This is because human biology is messy. The learning problem is much tougher than most people realize. Machine Learning algorithms work well in a context where the size of datasets is fairly large. This is a difficulty in the clinical area due to privacy issues such as HIPAA etc.
Clinical decision building is highly suited for rule-based systems because of the nature of the data, such as ICD-9 codes, medications, etc., which are discrete fields [often boolean] in the Electronic Medical record. Also, rule-based methods have performed extremely well in the past.
If you are a beginner and you are quite enthusiastic about Machine Learning, doing a Machine Learning Course will give you some possible good insights eventually making you master the domain.
And if you are more into youtube tutorials then here is an awesome video tutorial by Intellipaat which will clear all your doubts regarding the same:
Back