Ultrasound in Med. & Biol., Vol. 42, No. 2, pp. 447–450, 2016 Copyright Ó 2016 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter
http://dx.doi.org/10.1016/j.ultrasmedbio.2015.09.013
d
Original Contribution ARE RESULTS OF 4-D ULTRASOUND ANGIOGRAPHY EXAMINATIONS DEPENDENT ON THE DOPPLER TECHNOLOGY APPLIED? COMPARISON OF RESULTS OBTAINED FROM AN IN VIVO MODEL MAREK J. KUDLA,* ANDRZEJ LOS,y and JUAN LUIS ALCAZARz * Clinical Department of Oncological Gynecology, Chair of Woman’s Health, Medical University of Silesia, Katowice, Poland; y Department of Gynecology and Obstetrics, Regional Specialist Hospital No. 3, Rybnik, Poland; and z Department of Obstetrics and Gynecology, Clinica Universidad de Navarra, Medical School, University of Navarra, Pamplona, Spain (Received 19 March 2015; revised 3 September 2015; in final form 14 September 2015)
Abstract—We aimed to evaluate the agreement of results obtained by 4-D spatio-temporal image correlation (STIC) angiography with two options of Doppler technology (power Doppler [PD] and high-definition flow [HDF]) from an ovary as an in vivo model. Thirty-eight ovaries were recorded by trans-vaginal ultrasound examination in the first part of the menstrual cycle. Two STIC sequences (4-D HDF and 4-D PD) were stored. Volumetric pulsatility index, volumetric resistance index and volumetric systolic/diastolic index for each of these sequences were calculated, and their mean values were compared and correlated. Agreement between 4-D HDF and 4-D PD was assessed using the intra-class correlation coefficient. Intra-class correlation coefficients for all three indices were high, but 95% confidence intervals and limits of agreement were wide. We conclude that both 4-D power Doppler and 4-D high-definition flow may be used for calculating volumetric pulsatility index, volumetric resistance index and volumetric systolic/diastolic index from a STIC sequence, at least in ovaries used as an in vivo model. However, values obtained by both methods cannot be used interchangeably. (E-mail: marekkudla@ hotmail.com) Ó 2016 World Federation for Ultrasound in Medicine & Biology. Key Words: 4-D, Ultrasound, Doppler, Ovary, Agreement.
sequence can be calculated. Additionally, two other indices can also be calculated, the volumetric resistance index (vRI) and the volumetric systolic/diastolic index (vS/D) (Kudla and Alcazar 2012). These indices have been developed as an attempt at overcoming some of the problems of standard 3-D PD indices, such as the influence of machine settings and attenuation (Martins et al. 2013). A STIC sequence may be obtained using power Doppler (4-D PD) (Martins et al. 2011) or highdefinition flow (4-D HDF) (Kudla and Alcazar 2010). Some authors have advocated that power Doppler should be used instead of high-definition flow for calculating the above-mentioned indices because power Doppler is more widely available and can render results that are more robust (Martins et al. 2013) while HDF was reported to have better axial resolution, fewer blooming artifacts and improved sensitivity to small vessels compared to power Doppler and color Doppler (Kim et al. 2008; Alcazar and Kudla 2010). However, to the best of our knowledge, no comparison between these two methods had been performed so far. Therefore, we considered it worthwhile to check whether results of
INTRODUCTION The use of 3-D power Doppler (PD) or 3-D high-definition flow (HDF) combined with the option of spatio-temporal image correlation (STIC) allows for the assessment of changing values of standard 3-D indices (vascularization index [VI], flow index [FI] and vascularization-flow index [VFI]) throughout the cardiac cycle (Kudla and Alcazar 2010). This option, called 4-D Doppler angiography, resulted in a new 4-D index for the assessment of tissue volumetric vascularization changes, the so-called volumetric pulsatility index (vPI) (Martins et al. 2011). This index is derived from the analysis of a 4-D (STIC) record (Kudla and Alcazar 2010; Martins et al. 2011). A 4-D (STIC) data set is a sequence of 3-D volumes set in time order (time being the fourth dimension) during one heart cycle. From each of these 3-D volumes the value of VI can be calculated; then, vPI from the entire 4-D
Address correspondence to: Marek J. Kudla, Clinical Department of Oncological Gynecology, Medical University of Silesia, Panewnicka 65, 40-751, Katowice, Poland. E-mail:
[email protected] 447
448
Ultrasound in Medicine and Biology
vPI, vRI and vS/D values could be similar when derived from 4-D HDF or 4-D PD sequences. Their similarity would support clinical usefulness and reliability of 4-D angiography. The aim of our study was to assess the agreement of these two methods of 4-D angiography (4-D power Doppler and 4-D high-definition flow) in an in vivo model for calculating vPI, vRI and vS/D indices.
METHODS This was a prospective observational study. Fortythree consecutive, healthy, regularly ovulating women who reported for routine gynecologic check-up were invited to participate in this study. Inclusion criteria were as follows: 1. Ovary with no visible pathologies (e.g., simple or endometrial cysts, dermoids etc.). 2. Ovaries of typical size, enabling the total volume of the ovary to be recorded in a single 4-D (STIC) record as well as in a 3-D scan. 3. Distance between the surface of the transducer head and the surface of ovaries accepted for the evaluation could not exceed 10 mm. If both ovaries fulfilled the inclusion criteria, the one with better visibility was accepted. 4. All women had regular menstruations and were not taking any hormonal treatment at the time of being included in the study. Institutional Review Board consent was obtained for this study.
Volume 42, Number 2, 2016
All of the women underwent transvaginal ultrasound examination during the early stage of the first part of the menstrual cycle (day 4–9) to assess ovarian vascularization with 3-D and 4-D Doppler technology (STIC). A pre-determined protocol (Kudla and Alcazar 2012) was applied on Voluson E8 machine equipped with a volumetric endo-vaginal RIC 5-9 Mhz probe (GE Healthcare, Zipf, Austria). For this study, first a 4-D HDF sequence was acquired and stored for further analysis. Immediately after the 4-D HDF had been acquired, Power Doppler was activated and a 4-D PD (STIC) record was taken (PD and HDF settings: PRF 5 0.6 kHz; Gain 5 21; Freq. mid.; Qual. high, WMF low 1; STIC settings: vol. angle 90 , acquisition time 15 sec). Calculations were performed using dedicated software (4-D View, Version 10.2, GE Medical Systems, Zipf, Austria) on a personal computer. The first step in calculations for every 4-D (STIC) record was to activate STIC loop in order to work with a sequence of 3-D volumes (Fig. 1). Ovaries were virtually cut out from each 3-D record (rotation step 5 9 ). Histograms for these samples were automatically calculated for each 3-D volume of the complete 4-D (STIC) sequence. Mean VI value was calculated based on the results from all volumes. vPI was then determined according to the following formula: vPI 5 (maximum VI value2minimum VI value)/mean VI value. Next, vRI was determined according to the following formula: vRI 5 (maximum VI value2minimum VI value)/maximum VI value. Finally, vS/D was determined according to the following formula: vS/D 5 maximum VI value/minimum VI value. This assessment was done first with the 4-D HDF and then with 4-D PD sequences (Kudla and Alcazar 2010;
Fig. 1. STIC approach summarized diagrammatically. A 4-D loop of 3-D images representing tissue vascularization changes over one cardiac cycle. Simplified diagram based on four images only; the real STIC record contains more.
Agreement between two Doppler methods for calculating 4-D vascular indices d M. J. KUDLA et al.
Table 1. Results of volumetric indices measured by each method
449
DISCUSSION
Parameter
Mean
SD
Range
vPI 4-D PD vPI 4-D HDF vRI 4-D PD vRI 4-D HDF vS/D 4-D PD vS/D 4-D HDF
0.320 0.350 0.260 0.300 1.410 1.540
0.19 0.23 0.14 0.16 0.31 0.46
0.05–0.750 0.05–0.95 0.05–0.55 0.05–0.65 1.05–2.82 1.05–2.21
SD 5 standard deviation; vPI 5 volumetric pulsatility index; PD 5 power Doppler; HDF 5 high-definition flow; vRI 5 volumetric resistance index; vS/D 5 volumetric systolic/diastolic index.
Martins et al. 2011). All acquisitions and calculations were performed by a single examiner (MJK). The Kolmogorov–Smirnov test was used to assess normal distribution of vPI, vRI and vS/D derived from 4-D HDF and 4-D PD sequences. Agreement between both methods was estimated by calculating the intra-class correlation coefficient (ICC) (Bland and Altman 1999). Ninety-five percent limits of agreement were calculated as described by Bland and Altman (1999). Systematic bias between the two methods was determined by calculating the 95% confidence interval for the mean difference (mean difference 6 two standard errors). If 1 (when results were expressed as a ratio) lay within this interval, no bias was assumed to exist. A p value , 0.05 was considered to be statistically significant.
RESULTS Forty-three women were invited to participate. Thirty-eight women gave their oral informed consent and were included (mean age: 27.8 y, range 24–35 y); five did not fulfill the criteria or refused to participate and were therefore excluded. All indices had normal distribution. Data of measurements for each method are shown in Table 1. Mean difference between methods for all three indices and their limits of agreement are shown in Table 2. Agreement was good for vPI (ICC 5 0.850), vRI (ICC 5 0.821) and vS/D (ICC 5 0.817; Table 2).
In this study, agreement of vPI, vRI and vS/D obtained from 4-D PD and 4-D HDF sequences were analyzed. We found that agreement was good according to the ICC values obtained. However, 95% confidence interval for ICC and limits of agreement were quite wide for all measurements. This means that values might actually vary significantly depending on the method used. Our study has some limitations. As all acquisitions and calculations were done by a single examiner, and thus neither intra-observer nor inter-observer variability was assessed. However, recent research has shown that calculation of vPI, vRI and vS/D is reliable (Miyague et al. 2013). We decided to use an ovary in its early, first part of the cycle as a model in our study. Advantages of this model include ability to place the head of a highresolution transducer (endo-vaginal probe) in immediate vicinity of the examined organ, meaning that the distance between the transducer and the ovary was almost the same among all patients; therefore, there was no need to change system settings to obtain a good record of blood supply changes. The ability to visualize the entire organ in a single record, with its clearly visible boundaries, is of crucial importance for cohesion of results, and so an ovary proved to be the most appropriate choice. To the best of our knowledge, no other study has been reported so far making a similar assessment as ours. Therefore, we cannot compare our results with any other study. This means that our data need to be confirmed or refuted in further studies using other or similar in vivo models. To date, only one study has been reported using these indices in clinical grounds. We observed that these indices were significantly lower in women with polycystic ovaries compared with controls, indicating a higher vascularization in women with polycystic ovaries (Alcazar and Kudla 2012). However, the clinical value for using these 4-D vascular indices still needs to be established. CONCLUSION Both 4-D power Doppler and 4-D high-definition flow may be used for calculating vPI, vRI and vS/D
Table 2. Differences between the mean for each method and the intra-class correlation coefficient Parameter vPI vRI vS/D
Mean value measured*
Mean difference*
95% CI of mean difference
Limits of agreement
ICC
95% CI of ICC
0.334 (0.049–0.954) 0.285 (0.050–0.650) 1.479 (1.051–2.823)
0.03 (20.37 to 0.42) 0.04 (20.25 to 0.30) 0.13 (20.61 to 0.81)
20.02 to 0.08 20.01 to 0.08 20.03 to 0.22
20.27 to 0.33 20.19 to 0.27 20.48 to 0.73
0.850 0.821 0.817
0.711–0.992 0.655–0.907 0.648–0.905
CI 5 confidence interval; ICC 5 intra-class correlation coefficient; vPI 5 volumetric pulsatility index; vRI 5 volumetric resistance index; vS/D 5 volumetric systolic/diastolic index. * For all 76 measurements, range in parentheses.
450
Ultrasound in Medicine and Biology
from a STIC sequence, at least in ovaries used as in vivo model. However, values obtained by both methods cannot be used interchangeably.
REFERENCES Alcazar JL, Kudla MJ. Three-dimensional vascular indices calculated using conventional power Doppler and high-definition flow imaging: are there differences? J Ultrasound Med 2010; 29:761–766. Alcazar JL, Kudla MJ. Ovarian stromal vessels assessed by spatiotemporal image correlation-high definition flow in women with polycystic ovary syndrome: A case-control study. Ultrasound Obstet Gynecol 2012;40:470–475. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135–160. Kim SH, Lee JM, Kim YJ, Lee JY, Han JK, Choi BI. High-definition flow Doppler ultrasonographic technique to assess hepatic vasculature compared with color or power Doppler ultrasonography: Preliminary experience. J Ultrasound Med 2008;27:1491–1501.
Volume 42, Number 2, 2016 Kudla MJ, Alcazar JL. Spatiotemporal image correlation using highdefinition flow: A new method for assessing ovarian vascularization. J Ultrasound Med 2010;29:1469–1474. Kudla MJ, Alcazar JL. Spatiotemporal image correlation with spherical sampling and high-definition flow: New 4-dimensional method for assessment of tissue vascularization changes during the cardiac cycle: reproducibility analysis. J Ultrasound Med 2012;31:73–80. Martins WP, Welsh AW, Lima JC, Nastri CO, Raine-Fenning NJ. The ‘‘volumetric’’ pulsatility index as evaluated by spatiotemporal imaging correlation (STIC): A preliminary description of a novel technique, its application to the endometrium and an evaluation of its reproducibility. Ultrasound Med Biol 2011;37: 2160–2168. Martins WP, Welsh AW, Falkensammer P, Raine-Fenning NJ. Spatiotemporal imaging correlation (STIC): Technical notes about STIC triggering and choosing between Power Doppler or high-definition color flow. Ultrasound Med Biol 2013;39:549–550. Miyague AH, Raine-Fenning NJ, Polanski L, Martinez LH, Araujo Junior E, Pavan TZ, Martins WP. Assessing repeatability of 3D Doppler indices obtained by static 3D and STIC power Doppler: A combined in vivo/in vitro flow phantom study. Ultrasound Obstet Gynecol 2013;42:571–576.
