8 minute read

Interview to Laura Velásquez Herrera

America is Innovation; Colombian company develops a sophisticated algorithm for the early detection of diseases

Early diagnosis of various diseases potentially decreases their mortality rate by increasing the chances of success of their treatments; in addition, it increases the quality of life of patients and the economy of the population in which they reside. Currently, in Latin America, we work with technology based on Artificial Intelligence for the early detection of certain diseases that afflict thousands of people around the world. Therefore, on this occasion, we invite Laura Velásquez Herrera, president and co-founder of Arkangel AI, who shares with us the development and impact of her platform and its potential influence on more technologies used for healthcare.

Figure: Detección de retinopatía diabética.

Figure: Detección de retinopatía diabética.

Courtesy: Arkangel AI

Dear Laura Velásquez, could you explain what Arkangel AI is?

«Arkangel AI is a platform that automatically transforms medical data into predictive models; these types of data can be images, tabulated data, voice, video, or any other kind. It is an Ai as a service platform, where any company in the health sector, from pharmaceuticals, insurers, hospitals, laboratories, etc., can access the power that artificial intelligence brings without needing to know about data science or artificial intelligence itself.

We have currently developed predictive models to make early detection of diseases such as leukemia, other types of cancer, and respiratory diseases, among others. However, we are also developing models to identify potential claims or high-cost accounts for an insurer, know if there is clinical deterioration in a hospital, or create models to increase access to treatments for pharmaceutical companies, etc.

In short, we transform that data into AI models to make it a digital asset for companies. We take care of the whole process by having an obvious problem you want to solve. Today we focus a lot on the early detection of diseases, but it can be used for many possibilities and challenges in the health system.»

In simple terms, how does the algorithm they use work?

«It is straightforward, if an IPS (hospital), an insurer, a pharmaceutical or a laboratory, even in government stays, have a problem within their processes in any area of the institution, or want to optimize or automate a process, AI could be implemented. What we do is understand what that problem is, see how it is currently solved, and from there, we analyze what kind of data is required. Once we have that data, it is entered into our platform and automatically starts with the algorithm’s training. When it is ready, the system notifies you through an alert email that has already been created for review, where you see the accu-racy metrics, which we seek always to be higher than 90%, and from there to be able to deploy it in the institutions’ systems to start using it in the field.»

Figura: Detección de patología de rayos X de tórax.

Figura: Detección de patología de rayos X de tórax.

Cortesía: Arkangel AI.

What kind of diseases can Arkangel AI detect?

«Let’s say it depends on the predictive models that institutions want to create; today, we have made predictive models for more than 19 respiratory pathologies, we are doing for colon and lung cancer, we have already done one with UNICEF of leukemia in children, also for malaria and hospital clinical deterioration, that is, what probability there is that a patient who arrives at the emergency room may require hospitalization or transfer to an intensive care unit and even what mortality rate exists in that patient so that doctors can take better decisions in real-time, and so on.»

Can it be categorized as a personalized process?

«Yes, it is at the same time a customizable and scalable process; each institution has its challenges, and we want them to be able to implement this technology to benefit patients and institutions simultaneously.»

So far, where has this technology been implemented?

«Today, we have alliances in the United States, England, Spain, and Colombia, with allies who can deploy our technology in their countries and create new models for their challenges. We currently work with different health institutions such as pharmaceutical companies, insurance companies, hospitals, and large health organizations worldwide such as UNICEF, AstraZeneca, Novartis, Boston Scientific, Abbott, AWS, and Fundación Santa Fe de Bogotá, among others.»

What are your goals in 2022?

«Our goal is to be able to close 1.5 million dollars this year, but more than that, our goal is to impact millions of lives with that amount, as well as to go deeper into England and Spain, which are one of our focuses taking advantage of the current alliances we already have, and in the same way optimize 100% in Colombia and generate strategies for Mexico with some allies with whom we are touching base.»

What has been the human and economic investment force of the project?

«We have raised about a million dollars of non-participatory investment; in this industry, when you are focused on the clinical, you must have too much clinical and scientific validation that confirms that what you are creating has value. We have been around three years, we have already validated the technology from the clinical and scientific point of view, and we have already started sales.»

How could data intelligence help us cope with a future health contingency?

«Data intelligence could help us on every front the healthcare system faces today, from interoperability and administrative processes to molecule development in record time or early identification processes. I think it’s a tool that gives us agility when making decisions. To give an example, if you ask someone in the health system what the most important challenge is, I would probably say that knowing when one of us might develop a chronic or highcost disease for the system; with data intelligence and through risk factors, we can identify who may develop a chronic disease in the future, the same can be replicated with the next pandemic.»

- We know that in 2020 Arkangel IA won the Everis Global Award, in which more than 2,200 national and international projects from the health, biotechnology, and related industries participated; therefore, you could share with us

what distinguishes Arkangel IA from other technologies used for the detection of diseases?

«There are many companies that work with their predictive models, but we provide the service for institutions to create their own models for what they need automatically without knowing how to develop this technology. This brings ease and speed when implementing these technologies within their own companies. Second, we work with basic installed capacities and infrastructure at a fraction of the cost. Normally any institution that makes predictive models for certain diseases focuses only on the creation of the algorithm. »

According to your experience in scientific research and development, coupled with your business career, what strategies should be implemented in Latin America to strengthen scientific and technological education?

«There are many, but the main one would be connecting all the public spaces with the private ones and the academy. I think that by uniting the three entities and generating value from a sandbox, for example, interesting things could arise.»

What advice could you give to those immersed in scientific research to undertake new business models in the health sector?

«First, don’t just stay in research, but focus on how that research generates value. Because in the end, if a paper comes out,

it will be beneficial for scientists and fill us with more science, but it will not be applied, then I invite you to understand what is the real problem that can be solved and to investigate what can really be executed and generate value out there of the paper.»

- Finally, I would like you to address a topic of utmost importance, and that is entirely the responsibility of your company, and that is that because of the digitalization pro-cess that the sectors of the various industries have suffered as a result of the imminent modernization, and also clearly due to the COVID-19 pandemic,

what lines do you recommend following so as not to lose humanity amid the incredible robotization of processes?

«I believe 100% that humanity is not lost before it optimizes it. What allows this type of process is that doctors have more time to attend to the patient as it should; for example, in a medical consultation, 70% of the time is left filling administrative processes, but with these technologies, you can optimize the time of attention to the patient, not only do a biological checkup but as it was done before that also analyzed their behavior, their day to day, their habits, how they live, something that was done before and generated too much value.

What I want to tell you is that data intelligence and other technologies are what allow us to transform how we provide and access health today and in which we connect to generate a differential value that allows us to make decisions in an informed and more human way. The most beautiful thing about this type of technology is that the access gap between poor and rich countries can be reduced because today, half of the world’s population does not have access to health.

Therefore, the union of data science and the human allows for more human and close attention that helps move the system.»

Interview generated by Editorial CLIR Clinical Research Insider