The International Classification of Diseases (ICD) gives a ranking of the diagnostic codes for the classification of the diseases. The process of medical coding allocates the subset of diagnosis codes as per CMS guidelines for a patient’s visit which is critical for patient care and claims billing. Human coding intervention is error-prone, time-consuming, and expensive. This experimental study is conducted by adapting and combining the individual methods like OCR for text reading, statistical text mining, machine learning, and rule-based NLP algorithms. We tested statistical text mining and rule-based NLP approach by adapting and combining the individual techniques. Using these two distinct technology approaches applied in conjunction, developed technique has been evaluated on a set of 2500 clinical notes annotated be the human subject matter medical coding experts.