Artificial intelligence comes to the rescue in cancer research

Thanks to advances in artificial intelligence, great strides are being made in medicine in the (early) analysis of diseases. Christophe Deben and Edgar Cardenas, working at the Center for Oncological Research (CORE), developed an image-based analysis technique for 2D and 3D cancer cell cultures using artificial intelligence. During LabAutomation, Deben provided more insight into the operation of this platform.

 Smart software program for cancer cell analysis

The platform is called OrBITS, which stands for Organoid Brightfield Identification-based Therapy Screening. The software works together with the Tecan Spark Cyto multimode plate reader, in combination with a high-throughput, 384-well microtiter plate. Deben: “OrBITS is essentially a smart software program for analyzing cancer cells. The growth percentages of the cancer cell cultures can be followed via automated image processing.”

He continues: “The pictures are taken by microscopes that can photograph the kinetic growth of cancer cells. Over the years, you see the switch from 2D to 3D models for cancer cell cultures, where the cells grow in a gel as small 3D tumors. We do research on cancer, but the technology can also be used for other research, such as 'organs in a dish'.”

Treatment tailored to the patient

OrBITS should also enable better, patient-tailored treatments for different types of cancer: “We screen different therapies to see if there is a therapy response to a certain treatment,” Deben explains. “The associated organoids come from patients who have had a biopsy. So the goal is to test therapies in the lab to see which ones could be used in a patient.”

Artificial intelligence and algorithms

Artificial intelligence plays an important role in the software: “The OrBITS software contains an algorithm to analyze the images taken with a web interface on the front end. The experiments sometimes analyze many gigabytes of data. That is why we also provide data compression with OrBITS, but also storage in a cloud storage system. We are also working on converting the raw data into readable graphs and tables.” We will also make the software compatible with the most common microscopes so that researchers have access to advanced analysis software, without having to invest in new infrastructure.”

During LabAutomation, Deben gave a demonstration of the OrBITS platform. He also hopes to be able to launch the platform on the market this year: “We want to launch OrBITS on the market later this year. We are investigating whether we are going to set up another company for that. And there is already interest from the lab world, for example from customers of Tecan because we have already tuned the software to their device (Spark Cyto). We also want to send test versions to a number of collaborating labs.”