Problem: What problem is this project trying to address?
As medicines become increasingly available worldwide, a new key aim in the field of global health today is simple, fast and inexpensive diagnosis. Malaria, tuberculosis and cervical cancer are three deadly diseases that can be diagnosed by microscopic analysis of blood, sputum and pap smears, respectively.
While microscope image acquisition is inexpensive, the diagnosis protocols can be a manual and time-consuming task for trained specialists. There are currently no perfect automated methods for the image analysis of malaria, tuberculosis or cervical cancer samples, despite the massive incidence of these diseases. Worldwide, there are more than 600,000 new malaria cases per day, and a child dies every minute because of this treatable disease. Malaria, with over 230 million (known) cases per year, causes 1-2 million deaths yearly and is the disease with the lowest rates of accurate diagnosis in the world.
The current gold standard procedure for malaria diagnosis consists of first detecting parasites and then manually counting the number of parasites in blood smears through a microscope. This process can take more than 20 minutes of specialist’s time. However, worldwide there are not enough specialists and specialists are not always present in many malarial zones. This makes diagnosis availability and resources optimization critical. Accessible technology can provide assistance to remote places without specialists and optimize the resources involved in diagnosis procedures and or clinical trials, where a lot of effort goes into quality assessment of the specialists themselves.
Solution: What is the proposed solution? Please be specific!
Miguel is transforming the medical image analysis field by building a platform that integrates crowd-sourcing techniques, video games, artificial intelligence and mobile devices to make disease diagnosis available on a massive scale for people in low- and middle-income countries.
MalariaSpot is an online game that directly supports malaria diagnosis based on three pillars. The first pillar is the contributions of thousands of citizens connected through the Internet. Certain specific image analysis tasks – such as recognizing malaria parasites – can be rapidly learned by non-specialists, therefore exponentially increasing the potential global “workforce” devoted to image diagnosis, while saving the valuable time of medical specialists. Secondly, Miguel has designed his platform to take advantage of the users’ abilities to interact and play in digital worlds, a space which often poses more of a barrier than familiarity with the biomedical images themselves. Finally, Miguel has imbued his platform with a competitive edge, incentivizing and motivating players to make accurate diagnoses through gamification of the crowdsourcing approach.
The analyses of different players around the world combine into a single collective image diagnosis decision. Fundamentally, by combining the clicks from a significant number of non-expert volunteers, Miguel is creating an important and potentially massive new source of high precision diagnoses.
Miguel is thus developing a scalable, fast and inexpensive crowd-sourcing platform for medical image diagnosis. This platform has the potential for application in diagnosing several other diseases commonly diagnosed by imaging methods, including cervical cancer and tuberculosis. This idea can revolutionize a medical field dominated by siloed specialists and high barriers to knowledge by opening it up to external collaboration that leverages the power of the crowd.