Diseases detection in seconds with AI
Where are you located (country)?
Colombia & Canada
Where are you located (city or town)?
Bogota & Montreal
If part of an organization or group, what is the name?
What is your organization or group’s website? (If applicable)
How does your proposed solution support emerging middle-class families in urban areas in the Global South to adapt and thrive during the COVID-19 pandemic?
Today, at least 2 million people are expected to die this year from diseases like COVID. It is just one of the conditions that we can cure, but that is not being detected in time.
The problem is that the current detection systems: are NOT scalable to the most vulnerable areas of the world, are very expensive and require an installed infrastructure to which few countries have access.
That’s why we created Arkangel AI, and artificial intelligence software for the early detection of preventable diseases. We have developed algorithms that allow us to identify in seconds positive cases of two species of Malaria in blood samples taken from a microscope, along with 19 pathologies, including COVID-19, from chest x-ray images. which generates results 20X faster, 10X more scalable, 5X more cost-effective, and all this with an accuracy above 95%. Designed to generate explainability, optimize decision making in health centers and generate equality in access to detection.
What stage of development is your Idea in?
Pilot: I have started to implement the solution as a whole with a first set of real users.
What tier are you applying to?
Global South Open Prize Tier: Open to submissions focused on any country in the Global South, at any stage of development.
Do you intend to implement your idea in Peru?
In what countries do you expect to implement your idea initially (in the first 2 years)?
Colombia, Mexico, Brazil, Democratic Republic of Congo
Feasibility: where are you with understanding the feasibility of your idea? Describe what you’ve accomplished to date, what barriers to implementation might exist, and what next steps you plan to take.
In Arkangel AI, we have validated our algorithms with a bank of more than 500 thousand images diagnosed by a specialist to test our accuracy metrics. Moreover, we are currently doing a field validation with a target population of 2M people in 27 health centers in rural and urban areas of different municipalities in Colombia. Our main barrier is our implementation cycle, as these processes require government approval; in some countries, it can take time to begin implementation. However, we are doing initial work with communities before the approval, understanding how the workflow is in health centers and to match the culture. After many tests and many mistakes, the results we have had so far this year have been encouraging. We raised over 400 thousand dollars of investment; we have the support of the top 3 universities in Canada supporting our technology. We are officially strategic partners of Novartis, one of the world's largest pharmaceutical companies worldwide. We have a contract with the Colombian government, where we are validating our technology in the field with a target population of more than 2 million people in vulnerable areas of the country. On the other hand, Google chose us as one of the nine startups they will support this year with their tech and mentorship. And finally, we are initiating an alliance with two countries in Africa to implement our algorithms.
Our vision is to impact 300 million people in the Global South and expand to 15 diseases by 2025
Viability: what needs to be true for you to be able to implement your idea? What stakeholders, partnerships and resources might be required to implement this solution?
At this moment, we are already validating its feasibility. We are at the most opportune moment to implement our technology, where the major stakeholders are the government because it saves money and increases the scope of mandatory health care coverage in the countries. The insurance companies save money in the cost of detection, expand the care scope to rural areas and patients triage—besides, the pharmaceutical companies, which are interested in detecting diseases in time to promote treatment.
The resources needed to impact millions of people and save lives are to finish our validating pilots to prove that our algorithms work in the field and generate a solution 10x more effective than the current ones to scale up globally.
Adaptability: how does your idea adapt to frequent changes within the context you are implementing, due to COVID-19? Consider the rapidly shifting government policies, and healthcare economic realities that might influence end users.
Arkangel AI was founded as a disease prevention algorithm through early detection of patterns in humans and a program developed by the Center for Disease Control that reduces the development of Type II Diabetes in high-risk patients by 60%. It was born to end the crisis of preventable diseases in the world where 80% of North American people are at risk of developing Type II Diabetes and generates expenses of 300 billion dollars. Along the way, we discovered massive interest from over 9,000 active users in 22 countries in our first MVP. However, because of poor control and high risk, we were unable to find Product-Market Fit, but we discovered that reading patterns, in medical information, were useful and very effective for early disease detection, and we decided to pivot to this for neglected diseases southern countries. When we were closing our first alliance with Novartis for the malaria early detection, the pandemic just started. And we decided to start our algorithms training for COVID-19 along with other 16 respiratory pathologies, which generates results 20 times faster, 10 times more scalable and 5 times more profitable.