top of page

Artificial Intelligence for the medicine by AI Power of Audensiel

Indeed, Audensiel has developed AI Power to offer you an innovative and tailor-made approach to capitalize on this opportunity. Fernando C, Senior Partner of Cognodata of Audensiel group, shares with us the new project « In Silico Medicine » and the management of AI in this project. « In-silico Medicine » is a modeling tool for scientifics and biological systems.

Can you tell more about the project « In-silico Medicine ” ?

“In-silico Medicine” corresponds to the computational modeling of biological systems and processes of the human organism with application to the analysis of the evolution, as well as the control, of diseases. In this sense, we complement, for example, test tubes and cell cultures (in vitro) or animal preparations (in vivo) with computational models. In this project, we have also included Machine Learning (ML) so that these computational models are optimal. Thus, we analyze the problem of the regulation of these biological processes as a problem of intelligent adaptive control where intelligent means that the Artificial Intelligence algorithms (AI) and ML are capable of optimizing and estimating a series of parameters. So the regulated system (organisms and systems of the human body) can return to a normal (non-pathological) situation where the natural evolution of the disease is counteracted by mechanisms of internal and external regulation (e.g. treatments).

Within this framework, we have focused on the development of generative Artificial Intelligence systems that allow us to generate and simulate different patterns of the dynamics of an infectious disease. It allows us to analyze under which different conditions different temporal patterns can emerge spatial evolution of a disease. For example, a generative model of the spread of infection by a pathogen (virus, bacteria, etc.) in a patient with pneumonia can explain, under what conditions, the infection spreads through a specific volume of the lungs with an extension and determined propagation speed. Likewise, it also allows us to analyze the evolution of dynamic patterns of the immune response. In this sense, we have successfully carried out R&D in Intelligent Adaptive Control Systems (SCAI) that correspond to the integration of adaptive control theory and the latest artificial intelligence and machine learning techniques for the optimization of these control systems. Thus, for example, in the evolution of the infectious process during pneumonia or in the generation of influenza, we can model on the computer the dynamic process of contagion of pathogens (e.g. viruses and bacteria). It causes the appearance and future dynamic evolution of the population of the pathogenic germ, as well as the dynamic control process of generation of different types of defense mechanisms (e.g. leukocytes). These mechanisms are caused by our own immune system as an immunological response to this external attack.


What is the type of AI and how does this project work ?

This project incorporates several types of AI :

Generative AI allows to generate personalized simulations for each patient´s specific clinical data. Generative mechanistic models based on Control Theory reproduce and generate the illness dynamics.

Schema-based Machine Learning allows the learning of new components based on combinations of previous components as well as parameter optimization for the new components.

Computer Vision allows for disease diagnosis based on patient´s X Ray images.

Multivariable predictive models (such as Random Forest) allow for predictive disease prognosis based on multitude of clinical variables such as oxygen saturation, proteins, etc.


How is AI an added value for this project ?

The Causal Models (Bayesian Nets) allow to estimate the effect of interventions : for instance, the effect of different treatment plans.

Schema-based Machine Learning is able to produce structure learning that allows to uncover the structure of the problem. That is, how it is decomposed in simpler problems, whose partial solutions working together help to solve the larger overall problem.

Structure learning allows to infer underlying  illness and underlying (sub)-phenotipes by uncovering the hidden structure by hypothesizing hidden variables and relations.


How is the team that manages this project made up?

The project team is composed by several AI experts on different AI technologies and several Developers under the direction of the R&D Director. The AI experts are more involved in the functional aspects of the project and work closely with the clinicians. The Developers take the technical design and develop most of the code, the interfaces, the reports, etc.

The project also has a Business Owner which are the clinical partners from a partner hospital in North Spain.


Why do you think it’s important to integrate AI into projects ?

AI is essential to cope with ill-defined problems. That is, we do not know a priori what are the essential aspects and variables to effectively predict the dynamics of the different pathologies. Thus, AI can help us determine what are the hidden causes and the essential parameters for the evolution of different pathologies.

AI also allows to generate personalized dynamic simulations that we can then compare with real patient data.


How do you see the evolution of AI in the coming years ?

AI is currently very strong in Deep Learning and getting in pace with Generative AI since the appearance of ChatGPT. I also see an increase in the use of Reinforcement Learning to cope with problems of Sequential Decision Making, Planning, and Control.

Also, more modern AI paradigms will be getting more and more common such as, Causal Artificial Intelligence and Artificial General Intelligence. Causal Artificial Intelligence allows the measure the effect of different interventions in the system.

Artificial General Intelligence pretends the development of agents that can solve many different tasks in many diverse domains of discourse, similarly to the general human capabilities.

IA Power, Audensiel’s new AI-based program

This is an innovative approach to enable companies to capitalize on this technology. This support, led by Audensiel’s AI expert team, will help develop this evolution and adapt it to the business uses of companies.

Learn more about our AI Power program?👇


bottom of page