High-frequency ultrasound imaging to assess fetal pancreas: a promising application

Abstract

Objectives: The purpose of this study is to establish a comprehensive reference range of quantitative characteristics of the fetal pancreas using a high-frequency transducer, and assess the growth and development of the fetal pancreas.

Methods: Pregnant women referred to a tertiary center were recruited to undergo a detailed fetal scan from 16 to 37 weeks. We evaluated the visualization rate of the fetal pancreas with high-frequency and low-frequency transducers and measured the head, neck, body, tail, circumference, area, and abdominal circumference(AC) of the fetal pancreas at different gestational ages(GA) with the high-frequency transducer. Regression analysis was used to analyze the relationship between biological parameters and GA and AC.

Results: During the time periods of 16⁺¹∼21⁺⁶weeks and 22⁺¹∼27⁺⁶weeks, the visualization rate of high-frequency transducers was higher compared to low-frequency transducers (83.33% vs 45% and 95.65% vs 70%, respectively). However, in the third trimester of pregnancy, the performance of the two transducers was similar (70.37% vs 74.07% for 28⁺¹∼33⁺⁶weeks and 41.67% vs 53.85% for 34⁺¹∼37⁺⁶weeks). The head, neck, body, and tail as well as the circumference and area of the pancreas were significantly positively correlated with GA (R²＝0.87, 0.94, 0.92, 0.92,0.96, and 0.92) and AC (R²＝0.87, 0.93, 0.91, 0.93,0.96, and 0.92).

Conclusions: The high-frequency transducer was utilized to establish the normal reference, which can be used to evaluate normal pancreatic development and may help in the accurate diagnosis of fetal pancreatic abnormalities.

Keywords:

• Fetal pancreas
• high-frenquency transducer
• ultrasound imgaing

Introduction

The pancreas is a sophisticated organ formed in the early embryo with endocrine and exocrine functions. Abnormal development of the pancreas is often associated with pregnancy complications, genetic syndromes, and congenital malformations, such as gestational diabetes mellitus, intrauterine growth restriction, Beckwith-syndrome, and annular pancreas [1–4]. The lack of attention given to prenatal pancreatic imaging and diagnosis may be attributed to the poor resolution of ultrasound technology. The pancreas is difficult to distinguish in comparison to the surrounding tissue. Improved visualization of the fetal pancreas would facilitate the detection of abnormalities and aid in prognosis consultation.

Previous research showed the value of using fetal pancreas circumference as an additional tool for examining fetal abnormalities that coexist with or result from abnormal congenital development of the pancreas [1]. However, whether a single parameter can reflect the overall morphology of the pancreas is controversial. The development of the pancreas during the embryonic period is regulated by a variety of factors, and if its complex developmental mechanisms are disrupted or malfunctioned, many congenital abnormalities, including morphological variations, can result [5–7]. Therefore, only using the pancreas circumference to evaluate the pancreas with different shapes is not comprehensive. We hope to provide a reference range for the head, neck, body, and tail of the fetal pancreas, as well as the circumference and area for gestational ages(GA). If the general morphology of the pancreas is abnormal, it can be diagnosed through a thorough assessment of the entire organ with a variety of parameters.

Conventional curvilinear transducers, although adequately penetrating, are poor at visualizing the fetal pancreas. In the recent literature on the fetal pancreas, the overall satisfactory visualization rate using a curvilinear transducer was only 61.6% [1]. We need to choose an appropriate frequency range that not only has enough penetration but also clearly shows the fetal pancreas. The previous high-frequency transducer has inadequate penetration and cannot display the fetus in a deeper position, let all one the pancreas. With the technological development of ultrasound transducers, the high-frequency transducer can also reach lower frequencies, laying technical support for clearly displaying the fetal pancreas.

The objective of this study was to establish comprehensive reference ranges for the fetal pancreas using a high-frequency transducer (3–12MHz), evaluate the reliability of intraobserver and interobserver measurements, and investigate the correlation between fetal pancreas growth and other fetal growth parameters.

Materials and methods

Participants

We performed a cross-sectional prospective study of routine prenatal ultrasonography in pregnant women presenting between 16 and 37 weeks from February 2021 to December 2022. According to the cross-sectional rules, each fetus was included only once. If the fetal pancreas was measured several times during pregnancy, only one measurement was randomly selected, while the remaining measurements were not used to fit the statistical models. The study protocol was approved by the Ethics Committee of Jinan Maternal and Child Health Care Hospital (2023-1-004). All women were informed and agreed to the evaluation of the fetal pancreas.

The inclusion criteria for this study were:1) regular menstrual cycle, gestation age based on a known last menstrual period and confirmed by ultrasound estimation of crown-rump length in the first trimester, 2) singleton pregnancy,3)no chromosomal abnormalities,4)no obvious fetal malformations, and 5)no maternal complications. All fetuses were followed up to six weeks after delivery.

Philips EPIQ 7 C equipped with a curvilinear transducer (1-5MHZ) and a high-frequency transducer (3–12MHz) were used. Participants were randomly assigned to the high and low-frequency transducer groups according to 1:1. The participants in the two groups were divided into four subgroups according to GA: 16⁺¹∼21⁺⁶, 22⁺¹∼27⁺⁶,28⁺¹∼33⁺⁶, and 34⁺¹∼37⁺⁶, and the visualization rates in each group were estimated and compared.

Visualization and measurement of fetal pancreas

Considering that the pancreas is situated in front of the splenic vein, we first used a curvilinear transducer to display the splenic vein with Color Doppler, enlarge the image, and attempt to distinguish the entire pancreas. Six o‘clock is ideal for the spine. The pancreas was identified and imaged in the same way using a high-frequency transducer. Pancreatic visualization was assessed by two senior sonographers(T.T.S. and H.H.C.) with over 10 years of expertise.

The pancreas is located between the stomach and the spine, it is slightly more echogenic than the liver. Freeze images and measure the fetal pancreas head, neck, body, tail, circumference, and area. The measurement positions were located in front of the inferior vena cava(head), superior mesenteric vein(neck), superior mesenteric artery(body), and left kidney(tail), respectively. Measure the largest diameter from the anterior to the posterior edge of the pancreas (Figure 1A). The circumference and area were measured by the freehand trace (Figure 1B). Each measurement was repeated three times, and the average of the three values was calculated. These data were also used to assess inter- and intra-observer reliability.

Figure 1. Visualization of 24⁺⁶W fetal pancreas using high-frequency transducers (3–12MHz). Measurement method of fetal pancreas head, neck, body, and tail(a). the measurement positions were located in front of the inferior vena cava(head,

), superior mesenteric vein(neck,

), superior mesenteric artery(body,

) and left kidney(tail,

), respectively. Freehand tracing of pancreas circumference(B). the pancreas was traced from the left caudal edge to the right edge, and the circumference and area were measured.

AO: abdominal aorta; IVC: inferior vena cava; LK: left kidney; RK: right kidney; SMA: superior mesenteric artery;SMV: superior mesenteric vein; STO: stomach; UV: umbilical vein

Statistics

SPSS software version 26.0 and Medcalc Version 20.0.27 were used for all the statistical analyses. Continuous variables were described as mean ± standard deviation, while categorical variables were described with percentages. Chi-square was used for pairwise comparison of categorical variables. The normality of the distribution of the data was determined by the Shapiro–Wilk test. The intra-and inter-observer reliability was assessed by the intraclass correlation coefficient (ICC) and Bland-Altman plots. The agreement was considered slight when the ICC was ≤ 0.2, fair when 0.2 < ICC ≤ 0.4, moderate when 0.4 < ICC ≤ 0.6, substantial when 0.6 < ICC ≤ 0.8, and excellent with the ICC > 0.8. Regression analysis was used to analyze the correlation of the fetal pancreas head, neck, body, tail, circumference, and area with abdominal circumference(AC) and GA. A P value of <0.05 was considered statistically significant.

Results

From February 2021 to December 2022, there were 834 participants in total, of which 44 were unable to determine the exact gestational age, 29 multiple pregnancies, 14 fetal morphological abnormalities, 5 fetal chromosomal abnormalities, and 49 cases with maternal complications. Finally, a total of 693 participants were included in the study. 499 ultrasound images can visualize the head, neck, body, and tail of the fetal pancreas, with a total visualization rate of 71.9% (499/693).

To evaluate the performance of high-frequency transducers, a total of 160 participants were randomly selected and divided into two groups: a high-frequency group (n = 80) and a low-frequency group (n = 80), maintaining a 1:1 ratio. The general characteristics of participants are shown in Table 1, and there was no statistical difference between the two groups in clinical characteristics. Table 2 and Figure 2 showed the satisfactory visualization rates of the fetal pancreas by the high-frequency and the low-frequency transducers at different gestational weeks. The visualization rate of the fetal pancreas with a high-frequency transducer was significantly higher than that with a low-frequency at 16–27 weeks, and the difference was statistically significant. With the progress of pregnancy, the visualization rate of the two methods was similar, and the difference was not statistically significant.

Figure 2. Visualization rate of the fetal pancreas at different gestational weeks. Before 27 weeks, the visualization rate of the pancreas with a high-frequency transducer was significantly higher than that with a low-frequency. After 27 weeks, there was no significant difference between the two transducers.

Table 1. The general characteristics of participants.

	High-frequency (n = 80)	Low-frequency (n = 80)	P
GA(weeks)	26.75 ± 5.99	26.79 ± 6.06	NS
Maternal age(years)	29.56 ± 5.18	29.75 ± 4.75	NS
Gender,male(n)	45 (56.25%)	39 (48.75%)	NS
Fetal AC(mm)	225.81 ± 61.96	224.45 ± 63.83	NS
BMI (kg/m^2)	23.19 ± 3.38	22.69 ± 3.07	NS

GA: gestation age; AC: abdominal circumference; BMI: body mass index; NS:not significant.

Table 2. Satisfactory visualization of the fetal pancreas at different gestational weeks.

GA	High-frequency	Low-frequency	RR(95% CI)	p
16^+1∼21^+6	83.33% (15/18)	45% (9/20)	1.852 (1.094–3.136)	0.02*
22^+1∼27^+6	95.65% (22/23)	70% (14/20)	1.366 (1.012–1.844)	0.036*
28^+1∼33^+6	70.37% (19/27)	74.07% (20/27)	0.827 (0.579–1.183)	NS
34^+1∼37^+6	41.67% (5/12)	53.85% (7/13)	0.774 (0.335–1.788)	NS

GA: gestation age; CI: confidence interval; *statistically significant difference; NS: not significant.

To assess inter- and intra-observer reliability, a total of 51 randomly selected images were used. Intra- and inter-observer agreement was performed, with a correct 95% LoA for the mean bias estimates. The reliability of both intra- and inter-observer measurements was found to be excellent, with ICC values ranging from 0.85 to 0.99 and 0.90 to 0.98, respectively (Table 3).

Table 3. The inter- and intra-observer variability for the fetal pancreas.

	Interobserver			Intraobserver
	Mean Bias	95%LoA	ICC(95%CI)	Mean Bias	95%LoA	ICC(95%CI)
Head	0.09	−0.71–0.89	0.98 (0.96–0.99)	0.02	−0.61–0.65	0.99 (0.98–0.99)
Neck	0.03	−0.55–0.61	0.94 (0.90–0.97)	−0.01	−0.54–0.53	0.95 (0.91–0.97)
Body	−0.09	−0.57–0.39	0.96 (0.94–0.98)	−0.13	−0.49–0.23	0.97 (0.95–0.98)
Tail	−0.01	−0.52–0.49	0.96 (0.93–0.98)	−0.02	−0.33–0.29	0.99 (0.97–0.99)
C	0.17	−0.59–0.93	0.98 (0.96–0.99)	0.08	−0.57–0.72	0.99 (0.98–0.99)
Area	−0.11	−0.48–0.27	0.90 (0.83–0.94)	−0.07	−0.57–0.43	0.85 (0.75–0.91)

C: circumference; LoA: limits of agreement; ICC: intraclass correlation coefficient; CI: confidence interval.

Values for the fetal pancreas head, neck, body, tail, circumference, and area with GA were provided in Table 4. The anteroposterior diameter of the head, neck, body, and tail as well as the circumference and area of the pancreas were significantly positively correlated with GA (R²＝0.87, 0.94, 0.92, 0.92,0.96, and 0.92) and AC (R²＝0.87, 0.93, 0.91, 0.93,0.96, and 0.92) (Figure 3,4)

Figure 3. The relationship between the fetal pancreas head, neck, body, tail, circumference, area, and GA. The relevant statistical results are in the upper left corner of the image.

Figure 4. The relationship between the fetal pancreas head, neck, body, tail, circumference, area, and AC. The relevant statistical results are in the upper left corner of the image.

Table 4. Fetal pancreas thickness, circumference, and area at different gestational ages.

GA (W)	n	Head (mm)	Neck (mm)	Body (mm)	Tail (mm)	C (mm)	Area (mm^2)
16	20	3.92 ± 0.22	1.34 ± 0.10	2.32 ± 0.12	2.31 ± 0.11	41.5 ± 4.1	39 ± 4
17	20	4.18 ± 0.33	1.52 ± 0.11	2.46 ± 0.17	2.45 ± 0.14	42.7 ± 3.4	40 ± 4
18	20	4.68 ± 0.33	1.56 ± 0.15	2.57 ± 0.20	2.54 ± 0.18	45.8 ± 2.7	47 ± 2
19	20	5 ± 0.29	1.69 ± 0.17	2.82 ± 0.22	2.76 ± 0.18	46.8 ± 3.3	50 ± 3
20	20	5.37 ± 0.29	1.79 ± 0.21	2.83 ± 0.26	2.86 ± 0.2	52.8 ± 2.5	56 ± 5
21	20	5.6 ± 0.26	1.88 ± 0.20	3.06 ± 0.09	3.11 ± 0.1	55.2 ± 1.7	65 ± 5
22	21	5.91 ± 0.22	1.99 ± 0.22	3.01 ± 0.14	3.1 ± 0.16	62.9 ± 1.5	77 ± 6
23	24	6.86 ± 0.6	2.18 ± 0.17	3.1 ± 0.24	3.26 ± 018	65.7 ± 3.6	85 ± 8
24	38	6.86 ± 0.59	2.28 ± 0.19	3.19 ± 0.21	3.23 ± 0.14	67.2 ± 2.9	87 ± 10
25	22	7.1 ± 0.33	2.38 ± 0.14	3.55 ± 0.29	3.54 ± 0.27	72.7 ± 4	96 ± 11
26	22	7.72 ± 0.34	2.63 ± 0.12	3.97 ± 0.21	3.97 ± 0.17	75.1 ± 2.8	108 ± 10
27	24	7.99 ± 0.28	2.7 ± 0.11	4.14 ± 0.18	4.1 ± 0.18	81.7 ± 2.7	109 ± 16
28	23	8.01 ± 0.72	2.72 ± 0.18	4.14 ± 0.25	4.14 ± 0.23	81.8 ± 2.3	122 ± 14
29	21	8.54 ± 0.86	2.97 ± 0.17	4.21 ± 0.36	4.29 ± 0.22	83.1 ± 3.2	126 ± 14
30	20	8.61 ± 0.75	2.99 ± 0.12	4.26 ± 0.32	4.32 ± 0.43	84.4 ± 3.2	131 ± 18
31	20	8.81 ± 0.91	3.22 ± 0.14	4.57 ± 0.24	4.6 ± 0.28	88.7 ± 2.2	144 ± 14
32	23	8.68 ± 1.07	3.36 ± 0.15	4.54 ± 0.3	4.62 ± 0.3	92 ± 3.4	156 ± 12
33	20	9.21 ± 1.03	3.48 ± 0.19	4.81 ± 0.14	4.72 ± 0.28	95.8 ± 4.2	157 ± 8
34	20	9.41 ± 0.38	3.59 ± 0.17	4.88 ± 0.2	4.93 ± 0.2	100.4 ± 4.9	166 ± 8
35	20	9.75 ± 0.31	3.65 ± 0.16	5.11 ± 0.14	5.08 ± 0.13	103.3 ± 3	170 ± 13
36	20	10.44 ± 0.54	3.81 ± 0.21	5.08 ± 0.17	5.37 ± 0.15	105.4 ± 4.4	164 ± 20
37	21	11.23 ± 0.71	3.92 ± 0.15	5.39 ± 0.3	5.42 ± 0.17	112.6 ± 2.9	170 ± 15

GA: gestation age; C: circumference.

Discussions

The existing literature on prenatal imaging of the fetal pancreas is limited, with most studies focusing on obstructive deformities such as annular pancreas [3,4,8]. Evaluation of the size and echogenicity of the pancreas has revealed a potential association between the size and hyperechogenicity of the fetal pancreas and gestational diabetes [9,10]. Additionally, clear visualization of the pancreas can provide important clues for diagnosing tumors in the fetal abdomen, such as pancreatic hamartoma and abdominal lymphatic malformation [11,12]. Previous clinical practice struggled with locating abdominal tumors due to poor organ visualization. Some studies also emphasize the significance of studying underdeveloped pancreases in fetuses, as it can guide genetic diagnosis and facilitate accurate genetic counseling [13–15]. Cospain et al. described fetuses with pancreatic hypoplasia associated with specific CNOT1 variants [13]. Gilboa et al. evaluated four fetuses with Beckwith–Wiedemann syndrome based on genetic testing and found that three of them had pancreatic circumference > 90 percentile [14]. Body-Bechou et al. found that TCF2 deficiency leads to pancreatic hypoplasia, indicating that this gene is crucial for pancreatic development [15]. Therefore, obtaining detailed anatomical information about the size and shape of the fetal pancreas is highly valuable for prenatal diagnosis, counseling, targeted genetic testing, and post-natal care. Unfortunately, diagnosing pancreatic abnormalities in the fetus is challenging even for experienced sonographers due to unsatisfactory visualization of the fetal pancreas. When the fetus is in the prone position, the acoustic shadow of the spine will obscure the visualization of the pancreas, but when the fetus is in the supine position, because the pancreas is a retroperitoneal organ with a deep anatomical position, it is not easy to obtain a satisfactory visualization image. Therefore, the pancreas has always been a neglected organ, and there have been few studies on the fetal pancreas in the past.

In recent years, high-frequency transducers with a wide frequency range have been used in fetal pancreas visualization, which can perfectly solve the problem of pancreatic imaging. In our study, the visualization rate of the high-frequency image of the pancreas was significantly higher than that of the low-frequency in the second trimester, and there was no significant difference between the two in the third trimester. Hata et al. found that the visualization rate was 80% at 20–23 weeks of gestation, and decreased to 38.36% after 36 weeks of gestation [16]; Hill et al. reported that the overall visualization rate of the fetal pancreas was 51.68% between 14 and 42 weeks of gestation [17]; Kivilevitch et al. disclosed that the overall visualization rate of 297 cases was 61.6% [1]. Our research showed that the visualization rate of the pancreas can be increased to 70.38% by using the high-frequency transducer between 16 and 37 weeks. The high-frequency transducer with a wider frequency range enables reassessment of the neglected fetal pancreas.

Our results also indicated that visualization of the fetal pancreas could be advanced up to 16 weeks with the high-frequency transducer when the low-frequency transducer was not visible. However, in the third trimester, the two transducers were similar in fetal pancreas visualization, and the low-frequency visualization rate was slightly higher than the high-frequency. Xiao et al. also found similar findings that the anatomical structure of the pancreas is most easily recognizable in the 19–23 gestational age group [18]. This may be due to the gradual growth of the pancreas with the increase of the gestational age, and the low-frequency transducer can distinguish the pancreas from the surrounding tissues. Secondly, the penetration of high frequency in late pregnancy is affected by the depth of the fetus, resulting in a decline in the visualization rate. Therefore, we suggested that the best time for diagnosing abnormal fetal pancreas was the second trimester, especially at 22–27 weeks, when the anatomical structure and morphology of the pancreas can be imaged.

In our preliminary study, fetal pancreas imaging was unsatisfactory before 16 weeks, and the visualization rate dropped dramatically after 37 weeks. Therefore, we chose the gestational age to 16–37 weeks, which is wider than previous studies. Kivilevitch et al. proposed the reference of pancreatic circumference for normal fetuses from 19 to 36 weeks in European and American countries [1]; there is no relevant report on the reference of the fetal pancreas in Asians. For the first time, we have established a normal reference range for the fetal pancreas in an Asian population, useful in diagnosing congenital abnormalities of the fetal pancreas. We not only measured the pancreas head, neck, body, tail, and circumference, but also assessed the pancreas area, and provided a reference value of the normal fetal pancreas area from 16 to 37 weeks, which was more than the reference in previous studies. These parameters were chosen to evaluate the morphology of the fetal pancreas because there are various congenital variations of the pancreas, such as congenital complete or partial agenesis of the dorsal pancreas, abnormal protrusions of the pancreas, heterotopic pancreas, circumportal pancreas, intrapancreatic accessory spleen, and annular pancreas, etc [5–7].

Conclusion

Although challenging and not suitable for routine prenatal anatomical scans, evaluating the fetal pancreas is valuable in diagnosing anatomical abnormalities and genetic conditions related to pancreatic structure and function. Recognizing the normal shape and size of the pancreas is essential for detecting abnormalities, and the clear visualization of the pancreas using high-frequency transducers offers a useful approach for prenatal diagnosis and perinatal consultation.

Disclosure statement

No potential conflict of interest was reported by the author(s). This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

The data that support the findings of this study are available from the corresponding author, [G.W.T.], upon reasonable request.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.