Image-based assay using mark-assisted machine learning
The present disclosure relates to devices, apparatus and methods of improving the accuracy of an image-based assay. One aspect of the present invention is to sandwich a sample between two plates and add reference marks in the sample areas of the plates, with at least one of the geometric and/optical properties of the reference marks being predetermined and known, and taking images of the sample with the reference marks, and applying a machine learning model in the analysis of the image-based assay.
1. An apparatus for assaying an analyte in a sample, comprising:
(a) a sample holder comprising two plates, wherein each of the two plates has a sample contact area, wherein the two plates sandwich the sample between the sample contact areas of the two plates into a thin layer,
(b) a plurality of reference marks on the sample contact areas of the two plates, wherein the shape of each of the reference marks and/or the location of each of the reference marks are predetermined and known;
(c) an imager that images the sample contact area; and
(d) a non-transitory computer-readable medium storing a machine learning algorithm for analyzing the analyte, wherein the machine learning algorithm utilizes the images of the reference marks.
2. A method of assaying an analyte in a deformable sample, comprising:
(a) obtaining the apparatus of claim 1 ;
(b) placing the sample into the sample holder, wherein at least a part of the sample is between the two plates;
(c) imaging, using the imager, one or more images of the sample and the reference markers; and
(d) detecting the analyte by analyzing the one or more images and the machine learning algorithm.
3. The method of claim 2 , wherein the one or more images comprising a bright image and a fluorescent image.
4. The apparatus of claim 1 , wherein the reference marks comprise the spacers, wherein the spacers are between the two plates and regulate a gap between the two plates; and wherein the spacers are pillars.
5. The apparatus of claim 1 , wherein the reference marks are periodic and the period is predetermined and known.
6. The method of claim 2 , wherein the analyzing of each image comprises determining a rotational orientation of one or more of the reference marks about an axis relative to the imaging system.
7. The method of claim 2 , wherein the analyzing of each image comprises comparing image information about the reference marks to a prior knowledge about the reference marks.
8. The apparatus of claim 1 , wherein the sample contact area further comprises a detection agent comprising antibodies configured to specifically bind to protein analyte in the sample.
9. The method of claim 2 , wherein the sample contact area further comprises a detection agent comprising oligonucleotide probes configured to specifically bind to DNA and/or RNA in the sample.
10. The apparatus of claim 1 , wherein one or both plates is flexible, the reference marks comprise spacers having an inter-space-distance (ISD), and wherein the thickness of the flexible plate times the Young's modulus of the flexible plate is in the range of 60 to 550 GPa-um, and the fourth power of the ISD divided by a thickness of the flexible plate (h) and the Young's modulus (E) of the flexible plate, ISD 4 /(hE), is less than 10 6 um 3 /GPa.
11. The method of claim 2 , wherein the sample is selected from the group consisting of: amniotic fluid, aqueous humour, vitreous humour, blood (e.g., whole blood, fractionated blood, plasma or serum), breast milk, cerebrospinal fluid (CSF), cerumen (earwax), chyle, chime, endolymph, perilymph, feces, breath, gastric acid, gastric juice, lymph, mucus (including nasal drainage and phlegm), pericardial fluid, peritoneal fluid, pleural fluid, pus, rheum, saliva, exhaled breath condensates, sebum, semen, sputum, sweat, synovial fluid, tears, vomit, urine, and any combination thereof.
12. The method of claim 2 , wherein the reference marks comprise the spacers, wherein the spacers are between the two plates and regulate a gap between the two plates; and wherein the spacers are pillars.
13. The method of claim 2 , wherein the reference marks are periodic and the period is predetermined and known.
14. The method of claim 2 , wherein the sample contact area further comprises a detection agent comprising antibodies configured to specifically bind to protein analyte in the sample.
15. The method of claim 2 , wherein the sample contact area further comprises a detection agent comprising oligonucleotide probes configured to specifically bind to DNA and/or RNA in the sample.
16. The method of claim 2 , wherein one or both plates is flexible, the reference marks comprise spacers having an inter-space-distance (ISD), and wherein the thickness of the flexible plate times the Young's modulus of the flexible plate is in the range of 60 to 550 GPa-um, and the fourth power of the ISD divided by the thickness of the flexible plate (h) and the Young's modulus (E) of the flexible plate, ISD 4 /(hE), is less than 10 6 um 3 /GPa.
17. The method of claim 2 , wherein the sample is selected from the group consisting of: amniotic fluid, aqueous humour, vitreous humour, blood (e.g., whole blood, fractionated blood, plasma or serum), breast milk, cerebrospinal fluid (CSF), cerumen (earwax), chyle, chime, endolymph, perilymph, feces, breath, gastric acid, gastric juice, lymph, mucus (including nasal drainage and phlegm), pericardial fluid, peritoneal fluid, pleural fluid, pus, rheum, saliva, exhaled breath condensates, sebum, semen, sputum, sweat, synovial fluid, tears, vomit, urine, and any combination thereof.
18. The apparatus of claim 1 , wherein the reference mark comprises spacers having an inter-space-distance, and wherein the inter-space-distance is from 7 um to 200 um.
19. The apparatus of claim 1 further comprising a second camera.
20. The method of claim 2 , wherein the imaging images cells in the sample.
21. The method of claim 2 , wherein the imaging images red blood cells, white blood cells, and/or platelets in the sample.
22. The method of claim 2 , wherein the analyte is a protein.
23. The apparatus of claim 1 , wherein the machine learning algorithm comprises a comparison of image information about the reference marks to a priori knowledge about the reference marks.
24. The method of claim 2 , wherein the machine learning algorithm comprises a comparison of image information about the reference marks to a priori knowledge about the reference marks.
25. The apparatus of claim 1 , wherein one or both of the sample contact areas of the two plates comprise one or a plurality of storage sites that each store a reagent or reagents.
26. The method of claim 2 , wherein one or both of the sample contact areas comprise one or a plurality of storage sites that each store a reagent or reagents.
27. The method of claim 2 , wherein the analyzing the one or more images comprising an analysis of a bright image and a fluorescent image.
28. The method of claim 2 , wherein the analyzing the one or more images comprising an analysis of a bright image and a fluorescent image, and wherein the analyzing the one or more images aligns the spatial position of the bright image with the corresponding spatial position of the fluorescent image.
29. An apparatus for assaying an analyte in a sample, comprising:
(a) a sample holder comprising two plates, wherein each of the two plates has a sample contact area, wherein the two plates sandwich the sample between the sample contact areas of the two plates into a thin layer,
(b) a plurality of reference marks on the sample contact areas of the two plates, wherein a priori knowledge of the plurality of the reference marks is known;
(c) an imager that images the sample contact area; and
(d) a non-transitory computer-readable medium storing a machine learning algorithm for analyzing the analyte, wherein the machine learning algorithm utilizes the images of the reference marks.