Accurate and Efficient Quantification of Fluorescent Images using Unbiased Methods
Background: Immunofluorescent staining offers a range of benefits over conventional immunostaining, including: higher signal: noise (S:N); superior spatial resolution; specific co-labeling of multiple objects in the same sections; and direct correlation between fluorescent intensity and antigen concentration in tissue sections.
With funding from the National Science Foundation (NSF), SRC Biosciences developed the FluoroStereologer, the first approach that combines deep learning, a type of artificial intelligence (AI), with the optical fractionator and other methods of unbiased stereology. The hardware required for FluoroStereologer is a conventional fluorescent (wide-field with deconvolution) or confocal microscope equipped with a high resolution camera and motorized XYZ stage/focus motors.
FluoroStereologer’s novel integration of AI and fluorescent microscopy with unbiased methods represents the new state-of-the-art approach for quantification of the Number, Volume, Surface Area and Length of immunostained cells and other biological objects across the full spectrum of low-to-high resolutions.
How it works:Image Capture (SRC). In step 1, the FluoroStereologer software drives the microscopes hardware to capture high signal: noise (S:N) images of immunofluorescent-stained biological structures (targets, e.g., fluorescent cells) in 8 to 12 sections through any anatomically defined region of interest. These images can be captured at cellular (40x) or subcellular (60-100x) resolution, depending on needs of the study.
AI-basedDeep learning (SRC). In step 2, our histology experts generate a set of “ground truth” images by placing an “x” (i.e., annotating) the targets in the training images. Second, our development team uses these annotated images to train a deep learning “model” to automatically recognize the annotated targets. In the final step, we validate the model by comparing the accuracy of the deep learning data with that collected from the same images using manual stereology. Minimum performance standards relative to manual stereology of the same images are high accuracy (< 5% error); 100% reproducibility (Test/Retest); and high efficiency (10-20x faster).
How it’s used: The end-user collects data with a validated FluoroStereologer model in a two-step process. First, the user captures similarly immunofluorescent-stained images from the same region of interest in their study tissue. In the second step, the validated model quantifies the targets in semi-automatic (with user approval) or fully automatic (without user approval) modes. As a quality control step, the FluoroStereologer allows the end-user to spot check accuracy, precision and efficiency of their results by collecting manual stereology on the same study images.
For more information, check out our blog poston fluorescence and stereology integration.
Reference: Mouton, P.R. (2002) Principles and Practices of Unbiased Stereology: An Introduction For Bioscientists. The Johns Hopkins University Press, Baltimore, MD.
How it works: Image Capture (SRC). In step 1, the FluoroStereologer software drives the microscopes hardware to capture high signal: noise (S:N) images of immunofluorescent-stained biological structures (targets, e.g., fluorescent cells) in 8 to 12 sections through any anatomically defined region of interest. These images can be captured at cellular (40x) or subcellular (60-100x) resolution, depending on needs of the study.
AI-based Deep learning (SRC). In step 2, our histology experts generate a set of “ground truth” images by placing an “x” (i.e., annotating) the targets in the training images. Second, our development team uses these annotated images to train a deep learning “model” to automatically recognize the annotated targets. In the final step, we validate the model by comparing the accuracy of the deep learning data with that collected from the same images using manual stereology. Minimum performance standards relative to manual stereology of the same images are high accuracy (< 5% error); 100% reproducibility (Test/Retest); and high efficiency (10-20x faster).
How it’s used: The end-user collects data with a validated FluoroStereologer model in a two-step process. First, the user captures similarly immunofluorescent-stained images from the same region of interest in their study tissue. In the second step, the validated model quantifies the targets in semi-automatic (with user approval) or fully automatic (without user approval) modes. As a quality control step, the FluoroStereologer allows the end-user to spot check accuracy, precision and efficiency of their results by collecting manual stereology on the same study images.
For more information, check out our blog post on fluorescence and stereology integration.
Reference: Mouton, P.R. (2002) Principles and Practices of Unbiased Stereology: An Introduction For Bioscientists. The Johns Hopkins University Press, Baltimore, MD.