Chinese PE Multimodality Imaging Artificial Intelligence Study

CHinese pulmOnary Embolism Multimodality Imaging-artifiCial intelligencE Study

The CHinese pulmOnary Embolism Multimodality Imaging-artifiCial intelligencE Study (CHOICE) is a prospective observational multi-center study that will collect imaging text data and raw data of patients with pulmonary embolism (PE) in China. By combining artificial intelligence technology, it aims to identify imaging markers to assist in early diagnosis, differential diagnosis, risk stratification, and prognosis assessment of PE.

Study Overview

• Status: Recruiting
• Conditions: Pulmonary Embolism; Chronic Thromboembolic Pulmonary Hypertension; Chronic Thromboembolic Pulmonary Disease
• Intervention / Treatment: Device: Artificial Intelligence

Detailed Description

Pulmonary embolism (PE) represents a significant public health issue. Timely diagnosis and treatment during the acute phase, as well as appropriate long-term follow-up strategies, are crucial for the management of PE. PE is classified into three stages based on disease course: acute pulmonary embolism (APE), chronic thromboembolic pulmonary disease (CTEPD), and chronic thromboembolic pulmonary hypertension (CTEPH). APE can cause acute right ventricular failure and death if not diagnosed and treated early. CTEPD has the potential to significantly impair patients' quality of life. CTEPH is a rare and potentially life-threatening long-term sequelae of PE, characterized by persistent obstruction of pulmonary arteries by organized clots, leading to redistribution of blood flow and secondary remodeling of the pulmonary microvasculature. Early identification of PE and implementation of targeted treatment plans will significantly improve survival rates and prognosis.

Multimodal imaging tests play a crucial role in the management of PE (including computed tomography pulmonary angiography (CTPA), magnetic resonance imaging (MRI), echocardiography, and lung ventilation/perfusion (V/Q) scan). The guidelines have identified the right ventricle to left ventricle (RV:LV) ratio >1.0 on CTPA or right heart dysfunction signs from echocardiography as important indicators for risk stratification of APE. Patients stratified as high risk require closer monitoring in an inpatient setting. Whereas, those stratified as low risk are suitable for early discharge.

Therefore, exploring novel imaging markers and integrating these markers into radiology reports may have potential clinical significance. If no quantifiable evidence of right ventricular dysfunction is provided to clinicians to make treatment decisions, patients with high-risk APE may be considered "low-risk" and discharged home. In addition, patients with low-risk APE may require longer hospital stays and may not need to be hospitalized, which undoubtedly increases healthcare costs. For patients with CTEPD or CTEPH, treatment options are diverse, including multimodal therapies such as pulmonary endarterectomy, balloon pulmonary angioplasty and targeted medical therapy. Therefore, multimodal imaging evaluation is meaningful for clinical treatment decision-making and efficacy monitoring. Combined with artificial intelligence (AI) technology, it can provide a variety of metrics to assist in evaluating clots morphology, pulmonary ventilation-perfusion function, cardiac function, hemodynamics, and more. AI can not only assist in finding more clinically significant imaging biomarkers but also customize standardized radiology reports, which are expected to address the current challenges.

This study is a multi-center real-world study aimed at exploring novel imaging markers in combination with AI technology and integrating them into a software for clinical application to provide quantitative parameters, using imaging reports and raw data from Chinese patients with PE. It is hypothesized that AI technology can improve early diagnosis, differential diagnosis, risk stratification, and management of PE by increasing the ability to accurately evaluate PE in a real-world clinical setting. The researchers also hypothesized that the integration of AI technologies would be cost-effective and acceptable to radiologists and clinicians.

• Study Type: Observational
• Enrollment (Estimated): 1500

Contacts and Locations

• Study Contact: Name: Min Liu, PhD; Phone Number: +86-10-84205056; Email: mikie0763@126.com

Study Locations

• China
  • Beijing, China, 100029
    • Recruiting
    • China-Japan Frendship hospital
    • Contact: Min Liu, PhD; Phone Number: +86-10-84205056; Email: mikie0763@126.com
    • Contact: Zhenguo Zhai, PhD; Phone Number: +86-10-84206269; Email: zhaizhenguo2011@126.com
    • Principal Investigator: Min Liu, PhD

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

Patients suspected of PE in China

Description

Inclusion Criteria:

• 14 Years and older
• Patients suspected of PE

Exclusion Criteria:

• Pregnant women
• Refuse to follow up
• Incomplete or discontinued imaging scans
• Insufficient quality of image data to allow for analysis

Study Plan

How is the study designed?

Number of groups / cohorts

4

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Acute pulmonary embolism cohort; Patients objectively confirmed acute symptomatic PE or PE with deep vein thrombosis (DVT); PE was confirmed by CTPA, lung V/Q scan or pulmonary angiography.	Device: Artificial Intelligence; AI technology will provide novel imaging markers and generate a radiology report with relevant key slice imaging and evaluation results
Chronic thromboembolic pulmonary disease without pulmonary hypertension cohort; Patients with functional impairment despite 3 months of adequate anticoagulation therapy after APE.; CTPA/ pulmonary angiography or V/Q scan showed unresolved thrombi in the pulmonary vessels.; Without pulmonary hypertension at rest(mean pulmonary arterial pressure (mPAP) <20 mmHg), as measured by right heart catheterization.	Device: Artificial Intelligence; AI technology will provide novel imaging markers and generate a radiology report with relevant key slice imaging and evaluation results
Chronic thromboembolic pulmonary hypertension cohort; Patients with functional impairment despite 3 months of adequate anticoagulation therapy; CTPA/ pulmonary angiography or V/Q scan showed unresolved thrombi in the pulmonary vessels.; With pulmonary hypertension at rest (mean pulmonary arterial pressure (mPAP) >20 mmHg), as measured by right heart catheterization.	Device: Artificial Intelligence; AI technology will provide novel imaging markers and generate a radiology report with relevant key slice imaging and evaluation results
Other pulmonary vascular disease cohort; Patients diagnosed with other pulmonary vascular disease including Takayasu arteritis, pulmonary artery sarcoma, and fibrosing mediastinitis.	Device: Artificial Intelligence; AI technology will provide novel imaging markers and generate a radiology report with relevant key slice imaging and evaluation results

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Diagnostic rate of PE	Comparison before and after AI technique.	2 years
APE risk stratification rates (low, intermediate low, intermediate high and high risk)	Comparison before and after AI technique.	2 years
Disease severity of chronic thromboembolic pulmonary disease (CTEPD)/chronic thromboembolic pulmonary hypertension (CTEPH)	Comparison before and after AI technique. Assessment of disease severity is comprehensive, referring to the comprehensive risk assessment in pulmonary arterial hypertension (three-strata model) [DOI: 10.1183/13993003.00879-2022], including clinical observations and modifiable variables. The higher the score, the more severe the condition.	2 years
30 day mortality	Patient mortality (death) at 30-days post-PE diagnosis. Comparison before and after AI technique.	2 years

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Rate of discordant PE cases	False positive and false negative rate	2 years
AI failure rate for PE detection	Proportion of scans unable to be interpreted by AI despite suitable CTPA acquisition	2 years
12 month mortality	Patient mortality (death) at 12-months post-PE diagnosis. Comparison before and after AI technique.	2 years
Length of hospital stay for PE	Comparison before and after AI technique. Measured as number of days from admission to time of discharge from hospital.	2 years
Time from symptom onset to final diagnosis	Comparison before and after AI technique.	3 months
Hospitalization cost for PE using Markov model	Comparison before and after AI technique.	2 years

Collaborators and Investigators

• Sponsor: China-Japan Friendship Hospital
• Investigators: Study Director: Zhenguo Zhai, PhD, China-Japan Friendship Hospital

Study record dates

Study Major Dates

• Study Start (Actual): September 1, 2010
• Primary Completion (Estimated): September 1, 2027
• Study Completion (Estimated): September 1, 2028

Study Registration Dates

• First Submitted: June 15, 2024
• First Submitted That Met QC Criteria: July 28, 2024
• First Posted (Actual): July 29, 2024

Study Record Updates

• Last Update Posted (Actual): August 28, 2024
• Last Update Submitted That Met QC Criteria: August 27, 2024
• Last Verified: August 1, 2024

More Information

Terms related to this study

• Keywords: imaging; differential diagnosis; risk stratification; early diagnosis; prognosis assessment
• Additional Relevant MeSH Terms: Cardiovascular Diseases; Vascular Diseases; Respiratory Tract Diseases; Lung Diseases; Embolism and Thrombosis; Hypertension; Embolism; Hypertension, Pulmonary; Pulmonary Embolism
• Other Study ID Numbers: CHOICE

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: Information gathered for this study will not be disclosed to any other person or entity, or for other research.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No