Multiplexed imaging generates rich datasets where extracting meaningful insights requires more than predefined gates.
In this webinar, Indica Labs presents a real-world use case analyzing tertiary lymphoid structures (TLS) in non-small cell lung cancer (NSCLC) using whole-slide multiplex immunofluorescence images acquired on the Orion™ platform (RareCyte). Dr. Ghislaine Lioux will demonstrate how standard HALO® workflows, including AI-driven tissue classification and cell segmentation, can be seamlessly combined with unbiased high-dimensional analysis, all within the same software.
Using K-means, Phenograph, and UMAP, the demonstration will explore phenotypic diversity within TLS from dense tumor regions and highly infiltrated zones. HALO® High Dimensional Analysis reveals biologically relevant differences between these zones beyond supervised phenotypes. Importantly, the dynamic link between UMAP plots and the original tissue image provides biological meaning to clustering results and enables clear spatial validation.
Learning Objectives
Assess sample cell composition “at a glance” using K-means and Phenograph.
Use unsupervised analysis to quality control your analysis pipeline.
Navigate interactively between the UMAP and biological image.
Apply an iterative, data-driven strategy to identifyand discover biologically relevant phenotypes.
Presenter
Dr. Ghislaine Lioux Senior Field Applications Scientist Indica Labs
Dr. Ghislaine Lioux earned her PhD from the Spanish National Institute of Cardiovascular Research in Madrid as a Marie Curie Fellow, where she studied molecular mechanisms of heart development.Ghislaine joined Indica Labs in 2020, and in her current position as a Senior Field Applications Scientist, shesupports customers, distributors, and collaborators on the technical use of HALO software.
The latest from the Proscia blog: By Ritida Nanda | March 12, 2026 Proscia has been named Global 2026 Best in KLAS for Digital Pathology in Europe and earned the top performance score in the U.S. in the 2026 KLAS Digital...