Inter- and intra-observer variability analysis of completely automated cIMT measurement software (AtheroEdgeTM) and its benchmarking against commercial ultrasound scanner and expert Readers

Authors:
Luca Saba;Filippo Molinari;Kristen M. Meiburger;U. Rajendra Acharya;Andrew Nicolaides;Jasjit S. Suri
Affiliations:
-;-;-;-;-;-
Venue:
Computers in Biology and Medicine
Year:
2013

Citing 4
Cited 0

Review: A state of the art review on intima-media thickness (IMT) measurement and wall segmentation techniques for carotid ultrasound

Computer Methods and Programs in Biomedicine
Inter-Greedy Technique for Fusion of Different Segmentation Strategies Leading to High-Performance Carotid IMT Measurement in Ultrasound Images

Journal of Medical Systems
Completely Automated Multiresolution Edge Snapper—A New Technique for an Accurate Carotid Ultrasound IMT Measurement: Clinical Validation and Benchmarking on a Multi-Institutional Database

IEEE Transactions on Image Processing
Ultrasound IMT measurement on a multi-ethnic and multi-institutional database: Our review and experience using four fully automated and one semi-automated methods

Computer Methods and Programs in Biomedicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

The purpose of this study was to evaluate the measurement error and inter- and intra-observer variability of completely off-line automated and semi-automated carotid intima-media thickness (cIMT) measurement software (AtheroEdge(TM)). Two hundred carotid ultrasound images from 50 asymptomatic women were analyzed. AtheroEdge(TM) was benchmarked against a commercial system (Syngo, Siemens) using automated and semi-automated modes. The measurement error and inter- and intra-observer variability of AtheroEdge(TM) were tested using three readings. The measurement error of AtheroEdge(TM) compared to the commercial software was 0.002+/-0.019mm (r=0.99) in the automated mode and -0.001+/-0.004mm in the semi-automated mode (r=0.99). The measurement error of AtheroEdge(TM) compared to the mean value of the three expert Readers (cIMT bias) for the automated and semi-automated methods was -0.0004+/-0.158mm and -0.008+/-0.157mm, respectively. The Figure-of-Merit was 99.8% and 99.9% when compared to the commercial ultrasound scanner (using the automated and semi-automated method, respectively) and was 99.9% and 98.9% when compared to the mean value of the three expert Readers. Regarding inter- and intra-observer variability, the intra-class correlation coefficient of the three independent users using the semi-automated AtheroEdge(TM) was 0.98. AtheroEdge(TM) showed a measurement performance comparable to the commercial ultrasound scanner software and the expert Readers' tracings. AtheroEdge(TM) belongs to a class of automated systems that could find application in processing large datasets for common carotid arteries, avoiding subjectivity in cIMT measurements.