Applied science (1): Quibim, the hidden data after an image

I remember that a few years ago the Nature Neuroscience Magazine revealed in its always interesting Focus section a very significant fact: In the last five years, the number of scientific articles describing new neuroimaging techniques had increased by fifty percent. That editorial of Nature was written in 2013 and since then the development of this field has done nothing but grow, offering almost immediately innumerable applications in research, early diagnosis or more effective treatments for the patient.

Engineers, software developers, doctors or mathematicians are part of numerous research teams that are reaching levels of innovation unthinkable just a few decades ago in the field of image analysis.

In this new series in the Scientific Culture Notebook of the UPV / EHU, I will contact Spanish scientists who are working on internationally recognized projects and whose work can serve as an example of how basic research turns into applied science at the service of society ... The first interviewee is Ángel Alberich, founder of the QUIBIM ( QUantitative Imaging Biomarkers In Medicine ) company, focused on the analysis of biomarkers using image processing techniques.

One of the main reasons to contact Ángel is his recent choice by the prestigious MIT among the Ten Innovators under 35 Spain 2015 in the Spanish version of his Technology Review.

With only 31 years, this young engineer of Telecommunications and Master in Biomedical Engineering, married and with a daughter, manages daily to combine family with his work as scientific-technical director of the Biomedical Research Group in Image of La Hospital Fe de Valencia, as well as the biotechnological image processing company that he founded just over three years ago.

Quibim is a good example of the process that takes us from basic research to the application of this acquired knowledge and its implementation as a service accessible to society.

In the first stage, all the effort falls on the investigation and as the own Angel indicates: "During all that time we were developing a series of algorithms of image processing that we were published by means of scientific articles in specialized magazines. For example, a new method of image analysis to obtain information on cellularity, or a new algorithm to see the resistance of the bone or to detect the fat present in the liver "

Over the years, the results of this research are grouped and organized into a set of tools and algorithms that can be offered as a service for any hospital, doctor or research project that needs to extract certain information from a biomedical image. Thus Quibim was born and " just as clinical analysis laboratories analyze blood, we are dedicated to analyzing images ".

To learn more about his work Angel gives me a very graphic example: Imagine that a doctor observes an MRI of a patient's liver and appreciates a hypointensity. Well, we analyze that image and measure the biomarker you need, as the exact percentage of fat in that liver.

The imaging techniques offered by Quibim are based on two types of modeling, structural and dynamic algorithms. On the one hand, structural modeling allows us to extract volumes, shapes, textures or intensities and quantify these biomarkers very precisely. On the other hand, dynamic models can modify some value or property in the image, such as the passage of time, and in this way, they are able to perform a mathematical model to follow a tumor or any other injury over time.

A few months ago, Quibim's work received the Best Oral Communication Award presented by a partner under the age of 35 at the Congress of the Bone Research and Mineral Metabolism Society (SEIOMM) for the presentation of image processing of the bone structure of bone. a trabecular bone, which you can see in the image above.

To explain it with simpler words: Architects and engineers know well what the risk of a fracture is, this is not only applicable to materials, metals or beams but is equally important in a bone. Currently, the patient only has densitometry (a measurement of density) and little else ... and that is not enough, we need a microstructural study to know exactly how these beams and pillars of the bone are.

The company was founded in 2012 and currently works with doctors, clinics and especially research teams that develop new drugs and that need to extract from these images precise data of the different biomarkers they are working on.