Curve Fit Extrapolation for Ebola Outbreak

I've been following the ebola outbreak with interest, especially since hearing a number of competing calm assurances and hysterical predictions of catastrophe. To help understand the threat, I've taken the World Health Organization infection data and calculated the doubling period. Using an exponential model, I calculated a doubling period of 28.2 days, or four weeks. On October 8, the number of infections was 8100. So, on November 5 there will be around 16,200 infections. By Christmas Day there will be 55, 000 cases and January 15 there will be 95,000 cases. If uncontrolled, there will be 225 million infections (20% of Africa) by the end of 2015. I've plotted the model on the attached graph. The model is Io2^(days/28.2), where Io is the current number of infections and days is the number of days from now.

More info about ebola from the SciShow YouTube channel here:
BBC Ebola link:

ebola.png43.77 KB