Heart problems can be revealed by observing forests and trees- new study reveals

Heart-Forests-treesIndian researchers discovered that watching forests and trees could help in detecting heart problems.

A new study conducted at the PSG college of Technology in Coimbatore, India by C. Vimal and his team established that heart problems can be detected by considering various tools like Random Forests, Logistic Model Tree and neural Network. They have examined statistical data available on heart problems.

C. Vimal explained that Heart rate and Heart Rate Variability (HRV) reflects the state of cardiovascular system and is popular in the field of cardiology for diagnosing cardiac abnormalities.

Researchers are working on finding new techniques for detecting heart problems and finding the possible indicators of imminent heart failures more quickly than the current available techniques.

During their course of discovery the team explained that automated detection and classification of cardiac diseases help the physician to diagnose the cardiac abnormalities in a speedy manner.

Generally, Electrocardiogram is the first stage to detect any heart abnormalities as it records the heart's electrical activity but it is not a persistent technique.

The major drawback from which ECG suffers is that it cannot record the instant changes in the heart's behavior as symptoms of heart problems can be shown at any time.

Whereas, Heat Variability test (HRV) is more productive in diagnosing the problem although it takes more time. Experts believed that the results declared by HRV are clearer and proved to be better in providing aid to the patient.

Researchers also said that the most important HRV determinant of death and that helps in identifying the patients at risk is the low or high frequency electrical changes.

To reach the conclusion, researchers have analyzed heart data of various heart diseases from a website, physionet website, which is loaded with medical data of various diseases and their studies. Team has used three different approaches: Random Forests, Logistic Model Tree and Multilayer Perceptron Neural Network.

With the help of these approaches, team succeeded in, achieving 98.17% accuracy.

Researchers also said that after the successful discovery of this system it will aid the physicians in the classification of heart diseases. They also said that in future, the system will be used on real heart patients to check its performance and verify the observations.

The study is printed in the International Journal of Electronic Healthcare.