Automated sonographic evaluation of testicular perfusion
Contrast-enhanced ultrasound (US) imaging is potentially applicable to the investigation of vascular disorders of the testis. We investigated the ability of two automated computer algorithms to analyse contrast-enhanced pulse inversion US data in a rabbit model of unilateral testicular ischaemia and...
Saved in:
Published in | Physics in medicine & biology Vol. 51; no. 14; pp. 3419 - 3432 |
---|---|
Main Authors | , , , , , |
Format | Journal Article |
Language | English |
Published |
England
IOP Publishing
21.07.2006
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Contrast-enhanced ultrasound (US) imaging is potentially applicable to the investigation of vascular disorders of the testis. We investigated the ability of two automated computer algorithms to analyse contrast-enhanced pulse inversion US data in a rabbit model of unilateral testicular ischaemia and to correctly determine relative testicular perfusion: nonlinear curve fitting of the US backscatter intensity as a function of time; and spectral analysis of the intensity time trace. We compared (i) five metrics based on the algorithmic data to testicular perfusion ratios obtained with radiolabelled microspheres, a reference standard; (ii) qualitative assessment of the US images by two independent readers blinded to the side of the experimental and control testes to the radiolabelled microsphere perfusion ratios; and (iii) results of the algorithmically-derived metrics to the qualitative assessments of the two readers. For the curve fit method, the algorithmically-derived metrics agreed with the reference standard in 54% to 68% of all cases. For the spectral method, the results agreed in 70% of all cases. The two readers agreed with the reference standard in 40% and 35% of all cases, respectively. These results suggest that automated methods of analysis may provide useful information in the assessment of testicular perfusion. |
---|---|
Bibliography: | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
ISSN: | 0031-9155 1361-6560 |
DOI: | 10.1088/0031-9155/51/14/010 |