Multi-system Disease Diagnosis, Public Health, Public Health System Research
Conditions
Brief summary
This registered multicenter study aims to investigate the diagnostic efficacy of artificial intelligence-enhanced electrocardiography (AI-ECG) in detecting multi-system diseases. The research will utilize prospectively collected data from inpatient, emergency, and outpatient populations to develop ECG-based diagnostic, screening, and predictive models for multi-system diseases.
Detailed description
Recent advances in artificial intelligence (AI) have expanded the diagnostic capabilities of electrocardiography (ECG) beyond cardiovascular diseases. Emerging evidence demonstrates that AI-enhanced ECG analysis can provide valuable insights into age, gender, mortality risk, cardiac function, and systemic conditions such as electrolyte imbalances, renal dysfunction, and thyroid disorders. These findings position ECG as a promising tool for the identification and prediction of a broad spectrum of diseases. To further investigate the underlying mechanisms linking ECG abnormalities with multi-system diseases and to develop ECG-based diagnostic, screening, and predictive models, we initiated a multi-center, prospective, observational registry study involving patients undergoing ECG examinations. The goals of the project are as follows: 1\. AI-ECG Foundation Model Development 1. Diagnosis of traditional cardiovascular diseases (e.g., arrhythmias, myocardial infarction). 2. Screening of multi-system disorders, including: Circulatory, digestive, respiratory, and nervous system diseases, Endocrine/metabolic disorders, urogenital diseases, hematologic conditions, Neoplasms and mental health disorders. 3. Prediction of new-onset conditions (e.g., atrial fibrillation, heart failure, valvular diseases, NSTEMI, ventricular tachycardia) and 1-year mortality risk. 2\. Clinical Utility & Implementation Leveraging the portability, cost-effectiveness, and non-invasiveness of ECG, our AI foundation model enables: 1. Rapid, large-scale screening in outpatient, inpatient, emergency, and community settings. 2. Early detection of multi-system diseases, guiding targeted diagnostic workups. 3\. Mechanistic & Interpretability Research Elucidating the diagnostic, predictive, and risk-stratification logic of AI-ECG foundation models.
Interventions
OTHERECG screening
Each subject is subjected to ECG assessment.
Sponsors
RenJi Hospital
Study design
Observational model
CASE_CONTROL
Time perspective
PROSPECTIVE
Eligibility
Sex/Gender
ALL
Healthy volunteers
Yes
Inclusion criteria
1. Patients who visited the study hospital. 2. Patients included should have both ECG data and discharge diagnosis codes (ICD-10) for inpatients and emergency patients.
Exclusion criteria
1\. Patients who declined participation, cases with incomplete or missing clinical data, and pregnant individuals.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Multi-system disease predicting based on ECG | 1 month | Evaluating the effectiveness of ECG in predicting diseases across various systems, such as circulatory system diseases, respiratory system diseases, digestive system diseases, nervous system diseases, urogenital system diseases, endocrine and nutritional/metabolic system diseases, hematological diseases, infectious and parasitic diseases, tumors, and mental and behavioral disorders. This study initially uses the ICD-10 coding system for preliminary screening of target diseases. Subsequently, a committee of multidisciplinary clinical experts conducts a systematic review of candidate diseases based on the ICD-10 coding system framework, including the applicability of diagnostic criteria, the accuracy of ICD-10 classification, the reasonableness of exclusion criteria, and the assessment of the level of evidence. |
Countries
China
Contacts
Primary ContactJun Pu, MD,PhD
Outcome results
None listed