Ultrasound image quality assessment: a framework for evaluation of clinical image quality

Ultrasound in Medicine & Biology

Hansen

et al. 2015

Self Cite

Abstract-The purpose of the study was to perform a clinical comparison of synthetic aperture sequential beamforming tissue harmonic imaging (SASB-THI) sequences with a conventional imaging technique, dynamic receive focusing with THI (DRF-THI). Both techniques used pulse inversion and were recorded interleaved using a commercial ultrasound system (UltraView 800, BK Medical, Herlev, Denmark). Thirty-one patients with malignant focal liver lesions (confirmed by biopsy or computed tomography/magnetic resonance) were scanned. Detection of malignant focal liver lesions and preference of image quality were evaluated blinded off-line by eight radiologists. In total, 2,032 evaluations of 127 image sequences were completed. The sensitivity (77% SASB-THI, 76% DRF-THI, p 5 0.54) and specificity (71% SASB-THI, 72% DRF-THI, p 5 0.67) of detection of liver lesions and the evaluation of image quality (p 5 0.63) did not differ between SASB-THI and DRF-THI. This study indicates the ability of SASB-THI in a true clinical setting.

Section: Equipment and Data Acquisitionmentioning

confidence: 99%

“…They were informed that some of the images contained malignant focal liver lesions. An in-house developed software program (IQap) was used for the evaluation (Hemmsen et al 2010). Images obtained with the two techniques were shown separately, resulting in 254 evaluations by each radiologist for a total of 2,032 evaluations.…”

Section: Image Evaluationmentioning

confidence: 99%

Clinical Evaluation of Synthetic Aperture Harmonic Imaging for Scanning Focal Malignant Liver Lesions

Brandt

Ultrasound in Medicine & Biology

Hansen

et al. 2015

Self Cite

“…The simulation results are validated through measurements using a ProFocus scanner (BK Medical, Herlev, Denmark) equipped with a research interface for acquisition of beamformed RF data described in [8], [9]. Transmit-and receive delay profile, apodization, and excitation waveform are preset standardized settings from the manufacturer.…”

Section: B Measurementsmentioning

confidence: 99%

Compound imaging using Synthetic Aperture Sequential Beamformation

Jensen

2011 IEEE International Ultrasonics Symposium

et al. 2011

Self Cite

Abstract-Synthetic Aperture Sequential Beamforming (SASB) is a technique with low complexity and the ability to yield a more uniform lateral resolution with range. However, the presence of speckle artifacts in ultrasound images degrades the contrast. In conventional imaging speckle is reduced by using spatial compounding at the cost of a reduced frame rate. The objective is to apply spatial compounding to SASB and evaluate if the images have a reduced speckle appearance and thereby an improved image quality in terms of contrast compared to ordinary SASB. Using the simulation software Field II, RF data are acquired for a phantom with cysts at different sizes and scattering levels. 192 scanlines are recorded for five steering angles (0, ±2, ±4 degrees) using a 192 element linear array transducer. SASB is performed for each angle using a rectangular grid in the second stage beamformation. After envelope detection the five second stage images are added to form the compounded image. Using a ProFocus scanner and the 8804 linear array transducer (BK Medical, Herlev, Denmark) measurements of a phantom containing water filled cysts are obtained to validate the simulation results. The setup is the same as in the simulations and SASB second stage beamformation data are processed offline for each of the five angles. Contrast-to-noise ratio (CNR) and speckle-to-noise ratio (SNR) are extracted for the compounded image and the reference image (ordinary SASB). CNR was calculated for the simulated cysts at depths of 40, 50, 60, 70 and 80 mm. On average the CNR was improved by 33.2% compared to the values obtained from the reference image. For regions of increasing depth SNR was on average increased by 9.3%. Results from the simulation were confirmed by calculations on the measured data. CNR of cysts at depths from 18 to 78 mm with a separation of 10 mm was on average improved by 45.9%. On average an improvement of 16.6% in SNR was obtained. The calculations along with visual inspection revealed larger improvements in deeper regions, and the compounded image for the measured phantom showed a 3 mm diameter cyst not detectable in the reference image. Compounding applied to SASB improves CNR and SNR results in images with a reduced speckle appearance. This was shown for simulations and confirmed on measured data.

“…The measurement setup consist of a 2202 ProFocus ultrasound scanner (BK-Medical, Herlev, Denmark) connected to a standard pc through a X64-CL Express camera link (Dalsa, Waterloo, Ontario, Canada) and has previously been described in [5].…”

Section: B Measurementsmentioning

confidence: 99%

Simulation of high quality ultrasound imaging

2010 IEEE International Ultrasonics Symposium

Kortbek²,

Nikolov³

et al. 2010

Self Cite

Abstract-This paper investigates if the influence on image quality using physical transducers can be simulated with an sufficient accuracy to reveal system performance. The influence is investigated in a comparative study between Synthetic Aperture Sequential Beamformation (SASB) and Dynamic Receive Focus ( Further, it is shown that SASB successfully maintains a constant LR though depth at FWHM, and is a factor of 2.3 better than DRF at 80 mm. However, when using SASB the LR at FWOHM is affected by non-ideal element responses. Introducing amplitude and phase compensation, the LR at FWOHM improves from 6.3 mm to 4.7 mm and is a factor of 2.2 better than DRF. This study has shown that individual element impulse response, phase, and amplitude deviations are important to include in simulated system performance evaluations. Furthermore, it is shown that SASB provides a constant LR through depth and has improved resolution and contrast compared to DRF.