I'm in the midst of reading Eliezer's sequences and I have learned Bayes' Theorem because of it. I wanted to apply it to my friend's current uncertain situation. Her cat was diagnosed with a disease called FIP and she has treated her cat with remdesivir. However, there is a chance of about 10% according to studies that her cat might relapse after finishing the treatment. So she has been taking blood tests ever since to know whether her cat has relapsed or not. So I wanted to apply Bayes Rule to the blood cell counts and Biochemistry to update the prior probability of 10% each time she gets a new lab result.

Of course, my problem is with finding P(B|A) and P(B|~A) (A: is the probability that the cat has been cured). I have found a study that shows blood cell counts for cats that are infected with FIP. Their mean and Range and Median is also available. I know I can use a normal distribution to calculate P(B|~A). However I could not find a study that shows these values for Healthy cats so I can get the P(B|A). I don't exactly know what to search for really...

Healthy cat Complete blood counts or biochemistry didn't yield anything in Google Scholar.

I also know that this must be available somewhere because all blood tests come with a reference range.

New to LessWrong?

New Answer
New Comment