Read this first#
This was a research prototype, not a clinical diagnostic system. It was never clinically validated and must not be described as one.
The idea#
Irregular biological shapes are not always well described by ordinary Euclidean measurements. Fractal dimension gives another way to describe boundary irregularity, which made it worth asking whether it could separate one class of image from another.
An MVP, not a product.
The award#
The work won the poster competition at the AGSE (Applied Geoinformatics for Society and Environment) International Congress 2010, during her BSc in Systems Engineering with an Artificial Intelligence concentration at Universidad Católica de Santa María in Peru.
Why it stays in the portfolio#
Not because it shipped, and not because it is medically useful. It stays because it dates the interest: a 2010 research award, an AI-concentration degree, an accessibility venture in 2015, applied AI now.
This was a research prototype, not a clinical diagnostic system. It was never clinically validated and must not be described as one.
Frequently asked questions#
Is this a medical diagnostic system?
No. It is a 2010 research prototype. It was never clinically validated, never deployed for diagnosis, and must not be described as a diagnostic system.
What did the research actually do?
It used skin-lesion image samples and explored fractal dimension as a feature for telling one class of image from another. Image samples, a computed feature, then classification.
What was the award?
It won the poster competition at the AGSE (Applied Geoinformatics for Society and Environment) International Congress in 2010.
A long-standing interest
From a 2010 research award to applied AI now. The rest of the early work is in the archive.