from the ClinicalTrials.gov record
The purpose of this study is to supplement prior Kosmos Trio Guidance performance testing conducted under ECHO-020 Kosmos Guidance, Grading, and Capture Performance Evaluation Protocol. The Kosmos Trio Guidance algorithm is designed to provide guidance to help novice scanners acquire images and videos during ultrasound scanning of three cardiac views (Parastenal Long Axis, Apical 4 Chamber, and Apical 2 Chamber) with Kosmos Trio. The study will divide the guidance cues not evaluated sufficiently in the ECHO-020 study into four categories: 1) corrective guidance, 2) incorrect scanning window, 3) incorrect probe orientation, and 4) low confidence prediction. For each guidance cue evaluated in this study, expert annotators will review two clips: one suboptimal view, collected when the probe is positioned at an incorrect location; and one guided view, captured after the sonographer followed the probe motion cue displayed by the system. Annotators will review each pair of clips and respond indicating either i) Yes, the guided clip demonstrates the appropriate correction from the suboptimal view; ii) No, the guided clip does not demonstrate the appropriate correction from the suboptimal view; or iii) Unevaluable, the guided clip is technically inadequate for assessment. For each category of guidance cues, the outcome measure will be the rate at which annotators respond "Yes." Finally, this study will also evaluate the Kosmos Trio Auto-Capture algorithm's ability to detect and capture a diagnostic-quality image. Annotators will rate clips captured by Kosmos Trio Auto-Capture on an ACEP-based image quality scale. The outcome measure will be the image quality score. The study design involves the compilation of de-identified retrospective cardiac clips acquired using the Kosmos diagnostic ultrasound system. This data was collected for two previous studies; no additional data will be collected from previous study participants for the purposes of this study.