Visiopharm Partnership

AI-Driven Precision Pathology Meets Spatial Biology

NanoString has partnered with Visiopharm® A/S, a leader in the digital pathology space, to automate image analysis and Region of Interest (ROI) selection for the GeoMx® Digital Spatial Profiler. GeoMx customers can leverage Oncotopix® Discovery AI-driven software to enhance tissue segmentation, speed up the ROI selection process, and overall reduce pathologist time.

This integration of spatial multiomics with AI image analysis enables:

  • Improved accuracy for biomarker discovery
  • The ability to understand disease better
  • Enhanced and more accurate predictive and prognostic biomarkers
  • Improved success in drug discovery and development
  • Generation of new therapeutic hypotheses
  • Scale up to large-cohort spatial studies

Oncotopix Discovery for GeoMx DSP

Oncotopix Discovery for GeoMx DSP includes Visiopharm’s Oncotopix Discovery base package bundled with the Tissuealign™ module and heatmapping and is purchased separately from the GeoMx DSP. This combination enables you to:

  • Visualize tissue architecture and hot spots to select ROIs, then further segment ROIs into one or more Areas of Illumination (AOIs) based on staining patterns, intensity gradients, and/or cell types.
  • Leverage serial section images of other modalities
  • ‘Train’ a modular analysis unit (an APP) within Oncotopix Discovery to recognize specific ROIs and run your APP on one or more images once it’s trained.

Together, these tools enhance spatial multiomic studies and reduce hands-on time for GeoMx experiments.

Paint-to-Train AI
A Pathologist in Your Pocket
Enhanced Segmentation
Easy Image Overlay
Workflow Automation
Image Metrics
Paint-to-Train AI
Oncotopix Discovery provides an easy-to-use AI toolset to create image analysis modules (APPs) to find and measure what you are looking for within a tissue section.
  • If you can see it, you can train it – No need for image analysis knowledge.
  • Anyone with biological understanding can train a deep learning APP
  • Include and exclude different tissue types in your analysis and replace manual annotation steps
  • Improve the consistency of your results
A Pathologist in Your Pocket

Supercharge your spatial multiomics studies on GeoMx DSP with the power of automated tissue markup and ROI selection using Oncotopix Discovery:

  • Use an H&E and/or immunofluorescence (IF) stained image to mark up the tissue and identify functional structures and cell types, making ROI selection simpler and faster.
  • Overlay the marked-up image with the GeoMx image for rapid ROI transfer using the Tissuealign module.
  • Set up rules for ROI selection that allow you to cut down on analysis time and automate your GeoMx studies.
Oncotopix Discovery marks up the different functional compartments of a head and neck cancer FFPE tissue section from an H&E-stained image. Image courtesy of Arutha Kulasinghe, PhD, University
of Queensland.
Enhanced Segmentation

Oncotopix Discovery can enhance enrichment for GeoMx profiling to get a cleaner profiling signature from selected tissue compartments.

Exclusion of Unwanted Regions

Train deep learning algorithms to exclude cell-types or tissue regions that are unwanted, such as autofluorescent cells, regions of necrosis, blood vessels, etc.

Selection of cells by punctate staining

Select cells stained with RNAscope ISH or DNA FISH probes for deep exploration of expression in the microbiome, viral infected tissue, or the tumor microenvironment. Additional measurements of ISH spots can be made to quantify signals in profiled areas.

Contouring

Capture signaling gradients by creating profiling contours around structures, individual cells, or perturbation points. For sparse cell types, an annular ring around the sparse cells can serve as a local background measurement to more clearly detect expression of sparse cells in the tissue.

 

Easy Image Overlay
Use the Tissuealign module to mark key tissue landmarks to register H&E or other images with the GeoMx image for simple ROI transfer and multimodal image analysis. Images do not need to be from the same instrument or at the same magnification.
Workflow Automation
Use APPs to Segment the tissue based on expression patterns, then automate the placement of ROIs using desired rules such as >90% tumor, 50% tumor and 50% stroma, 100% stroma. Use Deep Learning apps to handle intensity variability across tissues, minimizing time during the GeoMx experiment the adjust thresholds. Deploy APPs across multi-site studies to improve reproducibility and minimize ramp up time.
Image Metrics
Phenotype and Quantify Cells in the Areas of Illumination
Perform whole slide analysis
Compare with H&E
Export tables of measurements

How It Works

Integrate digital pathology into your GeoMx workflow by importing the GeoMx IF stained image files (OME.TIFF) into Oncotopix Discovery, processing the images using a Visiopharm APP of your choice to select ROIs and define AOIs, and then loading the exported OME.XML files back into the GeoMx Control Center for barcode collection and subsequent counting using the nCounter Analysis System or an Illumina sequencer. Downstream expression analysis is done in the DSP Data Analysis Center or open-source tools while further image analysis such as cell counting and tissue phenotyping can be done using Oncotopix Discovery.

APP Center

Visit the Visiopharm APP Center to find pre-trained analysis protocols that can be used as templates to begin image analysis.  Users can get started with a proprietary GeoMx Epithelial Detection App to use as a starting point for segmenting Epithelial and Stromal compartments for samples stained using the GeoMx standard Solid Tumor Morphology Marker Kit, ultilize a world-class fluorescent nuclear detection algorithm, or choose from among 120+ other Visiopharm APPs to reduce your development time.

Request More Information

Contact our helpful experts and we’ll be in touch soon.

Related Resources

Product Bulletin GeoMx Human Whole Transcriptome Atlas – Product Bulletin
Product Bulletin GeoMx® IO Proteome Atlas Product Bulletin
Poster SITC 2023: Automating a spatial profiling workflow to explore the effects of hypoxia in the tumor microenvironment in head and neck…
Poster AACR 2023: Spatial profiling of the tumor microenvironment in head and neck cancer identifies immune checkpoints and proteins involved in cell…
Poster AACR 2022 Poster: High-plex spatial proteomic profiling of immunotherapy response groups in head and neck cancer identifies tissues signatures associated with…