Multi-center, Randomized Retrospective, Crossover design, Superiority, Pivotal study to evaluate the efficacy of ‘SKH-BCH-001’ software assisting the diagnosis decision regarding to identification and detection of intracranial hemorrhage on Brain CT scans
- Conditions
- Diseases of the circulatory system
- Registration Number
- KCT0006734
- Lead Sponsor
- Ajou University Hospital
- Brief Summary
Not available
- Detailed Description
Not available
Recruitment & Eligibility
- Status
- Completed
- Sex
- All
- Target Recruitment
- 296
All of the following selection criteria must be satisfied to participate in this clinical trial.
1.Brain CT images of men and women over 19 years of age.
2.Brain CT image taken for examination of suspected intracranial hemorrhage (ICH).
3.Images that can secure clinical information* of the image provider.
*Clinical information: Image acquisition year, number of slices, gender, date of birth, patient type (emergency, inpatient, outpatient)
? Except for personally identifiable information.
4.Images with a score of 3 or higher in the image quality survey (refer to the image quality survey criteria table)
Image quality survey criteria table
5-point scaleCriteria
1Non diagnostic due to excessive noise or artifacts
2Diagnosis questionable due to excessive noise or artifacts, moderate decrease in confidence
3Diagnostic with moderate but acceptable noise or artifacts
4Mild noise, no change in confidence
5Routine diagnostic image quality
Ref) Fletcher et al. AJNR Am. J. Neruoradiol (2019), Fletcher et al. Radiology (2015)
5.0.5 mm = z-axial slice thickness (mm) = 5 mm
If any of the following exclusion criteria are met, images cannot be used for this clinical trial.
1.Brain CT imaging using a contrast medium
2.Brain CT image that has already been used as a training set or validation set for the development of medical devices for clinical trials.
? In order to collect the images for this clinical trial as independent data apart from the training set or validation set, Seoul National University Hospital will use brain CT images excluding images taken from January 2009 to December 2015 as of the date of image recording, and Ajou University hospitals will use images taken on all dates, but for images from April 2004 to April 2020 as of the date of image recording, only the data, whose independence is confirmed by checking the image list provided to the institution for image identification, will be used as images for this clinical trial.
3.In case the principal investigator determines that participation in the clinical trial is inappropriate because it affects the efficacy assessment result or for other reasons (the specific reason is recorded in the case record).
Study & Design
- Study Type
- Interventional Study
- Study Design
- Not specified
- Primary Outcome Measures
Name Time Method Clinical Sensitivity for diagnosis of image provider (%) ,Clinical Specificity for diagnosis of image provider (%)
- Secondary Outcome Measures
Name Time Method Clinical Sensitivity for CT image slice reading (%),Clinical Specificity for CT image slice reading (%) ,Accuracy for image provider diagnosis and CT image slice reading (%),Concordance rate of intracranial hemorrhage type (%),Concordance rate of intra-group image reading results;[Assessment of clinical performance of test device] ?AUC of ROC Curve (Area Under the Curve of Receiver Operating Characteristic Curve) for the algorithm of the test device. ?Sensitivity for algorithm of test device (Clinical Sensitivity, %) ?Specificity for algorithm of the test device (Clinical Specificity, %)