Doubts, page-6

Martin,
Regarding the fact that the algorithym will not be learning DURING the USA trials .I believe this is simply due to the fact it is a double blinded trial.
To learn the algorithym requires that a human inserts the actual correct or incorrect result information into the system against each cough The system had diagnosed correctly or incorrectly. We do not know if that occurred continuously during the Australian trials, periodically or at the conclusion. As the results are blinded continous feedback is not possible in the USA trials, However at the conclusion of the trials those many unblinded results could be fed into the algorithym so that the learning benefit is not lost for the future.
I dont think its a question of objectivity so much as that simple difference between conducting unblided vs a double blinded process.
Fox