AI-FOX: Logic

AI-FOX searches the almost infinite MAP-space in an intelligent and efficient manner and within a limited search-domain that is defined by the active MAP and the measured psycho-acoustic outcome.  This yields millions of alternative MAPs.

For each new MAP, a Utility metric is defined reflecting the model’s prediction of the hearing outcomes with this particular MAP.  A high utility means a high likelihood of better hearing outcomes.  The MAP with the highest positive Utility becomes the one recommended to the audiologist.

AI-FOX can share the Utility of this MAP with the audiologist, together with the probabilistic predictions of how this MAP is anticipated to change the outcome.

AutoMaps

The AI-FOX MAPping workflow starts with a series of MAPs called “Automaps” which gradually increase in level. These MAPs are obtained by means of statistical analysis on the set of final MAPs used by the population of all CI users in the database and are filtered for optimal outcomes. AutoMaps represent the MAPs of all users with optimal hearing outcomes.

Starting from initial fitting, every CI user goes through the same series of Automaps until the comfort levels stabilize. From that point, the MAPs are personalized by means of the fine tuning process using the four tests previously described and driven by hearing outcomes. The whole clinical workflow takes four sessions in the first year.

Continue reading on AI-FOX’s learning capacity…