A Feasibility Study Demonstrated High Accuracy With Machine Learning
The first publication on eye-rubbing detection using only a wrist-worn device, without the need for additional sensors.
Read more: The latest cover article
The correlation between eye-rubbing and Keratoconus has now been widely accepted, yet it remains difficult to objectively assess the intensity and frequency of eye-rubbing habits.
This is therefore an important step forward in supporting the "No rub, no cone" paradigm of keratoconus progression and sets the groundwork for future studies to come.
What makes this work stand out is its high detection accuracy (97%) through wrist movements alone, offering a convenient and less intrusive way to monitor and prevent keratoconus progression.
This is the result of a collaborative effort of two highly esteemed teams, led by those who brought with them their expertise in AI to make this system possible.
As an added bonus, the paper was selected for the cover of TVST showing the recognition of the impact this research can have.