Evaluation of the LVivo Image Quality Scoring (IQS)

Overview

The study includes two parts: Part 1: * 100 examinations of patient referred for echo evaluation, containing clips that were acquired from the 4CH apical view were collected retrospectively. Each study includes 10 clips that represent typical user scanning errors * Offline evaluation of the system (by batch processing) shall be performed by comparing the system output to preliminary quality tagging by experienced sonographers Part 2: * Live scans of apical 4CH clips of patients with indication for POCUS examination will be performed by POC physicians * LVivo IQS shall be used (on Lumify) during the scan for patients that meet inclusion criteria until 50 exams will be collected. 3 sec of each scan shall be saved, and Image quality score (IQS) shall be documented * Saved scans shall be reviewed by an expert physician to determine whether they are clinically interpretable

Part 1 Total of 424 examination were collected from two medical centers. Participants that were referred to routine echo examinations and signed an informed consent were enrolled to the study. The image acquisition was done using ultrasound devices available in the medical centers. In addition Lumify ultrasound device was supplied by the sponsor. Each examination contained 10 clips, each representing a different scanning scenario where the image acquisition contained transitions between common user "acquisition error" to correctly acquired images. The examinations in DICOM format and echo reports were collected after anonymization, the patient details were documented in a study log that was stored in the medical center. The collected clips were tagged for image quality by experienced sonographers. The distribution of the collected data according to ultrasound devices, age, gender and LV function was analyzed. The data was randomly split into training and clinical validation sets, such that the proportion of data according to data distribution was maintained and the data set for the clinical evaluation will contain 100 patient examinations. The LVivo IQS algorithm will be applied to the clinical validation set automatically by batch processing. The image quality score provided by the LVivo IQS will be compared to the image quality tagging by the sonographers Part 2 The trial is designed as a prospective multicenter study, enrolling 50 patients at two medical centers. In each medical center Lumify ultrasound devices on which the LVivo IQS demo software will be installed will be provided for image acquisition. The image acquisition will be done by POC medical doctors, at different stages of their internship, that received POCUS training during their medical studies and are using ultrasound devices in accordance with their routine workflow. The scans will be done in different POC units including ER, ICU and internal departments.