocol shown in the flowchart is a motor 26, and is provided to provide control over each of the LEDs associated with the RT-PCR process performed on disk 68. Using the on-chip digital-analog converter, the CPU 41 is coupled to the reference and the speedometer input / output of the driver 52 via an operational amplifier 53, and directly supplies and disconnects the motor command directly to the driver 52. The CPU 41 command brushless motor driver 52 drives the main shaft or CD motor 26. The motor 26 includes an encoder that is fed back to the CPU 41 such that its speed and rotation direction is controlled in a closed loop servo mode. The driver 52 provides a three-phase drive signal to the motor 26, which includes a Hall effect sensor that returns the RPM feedback signal indicative of the motor RPM to the buffer 52. The CPU 41 is also coupled to an encoder feedback signal from the motor 26.

[0112] Such as Figure 5 As shown, the photoelectric control panel 40 is coupled to the power source 34 and includes a boost circuit 80 (an increase of 5V power supply to 6V) and low pressure difference regulator (LDO) 82 (3.3V, 1A). 41 clocked by the oscillator 43, and includes a memory 45, a temperature / humidity sensor 47, a Serial Programming (ICSP) and debugging interface 49, through the onboard analog CPU - a reference voltage source VREF is coupled to digital converter 41 and the CPU limit Bit switch 51, the limit switch 51 is built in the cover of the unit 10 such that the motor 26 and all other photoelectric elements are turned off whenever lifting cover.

[0113] It is now possible to understand the operation of the optical plate 40 relative to the disk 68. The movement and position of the disk 68 is tracked by the magnet 66 mounted on the disc, and the magnet 66 mounted on the disc is sensed by the magnetic and optical index drive 64 coupled to the CPU 41, and the angle orientation or position of the disk 68 is determined by the CPU 41. in Figure 8 In step 93, the sample is placed in Figure 7 Sample inlet 94. At step 95, the treated sample is transferred to the blood-plasma separation chamber 72, supplied from the sample inlet 94. Combined below Figure 7 with Figure 8 When the separation, the separated plasma is transferred to the receiving chamber 98, and then transferred to the microarray chamber 74 in which the microarray 92 is provided, in which it is Figure 6 The 593nm LED 268 shown on the LED ring plate 269 is activated, and the LED is driven by an LED driver 86 coupled to the CPU 41. in Figure 5 with Figure 6 In an embodiment, 10 LEDs are provided in the ring around the detecting chamber 209, each operating at 593 nm to provide a substantially uniform IR illumination field to activate fluorescence reading. The camera 32 captures the digital image of a fluorescent label sample through the low pass filter 76 and the lens 78, transmitting the image to the CPU 42, and transmits it from the CPU 42 to the cloud 134. The prepared sample can then be placed in the waste chamber 114. Biological reading is data of the camera 32 captures fluorescence activated microarray 92. The fluorescence intensity corresponds to the concentration of the sample. The camera 32 detects the fluorescence of the microarray 92. The camera 32 is focused on the microarray chamber 72 by lens 76 and the low pass filter 78 to perform fluorescence imaging. The LED driver 86 is included in the photoelectric controller 40, which drives 593 nm LED to activate fluorescence in the labeled tag in chamber 74.

[0114] A20-disk operation

[0115] Before discussing the diagnostic method of COVID-19 on the microarray, it is now considered that when using the microarray detector 92, the general operation of the disk 68, such as Figure 7 A top plan view shown. On disk 68 having a diameter of 70 mm, a thickness of 4.5 mm, a member, a two-form, two copies, such that redundancy can be redundantly measured or using different microarray 92 to perform two separate antibody detection simultaneously with the same patient. . The disc 68 is made of transparent plastic and has a plurality of chambers, channels, and valves, channels, and valves, as described in detail below. The disk 68 can be sealed on its top and bottom surface by a thin plastic laminate. The method starts from step 193, which will insert the sample inlet 94 from the sample from the patient from the patient. Such as Figure 8 During the flow chart, in step 193, the disk 68 is rotated at 5500 rpm for 1 minute to push the sample into the blood-plasma separation chamber 72, in the blood-plasma separation chamber 72, centrifugal action will be heavier blood components from Separated in plasma. The first laser valve 62 is opened by the positioning disk 68 such that the laser valve 96 (which is a plug) is aligned with the underlying laser 48 in the unit 10. The laser 48 is excited, open the valve 96, and plasma or serum flows from the separation chamber 72 to the microarray chamber 74 in which the microarray 92 is provided in the microarray 98 in about 0.5 minutes. At step 197, the transferred serum reciprocates 40 cycles in the microarray chamber 74, and reacted with the antibody point of the microarray 92 for about 5 minutes, then initiated at 170 rpm, then caused by initiating The siphon 93 rotates the discharge chamber 74 to the waste chamber 114 at 1000 rpm.

[0116] At 1999, the laser valve 106 is aligned with the laser 48 in the unit 10, and opened with 0.5 minutes. Thereafter, the cleaning buffer # 1 stored in the chamber 100 is transferred to the microarray chamber 74 by the step 201 at 2700-5428 rpm to the microarray chamber 74, in step 197, then at 170 rpm The chamber 100 was initiated, and then a rotating discharge room was about 2 minutes at 1000 rpm.

[0117] At step 203, the laser valve 108 is aligned with the laser 48 in the unit 10, and open with an exposure of 0.5 minutes. At step 205, the second antibody stored in chamber 102 is transferred to the microarray chamber 74 by approximately 5 minutes (20 cycles) of 2700-5428 rpm, and then initiates at 170 rpm, then The rotary emissions room of 1000 rpm is 74 for about 2 minutes. The second antibody is an anti-antibody. Antibodies in the blood bind to antigen. The second antibody is an antibody that specifically binds to the tail of the antibody in the blood sample. This second antibody carries fluorescent labels.

[0118] At step 207, the laser valve 110 is aligned with the laser 48 in the unit 10, and open with a 0.5 minute exposure. Thereafter, the cleaning buffer # 2 stored in chamber 104 is transferred to the microarray chamber 74 by 5 minutes (20 cycles) at 2700-5428 rpm at step 209, in step 197, then at 170 rpm The chamber 104 was initiated, then a rotating discharge room was about 2 minutes at 1000 rpm.

[0119] At step 211, the valve 112 is aligned with the laser 48 in the unit 10, and open with 0.5 minutes of exposure. At step 213, the disc 68 is rotated at 5500 rpm for about 1 minute to rotate the drying chamber 74, wherein the cleaning liquid # 2 is discharged to the waste chamber 114. The moving chamber 74 and the microarray 92 are then aligned with the camera 32 in unit 10. At step 215, one or more grayscale images of induced fluorescence were taken by the camera 32, and the CPU 42 stores and transmitted to the cloud in about 1 minute to perform data processing and diagnostic analysis, as described below.

[0120] The total time required for the measurement is about 16.5 minutes.

[0121] Cloud processing and diagnosis

[0122] The unit 10 uses the disc 68 and the detector provided in the disk 68 for physical assay detection. This produces some form of raw data. The unit 10 does not process the data, nor will it analyze it to give the patient's diagnosis, but transmit the original data to the cloud, in the cloud, the remote server provides data processing and diagnostic analysis. Use information associated with the patient's scan QR code or information in the QR code in the patient, then store the detection result in the database, and transfer the patient's computer, smartphone or other related medical providers The electronic address without the need for unit 10 to further participate.

[0123] Figure 9 It is a graph that performs data processing at a high level in A20. At step 216, the scan is assigned by the medical provider to the QR scan of his or her patient, associated with the detection, in step 218, the disk barcode is scanned to associate the disc with the same detection. Bind Figure 7 with Figure 8 The measurement is performed at step 220, and at step 222, the fluorescence excitation image of the captured microarray 92 is completed. Then, at step 224, the unit 10 sends one or more grayscale images which capture it to the cloud. At this time, the function of the unit 10 is ended in the detection.

[0124] The unit 10 operates under software control before transmitting the captured data. Figure 10 Indicated. The test can be activated by the touch screen display menu button 116 QA detection harness or activate operator interface (Human Machine Interface HMI) on the mass of 12,118 to start or wiring harness assembly unit 10, both of which are activated using Java Script ObjectNotation ( JSON) Operation. JSON is a type of data that contains human readable components. Using JSON is because it is independent of the operating system, safe and lossless (there is no data loss from the original data from the camera sensor). Figure 11A screenshot of the touch screen 12 is described. When the operation interface 118 is activated by activating the switch, the displaying QR code button and the detection button are displayed to scan the patient QR code, respectively, the patient is associated with the detection, and then Determination. Once the scan QR code button is activated, the operator will see Figure 12 The screen and you can access the gear icon to set the WiFi via the QR code. The gear icon is an icon on the human interface (GUI). When it is touched, it makes the user can manually or pass the WiFi information of the device through the QR code.

[0125] The unit 10 operates in the client / python module 122, the client / Python module includes a response operation to an external communication and a Linux Oracle programming protocol operation stored according to the airborne. The operation interface 118 communicates with the autonomous running backend software 120 that controls all the operations of the unit 10 through the device control module 124. The main function includes two-way communication of cloud module 126, hardware control module 128, and database module 130.

[0126] Figure 13 The backend operating architecture is shown. Database 130 is a SQLite device database module. SQLite is a widely used C language library that implements small, fast, self-contained, high reliability, and fully fully equipped SQL database engines. SQLITE is an embedded SQL database engine. Unlike most other SQL databases, SQLite has no separate server process. SQLite reads and writes a normal disk file directly. A complete SQL database with multiple tables, indexes, triggers, and views is included in a single disk file. SQLite is a compressed library. The Python device control 124 communicates with the database module 130 and includes a Python hardware control 128 as an operator module that controls the CD motor 26, the camera 32, the laser 48, and other electronic and electromechanical devices of the unit 10. Equipment Control 124 communicates with mild integrated graphic network (LVGL / C-HMI) 118 via JSON and advanced first out (FIFO), the mild integrated graphical area is an open source graphics library, which provides a touch screen 12 available Tools required for graphical elements, visual effects, and low memory occupied embedded graphical user interfaces (GUI). Equipment Control 124 and LVGL / C-HMI 118 are supported by the library of C executable library 132, which communicates with the QR reader 50. Oracle or Linux cloud passing module 126 communicates with the cloud 134 with the Red Hat Pack Manager (RPM) protocol, which is used to store mounting packages on the Linux operating system. The C executable library 132 communicates with the cloud 134 using JSON code.

[0127] Image processing in the cloud

[0128] As described above, the unit 10 generates the original digital image taken by the camera 32, and transmits it untreated to the cloud 134. The purpose is to convert the scanned microarray image to a scalar value of each microarray spot or site. Such as Figure 14 As shown, the image data processing is performed by the following steps: alignment 136, spots detection 138, and spot analysis 140. The microarray spot image analysis is described in detail in Bell et al., "An Integrated Digital Imaging System and Microarray)," GRACE BIOLABS, BEND, OREGON). Have written Figure 15 The program structure is modular in each stage of the analysis. The image 142 and JSON information file 144 are transmitted each phase. Each stage is working and hand over the results to the downstream process.

[0129] The main object of alignment step 236 is to correct the image inconsistency, including rotation angles, proportions, and background noise. Alignment algorithm filter all shapes in the image, look for objects that meet the spot or reference points. After finding any potential spots or reference points, the program looks for a spacing ratio between all potential reference points that match the reference point mode indicated in the JSON mode file. Once the reference point is found, the image is rotated and cropped at step 246 to include only the region of interest. All processes are done on grayscale images.

[0130] Import the initial or original grayscale image 142. Image 142 includes background information or noise that is independent of the processing of image 142. The alignment phase is intended to remove the non-interested region (NROI) information by identifying three bright base spots on the corner of the microarray. Applying bilateral filters to images to reduce noise, but maintain sharp edges for downstream processing. Next, the image 142 is processed by the adaptive threshold filter to obtain a binary image of the contour. Each contour is then filtered for a dimensional range or pixel region. The size range is pre-identified and scaled with the size of the image. Ignore too much or too small outline. The remaining contours have minimal fittings that are drawn around their peripherals; the area of ​​this circle is compared to the area of ​​the contour to determine how many "circles" are "round". The contour of the area and the area of ​​the boundary circle are retained. After identifying the potential reference point, the program compares the distance ratio between all sets (combinations) of three contours, looking for the ratio of the theoretical reference point pitch than the mode file ( Figure 22A , Dashed circle). Define the matching set of three contours as a reference point. After determining the reference point, the minimum fitting rectangle is drawn around three reference points. Figure 22B , Dashed rectangle). Then cropped and rotated the minimum fitting rectangle so that the reference point is on the left, left and upper right ( Figure 22C ). Adding the location of each reference point and the general information about alignment programs to JSON mode, the mode returns together with the cropped image, used for the next downstream application.

[0131] In the spot detection step 238, the main purpose is to determine where each microarray spot is located within the image of interest. It will be used to downstream to determine each spots. Using the reference point of the microarray and the known size, the cropped image is subdivided into a grid, where each square should contain spots. Apply an adaptive threshold within each square of the grid. The adaptive threshold image of each square is used to calculate an image moment, the image distance for determining the centroid of the spot:

[0132]

[0133] I p It is pixel strength in pixel P, with It is the distance vector of pixel P relative to the reference point, n is the total number of pixels in the grid region. If the spot is detected, the spot diameter is measured in each square. If the visible spot is not detected, the average diameter of the discovered spots is allocated for each spot.

[0134] The purpose of the spot analysis phase is to assign a single scalar value for each spot in the grid. Currently, this is done by calculating the inner bombs in the front desk and subtracting it from the background average intensity. Separate each spot and calculate the median of each spot. Similarly, calculate the average of each rear ring ring and subtracted from the median meditile number ( Figure 23 ). Pack the analytical output value into the JSON structure of each spot and return it as the final result.

[0135] Diagnostic processing in the cloud

[0136] The microarray used in the embodiments used in the embodiments are first considering the details of the treated image data in the cloud. The "multi-antibody array" in disk 68 provides individual viruses "exposed fingerprint", "conventional antibody spectrum", which reflects the history of past exposure and vaccination. Such array analysis methods have richer data (e.g., 67 antigens having four repetitive samples), and more quantifiable than the transverse flow detection of antibodies currently used to measure antiviral antibodies. In order to understand this, the inventors are Figure 16A The positive and negative 2019 nCov array sensitivity IgG results obtained from COVID-19 Washington State 820 have been shown in 2020.

[0137] Human and animal antibodies have previously been developed, and high throughput clones from more than 35 kinds of medical pathogens have been developed and the microarray, including bacteria, parasites, fungi and viruses (such as vaccinia, monkey pox). Herpes 1 and 2, chickenpox herpes, HPV, HIV, dengue fever, epidemic cold, Sinai River, base hole Kenya disease, adenovirus and coronavirus). The DNA microarray (also known as DNA chip or biocontrol) is a collection of microscopic DNA spots connected to the solid surface. The DNA microarray is used to simultaneously measure the expression level of a large number of genes or for genotyping of multiple regions of the genome. Each DNA spot contains Pi Mole (10 -12 Moore) specific DNA sequence called probes (or reported molecules or oligonucleotides). These may be a short segment for hybridization cDNA or cDNA or cDNA (also known as antisense DNA), sample (referred to as target) under high strict conditions. It is typically detected and quantitative probe-target hybridization by detecting a target of fluorophore, silver or chemiluminescent labeling to determine the relative abundance of the nucleic acid sequence in the target. The initial nucleic acid array is a macro array of about 9 cm × 12 cm, as well as the first computer image analysis published in 1981. The inventors have explored more than 25,000 samples from humans and animals from infective pathogens, and identified more than 1000 immunomotropic and candidate vaccine antigen against these pathogens. The inventors have shown that the single protein / antibody printed on these array 92 captures antibodies and / or antigens present in serum from infected individuals, and can quantify the amount of the captured antibody to quantify the amount of antibody of the fluorescent secondary antibody.

[0138] In this way, the comprehensive profile of antibodies generated after infection or exposure can be determined, which is characterized by infection types and disease stages. Array 92 can be produced and a large number of detection (500 serum or plasma samples per day), while each sample consumes <2 μL. This microarray method enables researchers to assess a large number of sample antibody libraries, while other technologies cannot be realized.

[0139] The coronavirus antigen microarray 92 (COVAM) is constructed, which contains 67 antigen causing acute respiratory tract infections. The virus anti-original from the array 92 comes from an epidemic coronavirus, including SARS-COV-2, SARS-COV, MERS-COV, ordinary cold coronary virus (HKU1, OC43, NL63, 229E), and epidemic cold, Multi-subtypes of adenovirus, meticulotic virus, subflu virus and respiratory syncytial virus. Such as Figure 17 The pattern, the SARS-COV-2 antigen on the array 92 comprises a printer (S), a receptor binding (RBD) S1 and S2 domain, intact protein (S1 + S2) and nuclear mats (NP) ). A group of antigens from SARS-COV, MERS-COV and four ordinary cold coronaviruses are present on the array.

[0140] In order to determine the antibody spectrum of SARS-COV-2 infection, SARS-COV-2 rehabilitation period from PCR positive individual (positive group) and the uninfected individual (negative control group collected prior to COVID-19) were evaluated. The serum is differentially reactive for these antigens. Such as Figure 17 The hot map is shown, the positive group is highly reactive to the SARS-COV-2 antigen. IgG is more obvious than IGA. Although the negative control has high reactivity to ordinary cold coronavirus antigens, it is not reacted with SARS-COV-2, SARS-COV or MERS-COV antigen. Such as Figure 17As shown, the positive group exhibits high IgG reactivity on SARS-COV-2NP, S2 and S1 + S2 antigen, and the degree of reactivity to SARS-COV-2S1 is low. The positive group also showed high IgG cross-reactivity against SARS-COV NP, MERS-COV S2 and S1 + S2 antigen, and the negative group indicated S1 + S2 and S2 from SARS-COV-2 and MERS-COV. The low cross-reactivity of the antigen has not been cross-reactive for other SARS-COV-2 antigens.

[0141] Table 1 contains Figure 18A The fluorescence intensity result of the IgG shown, Figure 18C The z score statistics of the fluorescence result of the fluorescence Figure 18B The fluorescence intensity result of the IgM shown and Figure 18D Z score statistics of fluorescence results. The Z fraction shows the standard deviation of the confirmed positive IgG or IgM sample above (positive) or lower than (negative) average negative results. The statistically significant Z score (5 or higher) has a shadow number.

[0142] Then use the recipient operation feature (ROC) curve to evaluate the original evaluation to distinguish the positive group and the negative group of the entire measurement truncated value range, wherein the surface area (AUC) is measured. As shown in Table 1, ROC AUC> 0.85 defines a high performance antigen for detecting IgG. Four antigens are listed as high performance antigens: SARS-COV-2NP, SARS-COV NP, SARS-COV-2 S1 + S2 and SARS-COV-2_S2. Additional high-performance antigens include SARS-COV-2S1 (with mouse FC tags) and RBD, and MERS-COV S2. The best sensitivity and specificity of the seven high-performance antigen is also evaluated based on the octopus index. YOUDEN's J statistic (also known as an agreement index) is a single statistic of the performance of the two-point diagnostic detection. The information is a summary of the multi-class situation and evaluates the probability of a sensible decision. It can be seen that the SARS-COV-2S1 is the lowest sensitivity, which is related to the relatively low reactivity of the antigen in the positive group. It can be seen that the specificity of SARS-COV-2S2 is the lowest, which is related to the cross-reactivity of the antigen as seen in the subgroup of the negative group. To assess the performance improvement by combined antigen, all possible combinations of up to four in seven high performance antigens are performed in the computer to distinguish positive and negative groups. Calculate each combined AUC, sensitivity, and specific ROC curve. By combining two or three antigens, performance is significantly improved. For IgG, the two antigen combinations using SARS-COV-2S2 and SARS-COV NP have achieved optimal distinguishing, similar performance (AUC = 0.994, specificity) when the SARS-COV-2S1 with mouse FC tag is added 1. Sensitivity = 0.944). The addition of the fourth antigen reduces performance.

[0143] 0.86 is marked.

[0144] Figure 18A -D shows an example of a single confirmation of positive patient results. Figure 18A The normalized fluorescence intensity of various IgG antibodies in serum is shown, and the two lines show the averaging results of the confirmated positive (above) and confirmed negative (below). Figure 18B The normalized fluorescence intensity of various IgM antibodies in serum is shown, and the two lines show a averaging result of confirmed positive (above) and confirmed negative (below). Figure 18C The Z score of the IgG antibody between the positive and negative results, three dashed lines represent each Z score threshold of mild, moderate, and significant response. Figure 18D The z score of the plot of the IgM antibody between the positive and negative results is shown, and three dashed lines represent each Z score.

[0145] More specifically, A20 serological detection is an optical microarray detection that performs indirect immunofluorescence to qualitatively detect IgM and IgG antibodies in human blood. Serological detection is intended to help identify individuals with adaptive immune responses to SARS-COV-2, indicating recent or previous infections. Serological tests currently produce intensity maps of the image of the microarray and the intensity of the spots on the array. In order to develop diagnostic criteria, known RT-PCR positive and negative samples were detected on the above apparatus. This is a closed threshold for each of the three SARS-COV-2 antigens in the microarray, so that the device can autonomously provide a qualitative "YES" (reactivity) or "no" (non-reactive) )result.

[0146] Microarray description

[0147] Serological detection includes two identical microarrays of the disk 68, one for detecting the presence of IgG, and the other for detecting the presence of IGM. These two types of antibodies were detected by using IgG or IGM report antibodies. Each of the two microarrays has Figure 19 The form shown in. The characteristics of the microarray spots are:

[0148] a. Negative control: buffer (10 spots): phosphate buffer containing 0.001% Tween-20 (polyethylene glycol sorbitan monochromerate, polyoxyethylene sorbitan monochromerate) Salt water (PBS). These spots are printed buffers and as a negative control to determine baseline fluorescence of the array.

[0149] b. Positive control 1: huigg (5 spots): Human IgG in 8 diluted concentrations of 0.3 to 0.001 mg / ml, a total of 40 spots. These spots are used as a positive control to indicate that when serum samples are detected, IgG's report antibodies apply appropriately to accurately determine the cutoff of the array. The concentration step can be used as a rough guide to explaining microarray fluorescence.

[0150] C. Huigm (5 spots): Human IgM, a total of 40 spots in 8 diluted concentrations of 0.3 to 0.001 mg / ml. These spots are used as a positive control to indicate that when serum samples are detected, the IgM's report antibody is appropriately actuated to accurately determine the cutoff of the array. The concentration step can be used as a rough guide to explaining microarray fluorescence.

[0151] d. positive control 2: A. Huigg (3 spots): Anti-human IgG printed with 0.3, 0.0.1 and 0.03 mg / ml concentration. These spots were used as a positive control to indicate that there were IgG antibodies in the sample. A. Huigm (3 spots): Anti-human IgM printed with 0.3, 0.0.1 and 0.03 mg / ml concentration. These spots were used as a positive control to indicate that there were IgM antibodies in the sample.

[0152] e. antigen: SGC-SPIKE19200701 (8 spots): SARS-COV-2 thorn protein (Oxford University). Take 0.2 mg / ml printing. SARS-COV2.NP (8 spots): SARS-COV-2 nuclear casing protein (sinobiological). Take 0.2 mg / ml printing. SARS-COV2.RBD.MFC (8 spots): SARS-COV-2 mining protein (SINOBIOLOGICAL). Take 0.2 mg / ml printing.

[0153] f. Benchmark Point (3 Spots): Streptavin, Alexa Fluor 647 Combination. These spots are designed to be the brightest spots on the array and are used to position and oriented arrays.

[0154] G.PBST cleaning liquid (21 spots): PBS + 0.05% Tween 20 is used to clean the needle.

[0155] H. Blank (2 Spots): Unused microarray locations.

[0156] Microarray result

[0157] Upload an image of each microarray in A20 serological detection to a server on the Oracle cloud. After positioning and directing the microarray using a corner base point, the image is analyzed, and the scalar value is generated for each spot in the microarray. These measurements are the median fluorescence intensity of each spot, minus the average fluorescence intensity of the surrounding ring. These measurements will be provided to the user online in JSON format, and also provide a graph that aggregates three SARS-COV-2 antigens on the microarray. The JSON file is a hierarchical file with the following top structure:

[0158] Top Overview of JSON

[0159]

[0160]

[0161] Each spot measurement is included in the list of "Spots" entries, which contains comprehensive details of each spot:

[0162] JSON Details

[0163]

[0164]

[0165]

[0166] An outline map of each microarray is a histogram that reports the value of each control spot, as well as the average value and standard deviation of each antigen, such as Figure 20 The example is shown. These results are established to establish a routine statistical model to distinguish between blood samples without SARS-COV-2 antibodies.

[0167] Whole system

[0168] Figures 21A-21E The entire user flow or the user's interaction is shown. in Figure 21A In the case, the operation of the patient, unit 10, the cloud 134, and the detection operator operating unit 10 are respectively shown in four horizontal rows. At step 400, the patient is logged in to the Internet to schedule diagnostic detection at the available detection location. At step 402, the remote server in the cloud communication of the portal is arranged to arrange the patient's detection and generate a unique QR code, which has: 1) detection time and place; and 2) to be detected, ie, whether to run and A10 The disk 68 associated with the A20 or A30. QR code is how patients control detection information and their privacy. At step 404, the QR code is sent to the patient, and the patient downloads it to his or her smartphone, laptop or computer. At the same time, at step 406, the remote server in the cloud 134 transmits the patient's reservation information and inserts it into the detection schedule. At this point, the program may be suspended or more days before taking further action.

[0169] in Figure 21B In step 408, at the time of the appointment, the patient goes to the detection location and scans his or her QR code entry unit 10 at step 410, in response to the screen prompt. At step 412, the validity of the QR code is checked in the cloud 134, and the remote server communicates with the detection site, which one of the disk 68 is detected, and the authorization unit 10 uses a particular type of disk 68. The detection operator checks the humidity and temperature level on the disk package to verify the integrity of the disk 68, in response to the screen prompt of the unit 10 in step 418 in step 418. At step 422, the cloud 134 transmits metadata of the disk to the unit 10, which includes a state of the disc batch, the JSON file for rotating the protocol, and the grayscale TIF of the microarray 92 generated during the disk 68 quality control test. file. At step 420, the unit 10 downloads the metadata of the disk from the cloud 134 and determines the permission of the disk 68 in the detection.

[0170]If it is determined in step 420 to be rejected, then at step 424, the operator is suggested to discard the disk 68 and replace it to another, and then the program returns to step 414. If you authorize the disk 68, at step 426, the blood sample is acquired from the patient from the patient, such as hand-finger blood, in step 428, by the detection operator to the disk 68, and then load the disk 68 in step 430 to the unit 10. in Figure 21C Step 432, the unit 10 displays the screen prompt to the detection operator to start detection. In response, at step 434, the operator touch the start button on the screen display.

[0171] In step 436, the pre-detection diagnostic data is collected, which includes at step 438, two microarray 92 in disk 68, by verifying the following content checking optical systems: 1) The three reference spots in each array are visible; 2) The reference point intensity is within 20% of the microarray original image; and 3) The reference spot is aligned. Also, in step 440, the monitoring program outputs diagnostic data from the camera 32, the LED 56, the motor 26, and the laser 48. Thereafter, at step 442, the unit 10 runs the rotation protocol on the disk 68 as described above, and takes a gradation image of each microarray 92 at the end of the detection. At step 444, the monitoring program in the CPU 43 continues to monitor the unit 10 during the measurement process, and generates an error message display when a fault occurs, and stops detection or determination when needed.

[0172] in Figure 21D In step 446, upload the grayscale TIF image of each microarray 92 taken by the camera 32 to the cloud 134, the two microarrays are used to provide background data, and after detection of images for providing detection Data, files include diagnostic data acquired before and after detecting. At step 448, the digital image of the microarray 92 is processed as described above to generate a JSON file, which lists each spots name, spot position, and fluorescence intensity, and is diagnosed therefrom.

[0173] in Figure 21e Step 452, by the JSON output file, determine whether the detection process is passed or failed. If the pass is passed, the diagnosis is predicted, and the confidence interval is calculated. Predict diagnosis is provided by statistical model, such as logic regression or random forest using fluorescence intensity or calculated Z score. At the same time, at step 454, an antigen list having an average fluorescence intensity value and a standard deviation is drawn. At step 456, these results are transmitted from the cloud 134 to the patient's smartphone, laptop, or computer, and according to the level of access, the patient can see the results, the drawings, and / or raw data. Then in step 458, the patient can choose to forward the detection data to him or her doctor, medical technicians, research institutions, government authorities, or patients think it must be necessary.

[0174] Data chain recognition

[0175] use Figure 22 The identification chain 300 implements the control of the data transmitted to the remote server in the cloud 134. The identification chain 300 can be seen as a tree map, such as Figure 22 As shown, each of the frames is a node that can be passed in any direction, and each node corresponds to the fabrication of the manufactured. You can recurrent each node in the query chain to generate information of its corresponding components, or its own manufacturing directory number, date, place of origin, and batches. Each time you detect 302 encoded in the transmissiond image analysis data file 304, the file 304 includes a TIFF packet 306 that is binding to a unique patient / detection code 308, unique machine ID 310, unique box code 312 And the UTC timestamp for detecting. Connect the detection code to the patient / detection code 308, the machine ID 310 and the timestamp 314 guarantee that the two detection results are not erroneously identified, since it cannot be performed on the same machine at the same time.

[0176] Additional unique box code 312 further ensures uniqueness of each detection and results, and a complete identification chain is created to connect a particular detection 302 and its result to each related assembly involved in the detection 302. Components. This provides complete traceability such that all component batch numbers used in the specific disk 68 can be identified, or all disc 68 using a specific component batch number. This makes it possible to obtain data from the COMPROSED detection and determine a defective component batch or recall all the discs using the defective component batch.

[0177] The machine ID 310 is uniquely defined by its camera serial number 316 and an airborne computer (PI raspberry) serial number 318. The machine ID 310 can then provide a hierarchy of all sub-components of all of its mechanical and electrical components.

[0178] The box code 312 can be tracked to the box assembly 320, which details the assembly date 328, microarray information 322, disc information 324, and reagent catalog 326 stored in the box. The disc information 324 includes details of disk design 330 and disk injection batch 332. The microarray information 322 includes a printing date 334, a microarray layout 336, a slide chip batch 338, a printed protein catalog 340, a printed protein catalog 340, a printing protein catalog 340, a printing protein catalog 340, a microarray batch 338, a printing protein catalog 340, and nitrocellulose batch 342. Tide Eventory Batch 338 refers to the slide batch 344.

[0179] Many of ordinary skill in the art can make many changes and modifications in the art without departing from the spirit and scope of the embodiments. Therefore, it must be understood that the embodiments illustrated are illustrated only for examples, and it is not to be construed as limiting the embodiments of the following embodiments and various embodiments thereof.

[0180] Therefore, it must be understood that the embodiments illustrated are illustrated only for examples, and it is not to be considered as limiting the embodiments as defined by the appended claims. For example, although the following sets the elements of the claims are set forth in some combination, it must be clearly understood that the embodiments include fewer, more or other combinations of different elements, even when they are not protected by this combination. They have been disclosed above. The teachings of the combination of two elements into the claimed combination should also be understood to allow such required combinations, wherein the two elements are not combined with each other, but may be used singly or in other combinations. Deletion of any disclosed element of the embodiment is explicitly taken into account within the scope of the embodiment.

[0181] The words used to describe various embodiments are not only to be understood from the meaning of common definitions, but also include special definitions of the present specification, materials or functions exceeding the meaning of the commonly defined meaning. Therefore, if an element can be understood in the context of this specification to include more than one meaning, it must be understood in the claims to be understood to cover all possible meanings supported by the instructions and words themselves.

[0182] Accordingly, definitions of words or elements in the appended claims are defined in this specification to include not only a combination of elements described above, but also include substantially the same functionality in substantially the same manner. All equivalent structures, materials, or effects of substantially the same results are obtained. Therefore, in this sense, it is possible to perform two or more elements of any of the appended claims, or two or more elements in the claims in the appended claims. Although the elements can be described above as a combination and even initially claimed, it should be clearly understood that in some cases, one or more elements of the claimed combination can be from the combination. Delete, and the required combination can involve sub-combination or sub-combinarity variants.

[0183] It is clearly considered equivalent to fall within the scope of the claims (now known or later design) as seen in the art. Accordingly, the clear replacement of ordinary skill in the art is defined within the range of defined elements.

[0184] Accordingly, the claims should be understood to include the content, conceptual equivalent content, conceptual equivalent content, which can be apparent, and can be significantly replaced, and the content of the basic idea of ​​the implementation.

[0185] Table 1

[0186]

[0187] Table 2

[0188]