Visualizing and sharing CyCIF images

For a ~10 x 11mm tissue FFPE specimen under 20X objective, each cycle of t-CyCIF generates around 160 individual image tiles. The assembly process involves stitching sequential image tiles from a single t-CyCIF cycle into one large image panel, flat-fielding to correct for uneven illumination and registration of images from successive t-CyCIF cycles to each other; these procedures were performed using ImageJ, ASHLAR, and BaSiC software.

The stitched images are stored on Amazon S3 and the zoomable image viewer implemented using OpenSeaDragon.

Generic placeholder image
Melanoma Pre-Cancer Atlas (HTAN)

Unpublished t-CyCIF images of melanoma and precursor fields from two patients. The specimen from patient 1 illustrates different regions of melanoma progression from relatively normal melanocytes to precursor melanocytic dysplasia to invasive melanoma. In addition, the specimen shows different regions representing immune responses to early melanoma (inflammatory regression) as well as to invasive melanoma in the form of a brisk immune infiltrative immune response (tumor infiltrating lymphocytes “TILs”). The excision specimen from patient 2 illustrates the histologic evolution of melanoma from a precursor field to melanoma in situ, and ultimately to polypoidal invasive melanoma.

Generic placeholder image
Tour of the melanoma dataset

The unpublished 13-plex t-CyCIF images of tissue specimens from four patients with BRAF-mutant metastatic melanoma resected before (left) and after (right) treatment with BRAF and MEK inhibitors (dabrafenib/trametinib). Each image is composed of 150-200 image tiles at a nominal resolution of ~0.9 μm.

Generic placeholder image
Figures in the 2018 eLife Publication

Tissue-based cyclic immunofluorescent microscopy (t-CyCIF) is a simple method for generating highly multiplexed optical images from formalin-fixed paraffin-embedded (FFPE) tissue samples routinely used for histopathological diagnosis of human disease. The method is based on previously described single-cell imaging approaches and readily implemented on existing instruments (Gerdes et al. 2013, Lin et al. 2015, 2016).