On the utility of subharmonic microbubble signals to detect portal hypertension

• Main Authors: Dave, Jaydev K., Halldorsdottir, Valgerdur G., Eisenbrey, John R., Merton, Daniel A., Ji-Bin Liu, Jian-Hua Zhou, Hsin-Kai Wang, Suhyun Park, Dianis, Scott, Chalek, Carl L., Feng Lin, Thomenius, Kai E., Brown, Daniel B., Forsberg, Flemming
• Format: Conference Proceeding
• Language: English
• Subjects: Acoustics; Catheters; Hypertension; noninvasive pressure estimation; portal hypertension; portal vein pressure; Portals; Radio frequency; Shape; Subharmonic aided pressure estimation; Ultrasonic imaging; ultrasound contrast agents
• ISSN: 1051-0117
• DOI: 10.1109/ULTSYM.2012.0163

In this work, the efficacy of subharmonic aided pressure estimation to detect portal hypertension (PH; i.e., portal vein (PV) pressures > 6-10 mmHg) is investigated. A Logiq 9 ultrasound scanner (GE Healthcare, Milwaukee, WI) with a 4C probe (transmit/receive: 2.5/1.25 MHz) was configured to obtain unprocessed radiofrequency (RF) data post pulse inversion. Fourteen canines were scanned. The main PV was accessed following a mid-line abdominal incision. Acute PH was induced using Gelfoam administration (8 canines; low-flow PH model) and using an arteriovenous fistula (6 canines; femoral artery to PV; high-flow PH model). A Millar pressure catheter introduced in the main PV indicated reference PV pressures. The RF data corresponding to 4 transmit cycles at 20% and 40% incident acoustic power (IAP) levels and pressure catheter data were acquired synchronously during infusion of Sonazoid (GE Healthcare, Oslo, Norway; 0.015 μl/kg/min) microbubbles, before and after inducing PH (5 s acquisitions; n = 3). The mean subharmonic amplitude from all acquired frames was extracted from the RF data corresponding to PV location. Linear regression analyses with leave-one-out cross validation technique were used to obtain PV pressures from the subharmonic amplitude (data from 3 canines was excluded due to extraneous conditions). The resulting PV pressures were compared to pressure catheter data. Overall, mean errors were -0.15 mmHg (p = 0.92) and -0.09 mmHg (p = 0.95) with 20 and 40% IAP levels, respectively. For baseline PV pressures, mean errors were 3.50 mmHg (p = 0.11) and 3.64 mmHg (p = 0.10), while for PH pressures the mean errors were -3.80 mmHg (p = 0.07) and -3.83 mmHg (p = 0.06), with 20 and 40% IAP levels, respectively. The sensitivity, specificity and accuracy of detecting moderate through severe PH (cut-off value: 16 mmHg PV pressure) in this cohort (n = 9) were 78%, 69% and 73% for 20% IAP and 78%, 77% and 77% for 40% IAP. Ambient pressure modulated subharmonic signals can be used for detecting PH in canines and thus, might be used to diagnose patients in the future.