Evaluating AI Predictive Models for Enhanced Decision-Making in Cardiovascular and Respiratory Diseases

Cardiovascular and respiratory diseases are among the leading diseases affecting mankind, and their occurrence is in the high incidence and mortality levels. The diseases cause high mortality rates, and precaution and early diagnosis are very significant to enhance the patients’ lives and minimize costs. Specifically, many of the conventional methods of modeling are inadequate to provide solutions to modern problems and tasks that involve large amounts of time-variant data and therefore, there are limitations in the amount of predictive power possible. Earlier research has often considered the normal ML techniques that are not powerful enough to address the dynamic sequence and dependency issues. To overcome these challenges, this paper introduces a novel method based on recurrent neural networks (RNN), which improve prediction accuracy because of their ability to handle time series and sequential data. The proposed RNN model takes patient information in sequential data and uses the hidden state to analyses the dependent or independent features like medical history, examination details, and lifestyles. The model obtains a significantly high accuracy of 99.8% compared to conventional approaches. Python is used to implement the form of the proposed model. This work suggests that RNNs can assist in early diagnosis and decision-making in the health care sector by showing a way of predicting cardiovascular and respiratory diseases with relative accuracy and reliability. The future work can build upon by assimilating more patient data into the model and, also can study about the real time use of this model in health care for extensive utility of the model.

Перевод пока недоступен