@inproceedings{4a1cf68264084b5fbdd8d8a70a60a040,
title = "In-vitro validation of a novel model-based approach to the measurement of arterial blood flow waveforms from dynamic digital x-ray images",
abstract = "We have developed a blood flow waveform shape model using principal component analysis (PCA) and applied this to our existing concentration-distance curve matching technique for the extraction of flow waveforms from dynamic digital x-ray images. The aim of the study was to validate the system using a moving-vessel flow phantom. Instantaneous recording of flow from an electromagnetic flow meter (EMF) provided the “gold standard” measurement. A model waveform was constructed from 256 previously recorded waveforms from the EMF using PCA. Flow waveforms were extracted from parametric images derived from dynamic x-ray data by finding the parameters of the shape model that minimized the mean value of our cost function. The computed waveforms were compared to the EMF recordings. The model-based approach produced narrower limits of agreement with the EMF data than our previously developed algorithms and, in the presence of increasing noise in the parametric images, it out-performed the other algorithms.",
author = "K. Rhode and G. Ennew and T. Lambrou and A. Seifalian and D. Hawkes",
note = "cited By 3; Medical Image Computing and Computer-Assisted Intervention : 4th International Conference, MICCAI 2001 ; Conference date: 14-10-2001 Through 17-10-2001",
year = "2001",
doi = "10.1007/3-540-45468-3_35",
language = "English",
volume = "2208",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer ",
pages = "291--300",
editor = "Niessen, {W J} and Viergever, {M. A}",
booktitle = "Medical Image Computing and Computer-Assisted Intervention",
}