Deep Learning–Based Clinical Decision Support System for Heart Disease Prediction Using ANN, CNN and RNN–LSTM

Q: What type of study is this?

This is a Validation Study study.

MD Samiulla; Yasir Khan; Dr P Privietha; Preethy Jemims P; Srideivanai Nagarajan

doi:10.5281/zenodo.18197159

December 30, 2025Open Access

Deep Learning–Based Clinical Decision Support System for Heart Disease Prediction Using ANN, CNN and RNN–LSTM

View Full Paper

Ask AI

Bookmark

Share

Key result

Deep learning models improve diagnostic accuracy over traditional machine learning for predicting heart disease.

Why the study?

Heart disease is a critical global health challenge, and early diagnosis is essential to improve patient survival rates.

Key Points

To develop and evaluate a multi-architecture deep learning framework using ANN, CNN, and RNN–LSTM models for accurate heart disease prediction based on patient clinical and physiological indicators.
Standardized clinical attributes from public repositories using label encoding for categorical variables and StandardScaler for numerical features, partitioning data into 80% training and 20% testing sets.
Implemented four deep learning models—ANN with dropout regularization, 1D-CNN for spatial feature extraction, LSTM for temporal patterns, and a combined RNN–LSTM network.
Trained all network architectures using the Adam optimizer alongside binary cross-entropy loss functions, evaluating performance across accuracy, precision, recall, and F1-score.
Deep learning–based models achieved better diagnostic performance than traditional machine learning methods for cardiovascular disease classification.
The integrated RNN–LSTM network effectively leveraged both sequential modeling and long-term memory to capture complex dependencies across physiological parameters.
The framework automated diagnostic classification to eliminate manual analysis errors and coupled positive disease detection with targeted lifestyle guidance modules.

Structured PICO

P

Population

Patient medical profiles from publicly available repositories (Heart.csv) containing health indicators like blood pressure, cholesterol, chest-pain type, ECG readings, and maximum heart rate.

I

Intervention

Deep learning algorithms (ANN, CNN, and combined RNN-LSTM) for heart disease prediction.

C

Comparator

Comparison among ANN, CNN, and RNN-LSTM models.

O

Outcome

Model performance metrics including accuracy, precision, recall, and F1-score for predicting the presence or absence of heart disease.

Deep learning models, particularly CNNs, demonstrate high accuracy in predicting heart disease risk from clinical and physiological data, offering a potential non-invasive clinical decision support tool.

Limitations

Need for larger and more diverse datasets to improve generalization
Need to incorporate additional medical features like real-time wearable sensor data

Implication

May enhance diagnostic tools; leaves open clinical adoption pending prospective validation.

Abstract

Heart disease is one of the most critical health challenges worldwide, and early diagnosis is essential for improving patient survival rates. This project presents a heart disease prediction system developed using deep learning algorithms such as Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN) and combination of Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) networks for effective classification of cardiovascular conditions. The system processes clinical and physiological patient data through data cleaning, normalization, and feature optimization before applying the learning models. These deep learning models are trained to identify complex patterns in medical data and provide reliable predictions. The proposed approach enhances diagnostic accuracy, reduces manual analysis errors, and supports healthcare professionals in making timely and informed decisions. The experimental results confirm that deep learning–based models achieve better performance than traditional machine learning methods, making the system suitable for real-world clinical applications.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Samiulla et al. (2025) studied this question. Deep learning models improved diagnostic accuracy and outperformed traditional machine learning methods in predicting heart disease.

synapsesocial.com/papers/6963221091e05aa366cb8845 https://doi.org/10.5281/zenodo.18197159

Topics

Coronary artery disease

STEMI management

View Full Paper

Ask AI

Bookmark

Share

Deep Learning–Based Clinical Decision Support System for Heart Disease Prediction Using ANN, CNN and RNN–LSTM

Key Points

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Topics

Deep Learning–Based Clinical Decision Support System for Heart Disease Prediction Using ANN, CNN and RNN–LSTM

Key Points

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

Topics

Keep reading