"The future looks bright: we are able to build AI models that can predict risk at the patient level and advise on personlized treatment options."
For over two years, Anne-Laure Budts has been working at EPCON as a data scientist and data lead for various country projects, including Nigeria, the Philippines, India, Bangui and Belgium. Her role involves using artificial intelligence to map and quantify health risks and predict disease burden hotspots.
So, what exactly does this involve? Anne-Laure gathers data from various sources, combines them into models, and then validates and compares them. The data is then visualized on platforms such as dashboards and geoportals, in a way that clients find comprehensive and easy to understand. At the same time, Anne-Laure also works to improve the AI engine by conducting literature reviews and research and development.
Given the diverse and stimulating nature of this work, it is not surprising that Anne-Laure has a wide range of pertinent skills. She holds Master's degrees in both biomedical engineering and artificial intelligence, with a focus on programming.
Anne-Laure is passionate about her work. She gets to improve her skills while doing meaningful work that clients respond to positively. The EPCON AI tool helps multiple sectors, including governments, industry, and NGOs, gain insights into various health risks. Anne-Laure enjoys working with the small but talented EPCON team, where the mix of different expertise and close collaboration boosts performance.
Anne-Laure believes that AI health models are capable of predicting risk at the individual or patient level, with personalized and customized treatment options. These models will be built from multiple indicators, including specific patient data and treatment protocols, to help with the optimal allocation of resources.
However, Anne-Laure notes that the models are only as good as the quality of the data inputted, which remains a challenge in the sector. EPCON spends a lot of time searching for good data sources and verifying their quality. Next, predictive models often have a lack of flexibility and adaptability: they need to be retrained when new scenarios arise. But Anne-Laure finds this to be a rewarding challenge as it gives her the opportunity to learn something new every day.