Release date: 2018-03-15
In the United States, more than 2,000 patients have a heart attack every day, and more than 400 of them have failed to receive treatment in time. When an object blocks an artery that supplies blood to the heart, it can cause a heart attack. If there is no blood, the heart will not function properly without the necessary nutrients and begin to experience depletion. The longer the patient waits for treatment, the more likely it is to cause irreparable damage to the heart. Therefore, timely detection and treatment of heart disease has become the key to saving patients' lives.
Although researchers have made some progress in detecting heart attacks, the basic methods of testing are no different from a century ago. At present, doctors commonly use electrocardiogram (ECG) developed in the early 20th century to monitor the electrical activity of the heart. Depending on the location and severity of a heart attack, certain areas of the ECG may change. However, these changes are small and unstable and include only a small fraction of all electrical signals of the heart. With the development of science and technology, researchers have applied different signal processing methods and other complicated mathematical operations to the electrocardiogram, but these treatment methods still can not show the difference of each person's heart.
Just as everyone's fingerprints are different, each person's heart shape and pulse power are slightly different, which makes their ECG signals different from others. What's more, the space between the body surface detection device and the heart may vary greatly depending on the patient's weight, gender, and body type. All of these changes make it difficult for automated systems to predict heart conditions at some point. To solve this problem, we need to develop a new system that can be adjusted to determine if there is a heart attack based on each individual's unique heart shape and signal.
To improve ECG measurement technology, the research team used computer science's latest machine learning technology to "teach" computers to read ECG signals. Combined with machine learning techniques, the ECG can tell us more about the heart than before.
How machine learning works
Researchers use machine learning techniques to train computers to identify features in data sets that are not easily visible to the naked eye. The researchers provided the computer with multiple sets of categorical data with different characteristics, allowing the computer to "learn" the characteristics of the data set that determine the data classification. These features detected by computers are often very small and complex and can be difficult to distinguish by the naked eye. Once the computer understands the corresponding characteristics of the different categories, it can apply this knowledge to determine the category to which the new data set belongs.
How do we apply machine learning?
The Scientific Computing and Imaging Institute (SCI) at the University of Utah is a world leader in biomedical computing and visualization. The overall research goal of the SCI Institute is to create new scientific computing technologies, tools and systems that provide solutions to many important issues in biomedical, scientific and engineering fields. At the same time, the SCI Institute is also committed to harnessing the power and versatility of modern computing to advance progress in all areas.
Researchers at the SCI Institute have begun using machine learning to detect changes in heart signals that reflect the initial characteristics of a heart attack. The researchers isolated the electrical signal from the heart and examined changes in the signal before, during, and after the simulated heart attack. The computer will then read the signals and classify the data into two categories: "have a heart attack" and "no heart attack." The computer can determine the start time of a heart attack faster, a 10% increase in speed compared to humans. At the same time, computers are 32% more accurate than humans in detecting early signs of heart attack.
The future of heart disease detection
Using machine learning to help doctors perform heart disease testing can advance the field of cardiology. Heart disease is one of the worst diseases in people's lives, and doctors use more advanced techniques such as artificial intelligence and machine learning to detect and help treat the disease. This technique can also be used for patients who are at risk for a heart attack due to genetic or environmental factors. The study provides a new way to understand and detect heart attacks and may even reduce the risk of heart attack.
We hope that future machine learning techniques can be applied to the detection and diagnosis of more diseases, so that more patients can benefit from it.
Reference materials:
[1] Machine learning could improve how doctors diagnose heart attacks
Source: WuXi PharmaTech
Canned Mackerel In Vegetable Oil
Canned Mackerel In Vegetable Oil,Vegetable Oil Canned Mackerel,Vegetable Oil Mackerel Tin,Vegetable Oil Mackerel Canned
ZHANGZHOU TAN CO. LTD. , https://www.zztancan.com