Segmentation-free analysis of live-cell imaging data reveals how T cell modifications influence cancer cell aggregation dynamics

2026-06-30 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Life Sciences & Biology, Mathematics & Computational Sciences, Health & Medical Research · Depth: Expert, short

Summary

A new unsupervised analysis workflow, segmentation-free live-cell behavioral analysis (SF-LCBA), has been developed to characterize live-cell imaging (LCI) data generated using the Incucyte imaging platform. This method bypasses traditional cell segmentation, which is often unsuitable for LCI videos due to low spatiotemporal resolution and extensive cell-cell contact. SF-LCBA instead focuses on identifying global aggregation patterns and local cellular keypoints to analyze multicellular interactions. The workflow was applied to TCR T cells from four donors, varying RASA2 knockout proportions, and different effector-to-target concentrations in co-culture with A375 melanoma cells. Results indicate that T cell modifications significantly impact the spatiotemporal dynamics of multicellular aggregate formation. Specifically, higher proportions of T cells with the beneficial RASA2 gene knockout resulted in fewer and smaller cancer cell aggregates. This approach offers more therapeutically-relevant measurements of T cell therapy behavioral phenotypes.

Key takeaway

For researchers analyzing T cell anti-cancer function from live-cell imaging, you should consider adopting segmentation-free live-cell behavioral analysis (SF-LCBA). This method overcomes limitations of traditional segmentation in low-resolution, high-contact videos, providing more accurate insights into multicellular aggregation dynamics. You can use the publicly available SF-LCBA software package. This will help characterize T cell modifications and their influence on cancer cell aggregation, leading to therapeutically-relevant measurements.

Key insights

SF-LCBA enables robust analysis of T cell-cancer cell interactions in low-resolution LCI data without segmentation.

Principles

• Cell segmentation is often unsuitable for dense LCI videos.
• Global aggregation patterns reveal T cell anti-cancer function.
• T cell modifications alter cancer cell aggregation dynamics.

Method

Develop methods to identify global aggregation patterns and local cellular keypoints, avoiding cell segmentation for LCI videos.

In practice

• Analyze LCI data with high cell-cell contact.
• Quantify T cell anti-cancer function from aggregation.
• Track aggregate formation in challenging datasets.

Topics

• Live-cell Imaging
• T Cell Therapy
• Cancer Immunotherapy
• Image Analysis
• Segmentation-free Analysis
• RASA2 Gene Knockout
• A375 Melanoma Cells

Code references

• bee-hive/sflcba

Best for: Computer Vision Engineer, AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• Non-invasive profiling of the tumour microenvironment with spatial ecotypes — A new machine learning framework enables non-invasive profiling of tumor microenvironment spatial ecotypes from liquid biopsies, predicting immunotherapy response.
• Evidential learning driven Breast Tumor Segmentation with Stage-divided Vision-Language Interaction — TextBCS enhances breast tumor segmentation in MRI by combining text prompts with evidential learning to address low contrast and blurred boundaries.
• Computational Methods and Challenges in Cell-Free DNA Analysis for Multi-Cancer Early Detection — Multimodal ensemble approaches for cell-free DNA analysis offer the strongest promise for multi-cancer early detection clinical integration.
• Querying Counterfactuals on Tissue Graphs with Supervised Disentanglement — Cellina disentangles intrinsic cell state from spatial context to predict counterfactual cell expression on tissue graphs.
• Label-free multiphoton microscopy and machine learning for recognition of hepatocellular carcinoma — Label-free multiphoton microscopy combined with machine learning accurately identifies hepatocellular carcinoma.
• Diffusion Attention Expert Model for Predicting and Semi-automatic Localizing STAS in Lung Cancer Histopathological Images — DAEM accurately detects STAS in lung cancer histopathology, offering strong generalizability and TME-based localization.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.