Contact Us

Transforming Multiplexed 2D Data into Spatial Insights Guided by AI

AI-based image analysis has significantly progressed in deep learning, segmentation, and detection, making it a powerful tool for gaining critical insights

2D slice of colon cancer tissue stained with 30 markers and imaged using the Cell DIVE system. Analysis performed using Aivia 13’s new multiplex cell detection recipe and automatic clustering tool. Each phenotype denoted in a different color. 2D_slice_of_colon_cancer_tissue_stained_with_30_markers_Aivia_Cell_DIVE.jpg

Generating Insights into tissue microenvironments is crucial to our understanding of normal and abnormal tissue development such as during cancer progression. Spatial biology methods using microscopy and multiplexing techniques have allowed researchers to bring together protein expression profile and spatial information to dive into the complex biology needed for these insights. However, image analysis of these large multiplexed 2D images remains a challenge.

About this webinar

Key Learnings

In this webinar our presenter, Quyen Tran, PhD, highlights new and improved AI features available in our latest release of Aivia. Aivia 13 provides a toolkit for transforming multiplexed 2D data into spatial insights.

  • How to accurately segment cells so you can spend less time on analyzing data and more time on the biology
  • Automatic and AI-guided tools to cluster your cells into phenotypes of interest to get a clearer understanding of your tissue 
  • Ways to dive deeper into the biology with spatial relational tools and various charts such as dimensionality reduction plot and marker-cluster dendrogram

Get more insights from large 2D images

In Aivia 13, significant improvements have been made to the visualization engine. This allows users to view large, multiplexed 2D images (85K x 57K in size with 30 channels) and millions of detected objects. The new Mini-Map feature helps researchers navigate to their region of interest from anywhere in their image. Automatic tissue detection enables users to automatically delineate Regions of Interest (ROI) based on any channels for efficient downstream analyses of large tissue sections and Tissue Micro Array (TMA) cores.

Additionally, Aivia offers two new recipes for cell detection: Cell Analysis - Cellpose and Multiplexed Cell Detection, to provide an accurate understanding of your data. Both recipes incorporate an optimized version of the Cellpose generalist deep learning cell segmentation algorithm [1], which has been demonstrated to accurately segment cells with diverse morphological characteristics.

Expert- and data-driven phenotyping

Understanding the types of cells in your image is critical to gaining deep insights. Aivia’s  AI-driven Phenotyper leverages your expert knowledge of the phenotypes present in your data to build a classifier. Simply select a few representative cells to train Aivia and Aivia will classify cells into different phenotypes automatically.

Aivia has two automatic clustering choices, k-means clustering [2] and PhenoGraph-Leiden [3] unsupervised automatic clustering methods, which only need a few user-guided inputs. These methods can generate traits that include intensity and morphological measurements for up to 30 biomarkers, aiding in data-driven phenotyping.

Complex Spatial Data Exploration

When working with millions of objects in large, multiplexed 2D images, it's essential to conduct a thorough quantitative analysis of various entities and characteristics, and to juxtapose them for comparison. Aivia now provides new charts optimized for data exploration of millions of objects.

In this new release of Aivia, an overview of data is generated for different hierarchical levels, which enables users to analyze the measurements for each phenotype and compare them across phenotypes. Dimensionality reduction is a mathematical method to gain better understanding and insights into data with a high number of dimensions by reducing them to two dimensions. Aivia 13 provides three unique methods for executing dimensionality reduction, including UMAP, PaCMAP, and t-SNE. In addition, Aivia allows users to observe their clusters or phenotypes in the plot of dimensionality reduction.

Want to learn more? Register below to watch the full webinar presented by Quyen Tran, Ph.D.

Watch the webinar

By clicking on the SUBMIT button, I confirm that I have reviewed and agree with Leica Microsystems GmbH Terms of Use and their Privacy Policy. I also understand my privacy choices as they pertain to my personal data, as detailed in the aforementioned Privacy Policy under ‘’Your Privacy Choices’’.

Scroll to top