Accuracy of automated measurement system for pleural effusion on chest CT

Accuracy of automated segmentation and measurement system for pleural effusion on chest CT: A prospective single-arm clinical trial in patients undergoing thoracentesis

Status

Recruiting

Phases

Unknown

Study type

Interventional

Source

CRIS

Registry ID

KCT0006683

Enrollment

Registered

2021-10-22

Start date

2021-10-25

Completion date

Unknown

Last updated

2021-11-02

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

None listed

Interventions

Medical Device : Participants will undergo low-dose chest CT before and after the thoracentesis that is performed for diagnostic or therapeutic purposes.

Eligibility

Sex/Gender

All

Inclusion criteria

Inclusion criteria: 1. Individuals with an age of 19 years or older 2. Patients undergoing thoracentesis for clinical (diagnostic or therapeutic) purposes

Exclusion criteria

Exclusion criteria: 1. Clinically unstable patients who cannot undergo chest CT (Decision by the radiologist who performs thoracentesis) 2. Patients with severe artifact on chest CT in which analysis by the artificial intelligence system is impossible. 3. Patients with pleural metastasis in which differentiation pleural nodule or mass between pleural effusion is impossible on precontrast CT. 4. Pregnant women

Design outcomes

Primary

Measure	Time frame
Absolute error between the difference of the AI-measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Absolute error between the difference of the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side	—

Secondary

Measure	Time frame
Percentage error between the difference of the AI-measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Percentage error between the difference of the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;The proportion of CT scans that is needed to edit the AI results manually .;Time spent to manually edit the AI results.;Correlation coefficient between the difference of the AI-measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Correlation coefficient between the difference of the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Absolute error between the AI-measured amounts of pleural effusion before and after thoracentesis at the contralateral side;Absolute error between the AI-measured/manually-corrected measured amounts of pleural effusion before and after thoracentesis at the contralateral side;Percentage error between the AI-measured amount of pleural effusion before and after thoracentesis at the contralateral side;Percentage error between the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis at the contralateral side;Correlation coefficient between the AI-measured amount of pleural effusion before and after thoracentesis at the contralateral side;Correlation coefficient between the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis at the contralateral side	—

Measure

Time frame

Percentage error between the difference of the AI-measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Percentage error between the difference of the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;The proportion of CT scans that is needed to edit the AI results manually .;Time spent to manually edit the AI results.;Correlation coefficient between the difference of the AI-measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Correlation coefficient between the difference of the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis and the amount of drained pleural fluid at the drain side;Absolute error between the AI-measured amounts of pleural effusion before and after thoracentesis at the contralateral side;Absolute error between the AI-measured/manually-corrected measured amounts of pleural effusion before and after thoracentesis at the contralateral side;Percentage error between the AI-measured amount of pleural effusion before and after thoracentesis at the contralateral side;Percentage error between the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis at the contralateral side;Correlation coefficient between the AI-measured amount of pleural effusion before and after thoracentesis at the contralateral side;Correlation coefficient between the AI-measured/manually-corrected measured amount of pleural effusion before and after thoracentesis at the contralateral side

—

Countries

Korea, Republic of

Contacts

Public ContactEui Jin Hwang

Seoul National University Hospital

ken921004@hotmail.com+82-2-2072-3617

Outcome results