Fibrosis Image Quantification Guide using ImageJ: Protocols for Masson's Trichrome and IHC

1. Introduction: Why is Image Analysis Necessary?

In fibrosis research, performing "bulk biochemical evaluations" like the hydroxyproline assay or ELISA is not enough; "histomorphological evaluation of spatial information" is absolutely critical. Proving whether fibrosis is occurring perivascularly, spreading through the interstitium, or causing severe structural remodeling (as assessed by the Ashcroft Score) can exclusively be achieved through image analysis.

Historically, pathological evaluations relied on "scoring" (e.g., semi-quantitative 0-4 scales), which suffered from high subjectivity and poor reproducibility. Today, the calculation of objective, continuous quantitative values (Area Fraction: %) using software like ImageJ (specifically the Fiji distribution) has become the gold standard for journal submissions and preclinical efficacy data.

This article provides step-by-step practical ImageJ analysis protocols for the staining methods most frequently used in fibrosis evaluation.

2. Preparation: Installing Fiji (ImageJ) and Basic Setup

It is strongly recommended to use Fiji (Fiji is just ImageJ), a distribution that comes bundled with many essential plugins pre-installed.

1. Download Fiji: Download the appropriate version for your OS from the official website (https://imagej.net/software/fiji/).
2. Prepare Images: Prepare your images acquired via digital slide scanners (NDPI, SVS, etc.) or microscope cameras (saved as uncompressed TIFF or high-quality JPG).
   • Crucial Note: Ensure that uneven illumination (shading correction) and white balance are properly adjusted during the image acquisition phase. This directly impacts the reproducibility of your quantification.

3. Protocol A: Masson's Trichrome (MT) Stain Quantification

In Masson's Trichrome staining, collagen fibers stain blue (or green), muscle fibers and cytoplasm stain red, and nuclei are stained dark purple/black. The analytical challenge is forcing the software to correctly separate the target blue collagen from the red cytoplasm and white background.

Step 1: Specific Dye Extraction via Color Deconvolution

Accurately thresholding "only the blue" from a full-color RGB image stringently is difficult. Thus, we use the standard Fiji plugin Color Deconvolution.

1. Open the image out, and navigate to Image > Color > Color Deconvolution.
2. From the Vectors drop-down menu, select "Masson Trichrome" (or a similar vector).
   • Note: If your slides have unique tinting due to different dye batches, it is better to create a custom vector by selecting representative ROIs (Regions of Interest) for blue, red, and background.
3. Three new 8-bit grayscale images (Colour_1, 2, 3) are generated. Keep only the image where the "blue component (collagen)" appears as dark/black pixels, and close the others.

Step 2: Binarization via Thresholding

Convert the extracted grayscale image into a pure black-and-white binary image isolating collagen (positive area) from the background.

1. Select the target image and go to Image > Adjust > Threshold (Ctrl+Shift+T).
2. Adjust the sliders while observing the histogram so that only the collagen structures are highlighted in red.
3. It is effective to test Auto-threshold algorithms (e.g., Huang or Otsu). Within the same project, systematically apply the exact same threshold algorithm (or fixed numerical limits) across all samples to eliminate subjective bias.
4. Click "Apply" to create the final binary mask (values of 0 and 255).

Step 3: Measuring the Area Fraction

1. Navigate to Analyze > Set Measurements, and check the boxes for "Area", "Area Fraction", and "Limit to threshold".
2. Run Analyze > Measure (Ctrl+M).
3. The Results window will display the %Area. This value represents the proportion of the fibrotic area (blue collagen) relative to the entire tissue area in the image.

4. Protocol B: Polarized Image Quantification of Picrosirius Red (PSR)

While Sirius Red staining stains collagen red under brightfield, when viewed under a Polarized light microscope, collagen exhibits birefringence glowing red, yellow, or green depending on fiber thickness and maturity. Because the glowing collagen appears vividly against a pitch-black background, this method is exceptionally well-suited for automated digital quantification.

1. Grayscale Conversion: Open the image and convert it via Image > Type > 8-bit.
2. Thresholding: Go to Image > Adjust > Threshold to filter out the black background noise, isolating only the brightly glowing collagen fibers.
3. Measurement: Execute Measure identically to Protocol A to secure the %Area.

• Advanced: If you wish to quantify Type I (thick, red-yellow birefringence) and Type III (thin, green birefringence) separately, apply an RGB Hue threshold before 8-bit conversion to extract only the pixels falling within those defined color spectrums.

5. Protocol C: DAB Quantification in Immunohistochemistry (IHC)

This protocol is used for slides stained using enzyme-labeled antibodies (DAB chromogen producing brown precipitants) and counterstained with hematoxylin (blue nuclei)—for example, detecting α-SMA (activated myofibroblast marker) or specific macrophage populations.

1. Color Separation: Go to Image > Color > Color Deconvolution and select "H DAB" in the Vectors menu.
2. From the three generated images, select the one corresponding to the "DAB (Brown)" signal.
3. Proceed identically with Thresholding → Measuring to output the DAB positive area fraction.

6. Automation and High-Throughput: Utilizing Macros

Manually opening, separating, thresholding, and measuring dozens or hundreds of images one by one is impractical and invites human error. It's vital to exploit ImageJ's powerful Macro (automated scripting) capabilities.

By navigating to Plugins > Macros > Record..., you can perform the workflow (Deconvolution → Threshold → Measure) on a single image, and Fiji will record these actions as script language. By incorporating this into Process > Batch > Macro, you can instantaneously apply identical analysis logic to entire folders of images seamlessly.

[!TIP] Advancing with CRO Outsourcing Simple thresholding-based ImageJ analysis carries the risk of overestimating fibrosis by inadvertently counting "artifacts (tissue folds, bubbles)" or "healthy perivascular collagen that shouldn't be counted as pathological interstitial fibrosis." Recently, modern CRO services utilize AI (Deep Learning) models to provide "AI Pathology." These advanced pipelines automatically exclude artifacts and selectively segment the pure interstitial spaces, yielding vastly more accurate and contextual fibrosis quantifications.

7. Conclusion

Fibrosis image quantification via ImageJ (Fiji) remains a robust methodology universally accepted in scientific literature. "Accurate color separation via Color Deconvolution" and "consistent thresholding criteria" are the unshakeable pillars of reliable data acquisition. By corroborating morphological/spatial data (%Area from ImageJ) with biochemical total collagen data (e.g., Hydroxyproline assays), researchers can curate profoundly convincing proofs of therapeutic efficacy.