Methods and systems for image processing of microfluidic devices

• US 8,849,037 B2
• Filed: 11/04/2013
• Issued: 09/30/2014
• Est. Priority Date: 09/13/2006
• Status: Expired due to Fees

• Current Assignee: Fluidigm Corporation
• Original Assignee: Fluidigm Corporation
• Inventors: Dube, Simant, Sun, Gang, Zhao, Lian-She
• Primary Examiner(s): Vanore, David A
• Application Number:
  US14/071,423
• Publication Number:
  US 20140133732A1
• Time in Patent Office: 330 Days
• US Class Current: 382/190
• CPC Class Codes: G01N 21/64 Fluorescence; Phosphorescence
• CPC Class Codes: G06T 2207/10064 Fluorescence image
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/74 involving reference images or patches

Abstract

A method of processing data associated with fluorescent emissions from a microfluidic device. The method includes performing an auto-focus process associated with a first image of the microfluidic device and performing an auto-exposure process associated with the first image of the microfluidic device. The method also includes capturing a plurality of images of the microfluidic device. The plurality of images are associated with a plurality of thermal cycles. The method further includes performing image analysis of the plurality of captured images to determine a series of optical intensities and performing data analysis of the series of optical intensities to provide a series of change in threshold values.

Claims

1. A method of baseline correcting a ratio data curve associated with a PCR process for a microfluidic device, the method comprising:

providing a plurality of baseline values for the ratio data curve;

computing a plurality of baseline removed curves by subtracting the plurality of baseline values from the ratio data curve;

computing a plurality of fitting errors for the plurality of baseline removed curves;

determining a minimum fitting error of the plurality of fitting errors, whereby a first baseline is characterized by having the minimum fitting error of the plurality of fitting errors; and

computing a baseline corrected ratio data curve by removing the first baseline from the ratio data curve.

2. The method of baseline correcting the ratio data curve of claim 1, wherein at least one of the plurality of baseline values is determined by averaging a number of cycles of ratio data.

3. The method of baseline correcting the ratio data curve of claim 2, wherein the number of cycles equals four.

4. The method of baseline correcting the ratio data curve of claim 2, wherein the number of cycles equals a cycle number characterized by an increase of ratio data between a reporter dye signal and a passive reference dye signal exceeding a noise level.

5. The method of baseline correcting the ratio data curve of claim 1, wherein at least one of the plurality of baseline values is determined by a polynomial fit to a predetermined number of cycles of ratio data.

6. The method of baseline correcting the ratio data curve of claim 1, further comprising determining a cycle number where ratio data between a reporter dye signal and a passive reference dye signal first exceeds a noise level.

7. The method of baseline correcting the ratio data curve of claim 1, wherein the plurality of baseline values are provided based on the plurality of baseline values having a fitting error less than a predetermined value.

8. The method of baseline correcting the ratio data curve of claim 1, the method further comprising:

providing a second ratio data curve; and

subtracting the first baseline from the second ratio data curve.

9. A method of baseline correcting a ratio data curve associated with a PCR process for a microfluidic device, the method comprising:

providing the ratio data curve, wherein the ratio data curve provides a value of a reporter dye signal compared to a passive reference dye signal as a function of a PCR cycle number;

computing a plurality of baseline values based on a number of cycles of the ratio data curve;

removing the plurality of baseline values from the ratio data curve to create a plurality of baseline removed curves;

computing a plurality of quality metrics by calculating a quality metric for each of the plurality of baseline removed curves;

selecting a first baseline value corresponding to a highest value of the plurality of quality metrics; and

identifying a first baseline removed curve, of the plurality of baseline removed curves, corresponding to a baseline removed curve that uses the first baseline value.

10. The method of baseline correcting the ratio data curve of claim 9, wherein at least one of the plurality of baseline values is determined by a polynomial fit to the number of cycles of the ratio data curve.

11. The method of baseline correcting the ratio data curve of claim 9, wherein at least one of the plurality of baseline values is determined by averaging the number of cycles of the ratio data curve.

12. The method of baseline correcting the ratio data curve of claim 9, wherein the quality metric for each of the plurality of baseline removed curves is determined by a goodness-of-fit value that ranges between zero and one.

13. The method of baseline correcting the ratio data curve of claim 9, the method further comprising:

providing a second ratio data curve; and

subtracting the first baseline value from the second ratio data curve.

14. The method of baseline correcting the ratio data curve of claim 9, further comprising determining a cycle number where ratio data between the reporter dye signal and the passive reference dye signal first exceeds a noise level.

15. A method of baseline correcting a plurality of ratio data curves associated with PCR processes, the method comprising:

providing the plurality of ratio data curves, wherein the plurality of ratio data curves plot a ratio between a reporter dye signal and a passive reference dye signal as a function of a cycle number in a plurality of PCR process chambers;

providing a plurality of baseline values;

computing a plurality of fitting errors for the plurality of baseline values with one or more of the plurality of ratio data curves;

determining a minimum fitting error of the plurality of fitting errors, whereby a first baseline is characterized by having the minimum fitting error of the plurality of fitting errors; and

removing the first baseline from the plurality of ratio data curves to compute a plurality of baseline corrected ratio data curves.

16. The method of baseline correcting a plurality of ratio data curves of claim 15, wherein the cycle number is determined by a plot location where ratio data between a reporter dye signal and a passive reference dye signal of a first ratio data curve of the plurality of ratio data curves first exceeds a noise level.

17. The method of baseline correcting a plurality of ratio data curves of claim 15, wherein at least one of the plurality of baseline values is determined by averaging a number of cycles of ratio data.

18. A method of baseline correcting a plurality of ratio data curves of claim 17, wherein the number of cycles equals four.

19. The method of baseline correcting a plurality of ratio data curves of claim 15, further comprising computing a plurality of quality metrics, wherein the plurality of quality metrics are based on the plurality of fitting errors, and each of the plurality of quality metrics ranges between zero and one.

20. The method of baseline correcting a plurality of ratio data curves of claim 15, wherein at least one of the plurality of baseline values is determined by a polynomial fit to a number of cycles of ratio data.