Voice Based Doctor Prescription

In India, thousands of people die as a result of wrong medication and normal ailment leads to severe ailments. The solution is voice recognition. It plays a vital role in effective interaction between humans and computers. The solution of recognition of voice can be used as a trigger to the number of the virtual application. This paper proposed a framework that signifies the use of voice recognition techniques for medical prescription. The goal is to avoid wrong medication for normal ailments viz. fever, cough, cold, body pain, etc. by designing virtual applications on voice base medical prescriptions. In this system, basic theories for voice recognition such as pre-emphasis feature extraction pattern comparison are implemented. The MFCC is used as a feature extraction technique and DTW as a pattern comparison. Since there is extensive fuzziness in medical prescription for a single disease or symptom, the fuzzy decision logic is introduced here for appropriate medication. System performance is analyzed by creating a symptom corpus.