Scientific research has been a field that benefited from the enhancements in AI. AI’s Deep Learning and Machine Learning algorithms are being used by researchers for the fast and accurate derivation of results. The recent development in the field of computer vision is ensuring a stressless study of cells, which free the biologists from month-long segmentation and tracking of cells on a video. The researchers at the Universidad Carlos III de Madrid (UC3M) developed a model that enables the automatic analysis of biomedical videos captured by microscopy. This system can be used to characterise and describe the behaviour of the cells in the image. Compared to the conventional manual method, this computer vision-based model is much more precise.

The above image shows the Neutrophil segmentation carried out using the ACME software. The segmentation is a 3D process. However, the image has an accumulated 2D version. Towards the left is the original microscopy image which has blood vessels (green), and neutrophils (red). On the right, segmentation using ACME is shown (one colour for each cell).

Result of the research

The new technique invented by the UC3M team has been used for measurements of living tissues. The research on the tissues was carried out by scientists from National Centre for Cardiovascular Research. Through the research, they came to know that a type of immune cell called neutrophils express different behaviour in blood while undergoing inflammatory processes. One of the cells caused by the Fgr molecule can cause the development of cardiovascular diseases. This work was recently published in the journal Nature. It was stated in the journal that this could come up with new treatments to minimise the consequences of heart attacks. There were researchers from the Vithas Foundation, the University of Castilla-La Mancha, the Singapore Agency for Science, Technology and Research (ASTAR) and Harvard University (U.S.), among other centres, who participated in the study.

According to Prof. Fernando Díaz de María, the head of the UC3M Multimedia Group, their contributions includes the design and development of a fully automatic system. He stated that the study which is based on Computer Vision techniques allows the researchers to characterise the cells under study, “by analysing videos captured by biologists using the intravital microscopy technique”. Using the technique automatic shape, size, movement and position relative to the blood vessel of a few thousand cells have been made. When compared with the conventional studies, more advanced analysis can be done with the technique with greater statistical importance.

Benefits of the Model

Reduced time consumed is one of the biggest benefits of this model. “our system only takes 15 minutes to analyse a 5-minute video”, stated Ivan González Díaz, Associate Professor in the Signal Theory and Communications Department at UC3M. The tools that the researchers primarily use are deep neural networks. These are algorithms that learn using examples. For the effective functioning of the system, sufficient examples must be generated for the machine to learn. The researchers stated that they have applied the technique in various other scenarios such as for studying the immunological behaviour of T cells and dendritic cells in cancerous tissues. The initial results are said to be promising according to the results. The team working on the project, can not emphasise enough on the importance of an interdisciplinary team. They regard that they need to communicate with mathematicians and engineers to understand the basic concepts of other disciplines before the real progress.