Exploring the Application Efficacy of Artificial Intelligence (AI) Diagnostic Tools in Medical Imaging (MI) of Respiratory(R) Infectious (I) Disease (D)

Not Applicable

Recruiting

• Conditions: Artificial Intelligence; Respiratory Infectious Diseases; Medical Imaging
• Interventions: Other: Artificial Intelligence-based medical imaging interpretation

• Registration Number: NCT06553911
• Lead Sponsor: Huashan Hospital

Brief Summary

The early identification and severe warning of acute respiratory infectious diseases are of paramount importance. Utilizing effective means to make correct diagnoses of the source of infection at an early stage is the premise of all effective measures. AI-MID is a research initiative that uses artificial intelligence tools to assist in the clinical medical imaging diagnosis of respiratory diseases, aiming to reduce the time doctors spend reviewing images, increase work efficiency, and enhance the sensitivity and specificity of pneumonia detection, thereby improving the detection rate of pneumonia at the grassroots level. This approach facilitates precise prevention, accurate diagnosis, and precise treatment.

Detailed Description

Not available

Study Timeline

• Study Start: 2022-04-01
• Status Verified: 2024-08
• Study First Submitted: 2024-08-07
• Study First Submitted QC: 2024-08-12
• Last Update Submitted: 2024-08-12
• Study First Posted: 2024-08-14
• Last Update Posted: 2024-08-14
• Primary Completion: 2025-12-01
• Study Completion: 2026-12-01

Recruitment & Eligibility

• Status: RECRUITING
• Sex: All
• Target Recruitment: 2000

Inclusion Criteria

1. 1-90 years old, gender not specified.
2. Exhibits symptoms of respiratory tract infection
3. Must have etiological examination results
4. Must have imaging data;

Exclusion Criteria

1. Severe artifacts in medical images
2. Clinical diagnosis indicates concurrent pulmonary edema
3. Dual review results in unclear diagnosis or potential misdiagnosis
4. Other situations that may cause difficulties in reading the films, or as determined by the researcher, the study participant is deemed unsuitable for enrollment.

Study & Design

• Study Type: INTERVENTIONAL
• Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
Artificial Intelligence-based medical imaging interpretation group	Artificial Intelligence-based medical imaging interpretation	Using clinical information, imaging data, and corresponding etiological results of the study participants, an AI diagnostic tool is established to specifically recognize patients' chest medical imaging and construct corresponding diagnostic conclusions.

Primary Outcome Measures

Name	Time	Method
Evaluating the Diagnostic Efficacy of Artificial Intelligence Diagnostic Tools in Medical Imaging of Respiratory Infectious Diseases	2 years	To evaluate the diagnostic efficacy of computer-aided detection (CAD) software in the identification of pulmonary infections, the study will employ the following methods: Imaging Criteria: Experienced radiologists will interpret the medical imaging of study participants, serving as the imaging standard.; Computer-Aided Detection: Concurrently, the CAD software will analyze the participants' medical imaging to generate diagnostic results.; Efficacy Assessment: The accuracy and consistency of the CAD software will be evaluated by comparing its interpretations with the diagnoses made by the radiologists.

Secondary Outcome Measures

Name	Time	Method
Utilizing artificial intelligence tools for early identification and severe warning of respiratory infectious diseases	2 years	By integrating the medical imaging of study participants with the corresponding respiratory pathogen detection results, these data will be used as the training set input into the AI diagnostic tool, enabling it to undergo deep learning. This process will establish an AI diagnostic tool based on pathogen imaging. After completing the data collection for both retrospective and prospective study sections, we plan to evaluate the disease progression and prognosis of the study participants based on survival analysis and predictive modeling. By integrating clinical data and imaging data, we aim to enhance the accuracy and precision of the prognostic assessment model. The model will be continuously optimized according to the changes in the conditions of study participants enrolled over different time periods.

Trial Locations

Locations (1)

Huashan Hospital; 🇨🇳 Shanghai, China