Dr. Soufiane Ajana is an epidemiologist at Bordeaux Population Health Research Center, based in Bordeaux, France. He did his PhD in Machine Learning in Public Health at Bordeaux University.
He is a laureate of the SPARK program, an initiative launched by the Stanford Faculty of Medicine, to start a deep-tech startup based on his research work to prevent the onset of a AMD, a blinding eye disease.
Summary
In this episode, Dr. Ajana talks through what made him choose this path and how his work in machine learning is about predicting progression to advanced Age-related Macular Degeneration (AMD) from clinical, genetic and lifestyle factors.
Key Questions
7:31.89: Could you explain to our audience what AMD is?
12:51.45: I see that you’re using machine learning to predict the progress to progression of the AMD disease, right? So, is that a reason why you think this is particularly suited for machine learning?
22:31.98: What more is sort of needed to bring your work? So that it’s part of the regular clinical workflow of say maybe an ophthalmologist so that it just really becomes useful to people?
34:09.58: That’s a different perspective…it is said that being a data scientist is the sexiest job of the century…but you’re saying that you’re spending most of your time cleaning and getting the data?
36.04.25: How representative of real-world data science or a data scientist’s work is let’s say participating in Kaggle and going through those problems?
Quotable Quotes
7:43.61: AMD (age related macular degeneration) – is the foremost cause of blindness in the elderly in industrialized countries.
16:58.73: If you’re just using machine learning, you won’t succeed. If you’re just using the domain knowledge, you will succeed, but you can always go further. But by combining both, you start seeing some magic happening.
18:13.09: You can’t say to a patient that you’re at a high risk of having breast cancer or developing AMD, like in my case and just tell him or her that the model says so.
30:05.20: …just break your limitations, know and believe that you can do everything and that everything is possible and really everything will be possible once you will convince yourself of that.
37:29.32: Kaggle is a kind of a place where you practice your skills and then you hope that what you learn there will help you to better analyse [00:38:00] real-world data.
40:25.41: Please whenever you have the opportunity, do an internship. Don’t go to beaches in the summer and do other stuff. Do your internship and you will see that you will be really grateful for that afterwards.
47:16.21: Whenever you have a problem to solve, always try to define the objective of your study i.e. What is the question that you want to answer?
Notable Mentions
3:17.71: Dr. House, a TV series focused on medicine.
48:09.46: Albert Einstein, as an example to drive home the importance of asking the right questions.
Connect
Twitter: @Soufiane_Ajana
LinkedIn: Soufiane Ajana
YouTube: Soufiane’s work in 3 minutes, Soufiane’s bike trip across Europe
Website: Assess your risk of developing AMD
The You+AI Vodcast
A companion video segment full of fun and candid moments.
Watch it here!