Waiting lists in healthcare: one of the great national challenges to be addressed by IMFD Innovation and Cero.ai
Using the knowledge and advances created in academia to solve problems that affect the People: this is one of the objectives of the Millennium Institute Foundational Research on Data, whose Innovation area, in alliance with Cero.airecently won the project "Prioritization of patients on the waiting list using Artificial Intelligence" of the Ministry of Health.
"The challenge is big: we want to use data science and artificial intelligence tools that are at the cutting edge of research, for a specific problem that affects a significant number of People," stresses Adrian Soto, the director of the Innovation area of the Millennium Institute Foundational Research on Data.
The project focuses on the area of patient prioritization, and consists of three stages: the first one is the diagnosis and evaluation of the information to be used, the quality of the data and the consistency of these data.. "One of the most decisive components when creating artificial intelligence or machine learning models is to have quality data that allows the training of the models," emphasizes the IMFD Innovation Director.
"We hope to take advantage of everything that the Minsal has been building with the Interoperable Waiting Time platforms. We are going to connect to this platform to obtain the information that will allow us to build prioritization models", adds Jorge Pérez, Co-founder of Cero.ai and part of the team that will be working on the project.
As a second stage, the creation of alternative the creation of alternative prioritization modelswith classical supervised learning techniques and modern techniques based on large language models (LLMs), "which we hope to be able to apply, for example, to the recognition and categorization of free text in medical records, in the recognition and categorization of the free text of medical recordsone of the aspects that we know is one of the most complex in the development of these models, since the texts can be irregular and difficult for a model to understand," says Adrián Soto.
After the tests, and as a third stage, the team expects to jointly develop a final prioritization model. to jointly develop a definitive prioritization model, that has the capacity to extract relevant information, while respecting the privacy and security of patient data.
"Since I left my academic position at the University of Chile in 2021, one of my goals has been to collaborate with research centers. Given all the time I was a research associate at IMFD, cooperating with them was the natural step," highlights Jorge Perez of Cero.ai. "My interest is to mix the best of both worlds: to combine the execution capacity that a health-focused startup like Cero has, with the applied research capacity that the IMFD has with the support of its academics and technology transfer department. It is difficult, you have to look for spaces and here with the award we have the opportunity. We are not going to miss it", adds the former researcher, who was one of the first researchers of the Millennium Institute Foundational Research on Data.
For the IMFD, this type of development is very relevant : "Undoubtedly, in Chile we have the capabilities to develop high-level models in artificial intelligence, which can and should respond to the needs of our society", emphasizes Adrián Soto.
About Zero
Zero helps to increase the care capacity of healthcare centers without creating a single new box or hiring a new specialist, coordinating the care of millions of patients every month in public and private care centers. Its solutions blend three aspects: 1) interoperability with hospital and clinic systems, 2) effective communication channels, mainly WhatsApp, and 3) 100% patient-friendliness, for a smooth and comfortable patient experience. As a startup Cero went through the Y-Combinator accelerator program in Silicon Valley, responsible for accelerating companies such as Airbnb, Dropbox, Rappi and Fintual. For more information, visit www.cero.ai.