Researchers at the Universidad Carlos III de Madrid (UC3M) have developed a system based on computer vision techniques that allows automatic analysis of biomedical videos recorded through microscopy to characterize the behavior of the cells appearing in the images and to describe.
These new techniques, developed by the UC3M engineering team, were used for measurements on living tissue in research carried out with scientists from the National Center for Cardiovascular Research (CNIC in its Spanish acronym). As a result, the team discovered that neutrophils (a type of immune cell) exhibit different behaviors in the blood during inflammatory processes, and identified that one of them, caused by the Fgr molecule, is associated with the development of cardiovascular disease. This work, recently published in the journal nature, could enable the development of new treatments to minimize the consequences of heart attacks. The study involved researchers from the Vithas Foundation, the University of Castilla-La Mancha, the Singapore Agency for Science, Technology and Research (ASTAR) and Harvard University (USA), among others.
“Our contribution consists of the design and development of a fully automated system based on computer vision techniques that will allow us to characterize the cells studied by analyzing videos recorded by biologists using the intravital microscopy technique” , says one of the authors work, Professor Fernando Díaz de María, head of the UC3M Multimedia Processing Group. Automated measurements of the shape, size, movement, and position relative to the blood vessel of a few thousand cells were performed, compared to traditional biological studies that are usually supported by analyzes of a few hundred manually characterized cells. In this way it was possible to perform a more advanced biological analysis with greater statistical significance.
According to the researchers, this new system has several advantages in terms of time and precision. In general, “it is not feasible to have an experienced biologist segment and track cells for months. On the other hand, to give a rough idea (because it depends on the number of cells and the 3D volume depth), our system only takes 15 minutes to analyze a 5-minute video,” says another researcher, Ivan González Díaz , Associate Professor in the Department of Signal Theory and Communications at UC3M.
Deep neural networks, the tools these engineers rely on for cell segmentation and detection, are basically algorithms that learn from example. In order to use the system in a new context, enough examples must be generated to enable their training. These networks are part of machine learning techniques, which in turn are a discipline in the field of artificial intelligence (AI). In addition, the system includes other types of statistical techniques and geometric models, all of which are described in another article recently published in Medical image analysis Diary.
The software that implements the system is versatile and can be adapted to other problems in a few weeks. “In fact, we are already applying it in other scenarios by studying the immunological behavior of T cells and dendritic cells in cancerous tissues. And the preliminary results are promising,” says Miguel Molina Moreno, another researcher on the UC3M team.
In any case, researchers conducting research in this area emphasize the importance of the work of an interdisciplinary team. “In this context, it is important to acknowledge the prior communication efforts between biologists, mathematicians and engineers that are required to understand the fundamental concepts of other disciplines before real progress can be made,” concludes Fernando Díaz de María.
YouTube video: https://youtu.be/EiTAvmQkyIo