VALIDATION OF A SENSITIVITY PERFORMANCE INDEX TEST PROTOCOL AND EVALUATION OF COLOUR DOPPLER SENSITIVITY FOR A RANGE OF ULTRASOUND SCANNERSINTRODUCTIONMETHODSSensitivity performance indexPenetration depth flow phantomUltrasound scannersSensitivity performance index test protocolValidation of sensitivity performance indexThe effect of instrument settings on the sensitivity per-formance indexTransducer frequencyClutter filter settingOutput powerPersistence settingEvaluation of the sensitivity performance index for a range of US scannersRESULTSValidation of sensitivity performance indexThe effect of instrument settings on the sensitivity per-formance indexEvaluation of the sensitivity performance index for a range of US scannersDISCUSSIONCONCLUSIONSREFERENCESdoi:10.1016/j.ultrasmedbio.2004.09.005● Original ContributionVALIDATION OF A SENSITIVITY PERFORMANCE INDEX TESTPROTOCOL AND EVALUATION OF COLOUR DOPPLER SENSITIVITYFOR A RANGE OF ULTRASOUND SCANNERSJACINTA E. BROWNE,* AMANDA J. WATSON,* PETER R. HOSKINS†and ALEX T. ELLIOTT**Department of Clinical Physics and Bioengineering, Western Infirmary Glasgow, Glasgow, UK; and†MedicalPhysics Unit, University of Edinburgh, Edinburgh, UK(Received 10 February 2004; revised 2 September 2004; accepted 16 September 2004)Abstract—The ability to detect flow is the most crucial aspect of an ultrasound (US) system because, if flowcannot be detected, no other aspect of performance matters. The objectives of this study were to validate aDoppler “sensitivity performance index,” a figure of merit, and to determine if it could be used to differentiatecolour Doppler sensitivity performance in scanners of varying complexity. The sensitivity performance index wasdeveloped to give a combined measure of related aspects of sensitivity, such as the lowest detectable velocity, thevessel size and the penetration depth. The colour Doppler sensitivity was evaluated objectively as the lowestdetectable velocity signal from the deepest achievable point within the Doppler sensitivity phantom free fromextraneous noise in a small diameter vessel (3.2 mm inner diameter). The effect of vessel size and mean velocityon the sensitivity performance index were investigated and it was found that the index was not proportional tovessel size, but this may be accounted for by considering the effect of the acoustic properties of the vessel material,the clutter filter and beam shape. The results obtained using flow phantoms with vessel sizes different from thoseused in this study are, therefore, not directly comparable to the results found in this study; however, a similartrend should be found in the results for the effect of control settings and a similar range of US scanners. It wasfound that the Doppler sensitivity performance index was a robust challenging test because none of the USscanners evaluated was capable of achieving the highest sensitivity performance index score, which would belimited by the lowest pump velocity and the deepest point of the vessel within the flow phantom. Therefore, thissuggests that this method of determining Doppler sensitivity performance is valuable in the absence of othersuitable methods, despite the fact that the relationship between the sensitivity performance index and vessel sizeis not proportional. Furthermore, use of the Doppler sensitivity performance index for the evaluation of a rangeof scanners demonstrated that curvilinear transducers have higher sensitivity performance indices than higher-frequency linear transducers, due to the higher achievable penetration depth. The effect of instrument settingswas assessed for two transducers, the 4C3 curvilinear general-purpose transducer (Aspen) and the PVM375ATcurvilinear general-purpose transducer (Nemio). The colour Doppler sensitivity performance was found to besignificantly dependent on the clutter filter setting and the output power setting for both transducers tested. Usersneed to be aware of the effect of these settings on the colour Doppler sensitivity performance of their US scannerwhen interpreting the clinical significance of the colour Doppler information. (E-mail: [email protected]) © 2004 World Federation for Ultrasound in Medicine & Biology.Key Words: Colour Doppler sensitivity, Doppler flow phantom, Sensitivity performance index, Penetration depth,Lowest detectable velocity, Instrument settings.INTRODUCTIONDoppler sensitivity is an important aspect of Doppler per-formance because, in clinical examinations, the questionoften arises as to whether or not blood flow is present, forexample: to distinguish between trickle blood flow (lowvelocity) and no blood flow in the case of the evaluation ofcarotid artery stenosis (Mattos et al. 1992); to verify paren-chymal perfusion by colour Doppler postrenal-transplant(Allan et al. 2000); or examine the venous system of thelower limb for venous insufficiency or thrombosis (Baxter1997). Therefore, sensitivity is one of the most criticalaspects of colour and power Doppler performance because,if blood flow cannot be detected, no other aspect of perfor-mance matters. Doppler sensitivity of ultrasound (US) scan-ners is defined in this study as a measure of the minimumAddress correspondence to: Dr. Jacinta Browne, Department ofNuclear Cardiology, Queen Elizabeth Building, Glasgow Royal Infir-mary, 16 Alexandra Parade, Glasgow G31 2ER UK. E-mail: [email protected] in Med. & Biol., Vol. 30, No. 11, pp. 1475–1483, 2004Copyright © 2004 World Federation for Ultrasound in Medicine & BiologyPrinted in the USA. All rights reserved0301-5629/04/$–see front matter1475signal strength (from different diameter vessels and fromdifferent depths within the body) for the lowest detectablevelocity that can be detected unambiguously in the absenceof an interference signal (AIUM 1993; Wang et al. 1992).Doppler sensitivity should not be confused with a Dopplersystem’s detectability, which is the system’s ability to dis-tinguish a small echo signal from an interference signal(AIUM 1993). The interference signal detected by theDoppler system is made up of system noise and clutter fromthe movement of blood vessel walls and slow-moving orstationary solid tissue within the sample volume. Dopplersensitivity is related to two factors: these are the detection oflow velocity blood flow (⬍ 10 cm s⫺1) and the presence ofa few moving scatterers. Therefore, both lowered Dopplerfrequency shift and