Natural Language Processing For Medical Text Analysis

Abstract

A lot of healthcare data is generated on a regular basis. This can be utilized to extract information required for disease occurrence prediction in a patient. For decision making and disease prediction, it is necessary to leverage the treatment history and health data present in the patient data most of which are 'buried' in patient data like EHR/EMR. The volume of data generated on a daily basis is extremely massive which requires leveraging data mining or machine learning methods. The hope of utilizing the analytical and ML techniques is to forecast clinical outcomes in advance so that supports the medical professionals for early diagnosis of disease and chronical diseases so that treatment can be initiated early or minimize risk of life to the patient. Early diagnosis and treatment can minimize the treatment cost to a major extent. Probabilistic modelling and deep learning method will train a Long Short-Term Memory recurrent neural network and a convolutional neural network to forecast the occurrence of the disease. The specific combination of deep learning methodologies and an abundance of data in the EHR is extremely beneficial in enhancing the understanding of human health.