Artificial Intelligence

Is there any mechanical ventilator using Artificial Intelligence?

Artificial intelligence applied to mechanical ventilators.

The search to create artificially intelligent (AI) computer systems has been one of the most ambitious and controversial. It also seems that very early on, scientists and doctors were captivated by the potential such a technology might have in medicine. With computers and software able to store and process vast stores of information, the hope was that they would become a system that gives a doctor’s answer, assisting or surpassing clinicians with tasks like diagnosis ¹ .

With these motivations, a small but talented community of computer scientists and healthcare professionals set about shaping a research program for a new discipline called Artificial Intelligence in Medicine (AIM). These researchers had a bold vision of the way AI would change medicine, and move forward the knowledge of technology. AI in medicine at that time was a largely US-based research community. Work originated out of a number of campuses, including MIT-Tufts, Pittsburgh, Stanford and Rutgers. The field attracted many of the best computer scientists and, by any measure; their output in the first decade of the field remains a remarkable achievement¹.

One of the first definitions used by this group was:

Medical artificial intelligence is primarily concerned with the construction of AI programs that perform diagnosis and make therapy recommendations. Unlike medical applications based on other programming methods, such as purely statistical and probabilistic methods, medical AI programs are based on symbolic models of disease entities and their relationship to patient factors and clinical manifestations.’

A lot has changed since then, and today this definition would be considered narrow in scope and vision. Today, the importance of diagnosis as a task requiring computer support in routine clinical situations receives much less emphasis² .  So, despite the focus of much early research on understanding and supporting the clinical encounter, expert systems today are more likely to be found used in clinical laboratories and educational settings, for clinical surveillance, or in data-rich areas like the intensive care setting. For its day, however, the vision captured in this definition of AIM was revolutionary.

After the first euphoria surrounding the promise of artificially intelligent diagnostic, programmers in the 80’s and 90’s have seen an increase in disillusion amongst many with the potential for such systems. Much of the difficulty has been the poor way in which they have fitted into clinical practice, either solving problems that were not perceived to be an issue, or imposing changes in the way clinicians worked. Regarding AI in mechanical ventilation, nothing has being reported so far. There is a company that offer as part of their portfolio software called “Intelligent solutions for mechanical ventilation”; however those are not more than marketing tools. After a small analysis anyone can tell that is only a system based in old fashion equations from 1920 (Otis equation). That only creates confusion into the medical market assuming that the product adds AI to the system when it only has a closed loop system as other companies may have (for example CLIO of Vyaire).

What is now being realized is that when Artificial intelligence fills an appropriate role, intelligent programmers could indeed offer significant benefits. One of the most important tasks now facing software engineers of AI-based systems is to characterize accurately those aspects of mechanical ventilation that are best suited for the introduction of artificial intelligence systems. AI-based systems could help nowadays when microprocessors can execute actions as fast as it is needed to make a decision in real time. That means that software could be added to medical equipment in order to help in situations where the user does not have the time or knowledge to make a good decision.

As an example, we could have software that would help a doctor to understand how the respiratory mechanics of a patient is, or to receive recommendations from the ventilator to set it according to the patient needs. Unfortunately even the fact that today we have the technology to move forward to use AI, most of the regulatory agencies (such us FDA in USA, Cofepris in Colombia or ANVISA in Brazil) do not have the structure to analyze if an application would improve the current state ventilated patients at the ICU.

Without any doubt we need changes not only in the medical device industry and regulatory agencies structure but also in the market. We already could have tools using AI for several medical tasks however we are not prepared to accept this new technology.

Tendencias: ¿Cómo se puede mejorar el tratamiento de pacientes en una terapia intensiva?

¿Se puede reducir el espacio muerto en pacientes ventilados?

Neonatología: Protección Pulmonar. Una herramienta simple para un problema complejo.

  1. Guide to Medical Informatics, the Internet and Telemedicine, Chapter 19, Enrico Coiera
  2. Durinck et al., 1994
Facundo Guillermo Bermejo

Facundo Guillermo Bermejo

Biomedical Engineer and Magister in Software Engineering (specialist in software management, expert systems and neural networks) with more than 20 years of experience in development and commercialization of mechanical ventilators

More Posts

Follow Me:

Leave a Reply