At this moment, we are at the mark of 75%. Where are we failing? We are mostly missing the mark with mild CFS patients. With controls our failure rate is 10% and that is mostly with EBV positive non-CFS patients. Long way to go as we are testing many different parameters.
That's sounds great to me, certainly an improvement on the status quo. If I understand correctly:
- a 25% false negative in mild CFS
- a 10% false positive in controls (if they have EBV)
Thank you so much for your persistence & innovation. & patient engagement
A challenge is perhaps whether to start grouping by symptom subsets, or to delineate subgroups based on clusters of markers.
For the latter, AI "unsupervised learning" could work if data set is large enough...
We are able to identify severe ME/CFS patients with neurological issues using many parameters with a success rate of 100%. That means we have a clear sub-grouping with these patients. Mild and moderate patients are difficult to group as many parameters overlap.