AI Enters medical industry- AI-ECG platform diagnoses ECG report in real time

  (photo credit: PR)
(photo credit: PR)

Artificial intelligence in the medical industry

In recent years, artificial intelligence (AI) technology has made tremendous progress and has become a reality in many areas of our daily lives. In the field of healthcare, efforts are being made to apply artificial intelligence technology to actual medical treatment. AI is seen as a method to solve the long-term challenges faced by the medical system. AI can provide the medical system with direct support to the most remote areas, thereby reducing the pressure of the aging population on the medical system's diagnosis and treatment capabilities and helping to correct the uneven distribution of medical resources and high-level medical personnel between different regions.

At present, artificial intelligence has shown its value in many fields, including auxiliary diagnosis, imaging, and accurate matching between patients and doctors. In the future, AI will show its value in more and more scenarios to help patients and doctors. Among them, artificial intelligence technology helps ECG analysis get accurate data faster and more intelligently. What is the difference between artificial intelligence ECG algorithms and traditional ECG algorithms? 

Artificial intelligence ECG algorithm VS traditional ECG algorithm

Globally, arrhythmia and sudden cardiac death are the biggest risk factors for human health. The clinical electrocardiogram is a method widely recognized by the medical community for screening and diagnosing an arrhythmia. However, traditional ECG data are highly dependent on manual feature extraction. Due to long-time work, manual diagnosis is prone to fatigue, which affects the accuracy of ECG report. In contrast, the AI-ECG platform combines artificial intelligence technology with ECG diagnosis to accurately identify heartbeats and determine various types of arrhythmias in ECG so as to help clinicians read and analyze ECG quickly and accurately.

Artificial intelligence ECG algorithm is more advanced than traditional ECG algorithm. Its artificial intelligence (AI) algorithm based on deep learning has recently achieved the most advanced performance in many fields, including medical imaging and diagnosis. One advantage of deep learning is to automatically learn features and relationships from given data without medical domain knowledge. The artificial intelligence ECG algorithm uses a fully automated AI model to mimic human-like interpretation of ECG. Compared with traditional rule-based computer interpretation, the artificial intelligence ECG algorithm has greater diagnostic fidelity and higher workflow efficiency. Take traditional algorithms and the AI-ECG platform for a comparison example in analyzing. 

  1. Workflow efficiency:

Traditional algorithms mainly identify the RR intervals of ECG and easily ignore P and QS waves. The electrocardiographic phenomena in the human body are complex. The traditional algorithms are easy to misdiagnose when the multiple ECG phenomena are in the body and interfere. 

AI-ECG platform is based on waveform image imput, which can capture the entire waveform effectively. What's more, the multi-layer processing of convolutional neural networks can effectively remove the influence of interference information on ECG diagnosis.

  1. Diagnostic Fidelity: 

Mainly using the traditional automated 12-lead, the positive predictive value of traditional algorithms for screening atrial fibrillation based on photoplethysmography (PPG) is 91.6%. And the analysis is mainly based on the subjective experience of doctors and cannot be made based on previous clinical cases, so the analysis results have certain limitations and error rates.

Compared with the traditional one, the AI-ECG platform has above 99% sensitivity for all 12 types of arrhythmic events, and the positive predictive value for screening atrial fibrillation is 98.67%. AI-ECG platform combines more than seven thousand applications and fourteen million ECG service volumes, making analysis results more reliable. 

AI-ECG platform is the first artificial intelligence medical technology for ECG automatic analysis and diagnosis in China. Its diagnosis projects cover the main ECG diagnosis events, such as arrhythmia events, especially premature beats, tachycardia, escape beats, conduction block, atrial fibrillation, and atrial flutter. Compared with the traditional ECG analysis and diagnosis methods, it has great advantages, with an accuracy of more than 95.0%, which is comparable to the level of ECG medical experts.

AI-ECG application scenarios

As a useful and intelligent ECG analysis method, AI-ECG has been vigorously used by hospitals and families.

In hospitals, the applications of AI-ECG can make clinical workflows more efficient and improve the safety of patients who stay in intensive care units (ICU) or operating rooms. Bedside ECG monitors are currently only used for bedside monitoring of critically ill patients due to their high price, large size, and need for power connection. Therefore, many patients with cardiovascular disease or older people are often not effectively monitored during gastroscopy, hysteroscopy, or minor surgery such as tooth extraction.

At home, AI-ECG helps seniors with a more independent daily life. Using AI-ECG is helpful for physical rehabilitation treatment by its daily data collection in managing chronic diseases. In such strained medical conditions, it can relieve the hospital's stretched resources and save a huge amount of expenses for the patient's family if those elderly patients carry out daily chronic disease management at home. The results of AI-ECG monitoring at home are equally accurate and effective and do not interfere with the normal daily life of elderly patients.

AI-ECG provides full-process solutions covering pre-hospital screening, in-hospital inspection and monitoring, and post-hospital rehabilitation monitoring. Smart healthcare is data-intensive. Through big data analysis and artificial intelligence, the AI-ECG Platform can assist doctors in improving the accuracy and efficiency of diagnosis and play a role in improving the level of medical services and alleviating the shortage of medical resources.

At the same time, the AI health all-in-one machine of the AI-ECG Platform artificial intelligence electrocardiogram diagnosis technology system has been widely used in primary care, social health, and nursing homes to provide convenient services for the elderly.

Home medical treatment

Intelligent medical care is approaching consumers. AI-ECG can be combined with er1. Users can measure by themselves with the help of the AI-ECG Platform, and the platform will perform diagnosis. AI-ECG Platform can provide 24-hour artificial intelligence-based remote heart disease services for the vast majority of heart disease patients who are not in front of doctors.

  (credit: PR) (credit: PR)

AI-ECG Platform

The ECG analysis software can monitor and warn up to 45 kinds of abnormal ECG events such as atrial fibrillation and atrial flutter in real-time, providing more comprehensive, more diverse, and more authoritative AI-ECG analysis services.

  1. Massive database

At present, the AI solution has been applied to more than 7100 medical institutions in China, with a total of 1,400 million ECG real-time data and 21.2 million dynamic reports, which has an early warning on about 350,000 high-risk patients. The amount of ECG data storage ranks among the top in the world. 

What's more, the AI team published an AI ECG analysis paper, "Arrhythmia Classification with Heartbeat-Aware Transformer", at the IEEE conference. The paper is based on the transformer model architecture, which enhances the correlation between heartbeats by the mechanism of Cardiac attention to improving the model's ability to recognize ECG heartbeats, thus further enhancing the accuracy of  AI-ECG algorithm. The paper also explains the inference mechanism of the model to help understand some internal logic of the AI-ECG algorithm.

  1. CE/FDA/NMPA certification

AI-ECG platform has become the first approved and registered AI ECG product in China and the only AI ECG product with NMPA approval, FDA approval, and CE certification at the same time.

  1. Case analysis

"Information integration is the meaningful direction for medical data to develop," a doctor from a hospital that has been utilizing the AI-ECG Platform for six years said. AI-ECGPlatform provides more convenient services for patients with Dynamic ECG. Patients wearing the Dynamic ECG devices will be monitored online and diagnosed by artificial intelligence by data. 

 In May 2020, a patient who uploaded physical health data through a Dynamic ECG device was found to be in a coma by a heartbeat monitor. AI-ECG preliminarily diagnosed arrhythmia. The monitor immediately sent a real-time alarm and submitted a preliminary AI diagnosis. The doctor immediately implemented treatment based on the AI-ECG diagnosis, saving the patient's life.

By monitoring patients' heartbeat condition, The Dynamic ECG has protected 30 thousand patients in this hospital and alarmed more than 7700 times. AI-ECG platform provides patients with faster and more accurate services.

About Viatom

Viatom is affiliated with the AI division of Lepu Medical Group. Viatom is a high-tech company specializing in the development and production of home medical devices. Viatom also provides OEM services. Viatom has focused on different medical fields for many years and has been committed to innovative research and development of mobile medical equipment. At the same time, Viatom is committed to bringing the ECG analysis and diagnosis technology and services into hospitals, grassroots, and homes to meet the growing demand for ECG examinations worldwide. 

Content Disclaimer:

This information does not constitute advice or an offer to buy. Any purchases made under the above press releases are made at your own risk. You should consult an expert advisor/health professional before purchasing. Any purchase from this link is subject to the final terms and conditions of the above website sales as the source. The advertising agency and its downstream distribution partners are not directly or indirectly responsible. If you have complaints or copyright issues related to this article, please contact the companies covered in this news.

This article was written in cooperation with Viatom