Abstract:Pulmonary hypertension (PH) is a condition of high blood pressure that affects the arteries in the lungs and the right side of the heart (Mayo Clinic, 2017). A mean pulmonary artery pressure greater than 25 mmHg is defined as Pulmonary hypertension. The estimated 5-year survival rate from the time of diagnosis of pulmonary hypertension is only 57% without therapy and patients with right heart failure only survive for approximately 1 year without treatment (Benza et al., 2012). Given the indolent nature of the disease, early detection of PH remains a challenge leading to delays in therapy. Echocardiography is currently used as a screening tool for diagnosing PH. However, electrocardiography (ECG), a more accessible, simple to use, and cost-effective tool compared to echocardiography, is less studied and explored for screening at-risk patients for PH. The goal of this project is to create a neural network model which can process an ECG signal and detect the presence of PH with a confidence probability. I created a dense neural network (DNN) model that has an accuracy of 98% over the available training sample. For future steps, the current model will be updated with a model suited for time-series data. To balance the dataset with proper training samples, I will generate additional data using data augmentation techniques. Through early and accurate detection of conditions such as PH, we widen the spectrum of innovation in detecting chronic life-threatening health conditions and reduce associated mortality and morbidity.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to predict pulmonary hypertension (PH) using machine - learning techniques through electrocardiogram (ECG). Specifically, the author aims to create a neural network model that can process ECG signals and detect the presence of PH and its probability. The following are the key points mentioned in the paper: 1. **Background and Challenges**: - Pulmonary hypertension is a hypertensive disease that affects the pulmonary arteries and the right side of the heart. Its 5 - year survival rate is only 57%, and untreated patients can only survive for about 1 year after right - heart failure occurs. - Early detection of PH is a major challenge because the symptoms of this disease develop slowly and are often overlooked. - Currently, commonly used diagnostic tools such as echocardiography, although effective, are costly and require professional operation, which limits their wide application in resource - limited areas. - Electrocardiogram (ECG), as a simpler and lower - cost tool, has not been fully studied for PH screening. 2. **Research Objectives**: - Create a neural network model that can detect the presence of PH through ECG signals and give a confidence probability. - Improve the accuracy of ECG in the early detection of PH, thereby broadening the scope of innovation in the detection of chronic life - threatening diseases and reducing related mortality and morbidity. 3. **Methods**: - Use ECG data in the public dataset PTB - XL for model training. - Optimize data quality through techniques such as data pre - processing, standardization, and data augmentation. - Use a dense neural network (DNN) model for training and evaluate the model performance through the validation dataset. - Adopt multiple evaluation metrics, such as training and validation accuracy, loss function, ROC curve, and F1 score, to ensure the robustness and generalization ability of the model. 4. **Results**: - The accuracy of the model on the training dataset reaches 98%. - The accuracy on the validation dataset is also between 97% and 99%, indicating that the model is not over - fitted. - The loss function gradually decreases, and the validation loss is lower than the training loss, further verifying the effectiveness of the model. 5. **Future Work**: - Update the model to adapt to time - series data and improve the ability to process ECG signals. - Generate more training samples through data augmentation techniques to balance the dataset. - Develop a user - friendly application so that medical workers and patients can conveniently use this model for PH screening. In conclusion, the main objective of this paper is to use machine - learning techniques to achieve early and accurate detection of pulmonary hypertension through ECG signals, thereby improving the prognosis and quality of life of patients.

Predicting Pulmonary Hypertension by Electrocardiograms Using Machine Learning

Machine Learning for Diagnosis of Pulmonary Hypertension by Echocardiography

A Combination of Clinical, Electrocardiographic, and Echocardiographic Parameters Predicts Pulmonary Hypertension Occurrence in Patients with End-Stage Renal Disease

Deep Learning-Derived 12-Lead Electrocardiogram-Based Genotype Prediction for Hypertrophic Cardiomyopathy: a Pilot Study.

An electrocardiogram-based AI algorithm for early detection of pulmonary hypertension

Automatic Echocardiographic Evaluation of the Probability of Pulmonary Hypertension Using Machine Learning

Pulmonary Hypertension Detection Non-Invasively at Point-of-Care Using a Machine-Learned Algorithm

Deep Learning Based Prediction of Pulmonary Hypertension in Newborns Using Echocardiograms

Deep Learning to Detect Pulmonary Hypertension from the Chest X-Ray Images of Patients with Systemic Sclerosis

Interpretable Prediction of Pulmonary Hypertension in Newborns using Echocardiograms

Pulmonary Disease Classification on Electrocardiograms Using Machine Learning

Noninvasive Prediction of Pulmonary Hypertension Based on Finite Element Analysis and Machine Learning

Electrocardiogram-based heart disease prediction using hybrid deep feature engineering with sequential deep classifier

Echocardiographic artificial intelligence for pulmonary hypertension classification

Deep learning to predict elevated pulmonary artery pressure in patients with suspected pulmonary hypertension using standard chest X ray

ECG-guided non-invasive estimation of pulmonary congestion in patients with heart failure

A Hybrid Neural Network Model for Predicting Kidney Disease in Hypertension Patients Based on Electronic Health Records

Classification of pulmonary arterial pressure using photoplethysmography and bi-directional LSTM

A promising approach for screening pulmonary hypertension based on frontal chest radiographs using deep learning: A retrospective study

Predicting hypertension control using machine learning

Noninvasive Estimation of Mean Pulmonary Artery Pressure Using MRI, Computer Models, and Machine Learning