AI imaging tool tracks immune cells in real time

Immunotherapy is one of the fastest growing fields in cancer medicine. It uses immune cells to locate and destroy tumour cells. To improve these therapies, researchers need to understand precisely how immune cells move, respond and attack. Until now, analysing such imaging data was slow, laborious work prone to human error.

A team of scientists introduced Celldetective, AI-driven software that automatically analyses time-lapse microscopy images. The study was published in the journal eLife. The programme identifies individual cells, tracks their movements, and automatically detects specific events, such as the moment an immune cell kills a target.

What makes it different

Celldetective combines AI segmentation (the automatic delineation of individual cells in an image) with movement tracking over time. No programming knowledge is required: the programme has a graphical interface accessible to non-programmers. Researchers can also train it on their own imaging data.

The software was tested on experiments where immune cells were brought into contact with activating surfaces via so-called bispecific antibodies. Bispecific antibodies are molecules that bind to two different targets simultaneously, effectively directing immune cells toward a tumour.

Broader than cancer research

Although Celldetective was designed for immunology experiments, it works in principle for any biological system involving interacting cell populations. From a longevity perspective, this is relevant: ageing of the immune system contributes to many age-related diseases, and better analytical tools can help researchers understand how immune cells change with age. It represents a methodological step that may accelerate future research.

Read the original article