Search

US-12626364-B2 - Materials and methods related to image processing

US12626364B2US 12626364 B2US12626364 B2US 12626364B2US-12626364-B2

Abstract

The present disclosure provides materials and methods related to image processing. In particular, the present disclosure provides methods for enhancing target signal detection using imaging processing analysis that identifies and removes non-specific background signals. The image processing methods of the present disclosure are useful for enhancing target signals in a variety of assays that involve fluorescent detection (e.g., fluorescent in situ hybridization, immunofluorescence).

Inventors

  • Ching-Wei Chang
  • Xiao-Jun Ma
  • Han Lu
  • Bing-Qing Zhang
  • HaYeun Ji
  • Ming Yu

Assignees

  • ADVANCED CELL DIAGNOSTICS, INC.

Dates

Publication Date
20260512
Application Date
20220415

Claims (18)

  1. 1 . A method for enhancing detection of a target, the method comprising: imaging a sample comprising a target signal to create a probe image; imaging the sample comprising no target signal to create a background image; modifying the background image to create an adjusted background image based on at least one image metric; subtracting the adjusted background image from the probe image to create a final image comprising an enhanced target signal; determining an estimated location of the target signal in the probe image; and removing the estimated location from the probe image to create a first background-only image and removing the estimated location from the background image to create a second background-only image.
  2. 2 . The method of claim 1 , wherein the method further includes displaying the final image on a display.
  3. 3 . The method of claim 1 , wherein the target signal is obtained by subjecting the sample to a fluorescent in situ hybridization assay and/or an immunofluorescence assay.
  4. 4 . The method of claim 1 , wherein the background image comprising no target signal is obtained by removing the target signal from the sample.
  5. 5 . The method of claim 1 , wherein the at least one image metric is a ratio factor of the first background-only image and the second background-only image.
  6. 6 . The method of claim 5 , wherein the ratio factor is a first intensity to a second intensity, wherein the first intensity is determined from the first background-only image and the second intensity is determined from the second background-only image.
  7. 7 . The method of claim 6 , wherein the first intensity is the mean of a plurality of pixel intensity values in the first background-only image, and wherein the second intensity is the mean of a plurality of pixel intensity values in the second background-only image.
  8. 8 . The method of claim 7 , wherein the first intensity is the mean of all the pixel intensity values in the first background-only image, and wherein the second intensity is the mean of all the pixel intensity values in the second background-only image.
  9. 9 . The method of claim 6 , wherein the first intensity is the median of a plurality of pixel intensity values in the first background-only image, and wherein the second intensity is the median of a plurality of pixel intensity values in the second background-only image.
  10. 10 . The method of claim 6 , wherein the first intensity is the mean of a central 80% of all the pixel intensity values in the first background-only image, and wherein the second intensity is the mean of a central 80% of all the pixel intensity values in the second background-only image.
  11. 11 . The method of claim 5 , wherein modifying the background image to create an adjusted background image includes scaling the background image by the ratio factor.
  12. 12 . The method of claim 1 , wherein the at least one image metric is a multiplication factor to account for local intensity differences.
  13. 13 . The method of claim 1 , wherein the at least one image metric is a local maximum value transform.
  14. 14 . The method of claim 1 , wherein the at least one image metric is a block-matching transform.
  15. 15 . The method of claim 1 , wherein the method further includes registering the probe image and the background image.
  16. 16 . The method of claim 1 , wherein the target signal comprises a fluorescent label bound to a target nucleic acid, or to a target peptide or polypeptide.
  17. 17 . The method of claim 1 , wherein imaging the sample with no target to create the background image is performed after imaging the sample with the target to create the probe image.
  18. 18 . The method of claim 1 , wherein imaging the sample comprising no target to create the background image is performed before imaging the sample with the target to create the probe image.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application claims priority to and the benefit of U.S. Provisional Patent Application No. 63/175,737 filed Apr. 16, 2021, which is incorporated herein by reference in its entirety and for all purposes. FIELD The present disclosure provides materials and methods related to image processing. In particular, the present disclosure provides methods for enhancing target signal detection using imaging processing analysis that identifies and removes non-specific background signals. The image processing methods of the present disclosure are useful for enhancing target signals in a variety of assays that involve fluorescent detection (e.g., fluorescent in situ hybridization, immunofluorescence). BACKGROUND Background fluorescence, such as tissue autofluorescence, is a long-standing challenge in the field of fluorescent-based imaging of cells and tissue samples, including protein detection methods (e.g., immunofluorescence (IF) and RNA detection methods (e.g., fluorescent RNA in situ hybridization (FISH), and co-detection of proteins and RNAs. Such assays utilize fluorescent dyes and fluorescent microscopes for target labeling and detection in the tissue context, respectively. Due to the inherent nature of autofluorescence within tissue samples, it interferes with detection of true signals in IF and FISH assays, especially when the signal of interests are low (e.g., less abundant). Tissue autofluorescence is widely observed in pathology samples prepared by widely used methods, such as in formalin-fixed paraffin-embedded (FFPE) tissue specimens, which is particularly prone to high tissue autofluorescence due to long fixation and processing steps involved. In addition, autofluorescence is present in some fixed frozen and fresh frozen tissues. For example, tissue autofluorescence has been attributed to natural fluorescence that come from endogenous components in cells such as aromatic amino acids, lipopigments, the extracellular matrix components, as well as fluorescence generated during fixation procedures. In FFPE tissues, autofluorescence is observed from lysosomal digestion residues called lipofuscins, the ECM protein collagens and elastins, red blood cells, as well as from formalin fixation. Autofluorescence has broad excitation and emission spectra around the FITC and Cy3 channels, and thus hampers the clear visualization of fluorescent RNA and protein signals in these spectra. Several methods have been developed to mitigate tissue autofluorescence. For example, the difference between signal intensity and autofluorescence background can be experimentally increased by boosting signal intensity (e.g., through the use of tyramide signal amplification), and/or by reducing autofluorescence background while preserving signal. Current commercially available products that employ an autofluorescent reduction strategy include Sudan Black, TrueBlack® and TrueView®. While they all reduce autofluorescence to certain extent, there are significant limitations with these products. For example, Sudan Black produces an undesired shifting of spectrum into the far-red channel; TrueBlack® specifically quenches lipofuscin autofluorescence but is not effective on autofluorescence produced from fixative background. TrueView efficiently quenches all sources of autofluorescence, but significantly dampens target signal intensity. Furthermore, these treatments have significant variation across different tissues, making it difficult to apply a universal and robust autofluorescence blocking procedure. Additionally, post-image processing can be performed to reduce autofluorescence using various methods, including spectral imaging and unmixing or autofluorescence removal software, (e.g., dotdotdot and AFid). However, these methods either require special, high-end microscopic setups or are only effective in removing highly autofluorescent sources such as red blood cells, but not in removing fixative-induced general background. Hence, there is a critical need to develop methods that mitigate the impact of tissue autofluorescence while maintaining the fluorescent signals from desired targets. SUMMARY Embodiments of the present disclosure include a method for enhancing detection of a target. The method includes imaging a sample comprising a target signal to create a probe image and imaging the sample comprising no target signal to create a background image. The method further includes modifying the background image to create an adjusted background image based on at least one image metric, and subtracting the adjusted background image from the probe image to create a final image comprising an enhanced target signal. In some embodiments, the method further includes displaying the final image on a display. In some embodiments, the target signal is obtained by subjecting the sample to a fluorescent in situ hybridization assay and/or an immunofluorescence assay. In some embodiments, the background image compris