Introduction: The aim of this study was to use computerized analysis of the grayscale spectrum (histogram) to provide an objective assessment of the echogenicity of the fetal bowel. Moreover, we investigated the role of histogram analysis in the prenatal prediction of postnatal outcomes in fetuses with echogenic bowel (fetal echogenic bowel [FEB]). Methods: This is a single-center retrospective study including all fetuses with a diagnosis of echogenic bowel (FEB) in the mid-second trimester between 2015 and 2021. Ultrasound images were analyzed using ImageJ software. The mean of the grayscale histograms of the bowel, liver, and iliac/femur bone was obtained for each patient, and the ratio between these structures was used to overcome gain variations. We compared these values with those of a matched control group of singleton uncomplicated pregnancies and with a group of patients referred for FEB, where the FEB was not confirmed by the expert operator (FEB false-positive). Results: There was a statistically significant difference between bowel/liver and bowel/bone histogram ratios between the FEB group and the control groups (p < 0.05). Mean ratio cutoffs were provided for the diagnosis of FEB. Among the patients with confirmed FEB, both ratios were not able to discriminate the cases with adverse outcomes. In contrast, the presence of dilated bowel or other markers was associated with an adverse outcome. Conclusions: Histogram analysis may refine the diagnosis of FEB and reduce the number of false-positive diagnoses. For the prediction of the fetal outcome, the presence of additional features is clinically more significant than the degree of bowel echogenicity.

Fetal echogenic bowel (FEB) is a common, nonspecific sonographic finding, defined as increased echogenicity of the fetal intestine. It can be isolated or associated with other structural abnormalities. It is usually diagnosed in the second trimester with an incidence in the literature of 0.2–1.8% [1, 2]. Among fetuses diagnosed with FEB, the perinatal outcome differs considerably. In the majority of cases, FEB is a soft marker without clinical significance, but in some cases, it is associated with aneuploidies, congenital infections, cystic fibrosis (CF), and other intestinal anomalies [3‒5]. Aneuploidies primarily include trisomy 21 and sex chromosomes anomalies with a cumulative occurrence of 2.4% and 0.7%, respectively [3]. FEB is also related to fetal growth restriction (FGR), or it can be a sign of fetal blood swallowing due to intra-amniotic bleeding [6]. Intrauterine fetal demise (IUFD) is reported in 3.2% of cases, mainly due to severe FGR [3].

Among fetuses diagnosed with isolated FEB, 72.3% experience regression or disappearance of echogenicity at follow-up scans with no adverse outcomes. However, 27.7% do not regress, and in these cases, there is a chance of an adverse outcome [3].

Pathophysiology of hyperechogenicity is related to the underlying cause leading to the appearance of increased bowel echogenicity at US imaging. Bowel hypotonia, with increased water absorption from the meconium, seems to be the primary cause in fetuses with aneuploidies. Proximal mechanical bowel obstruction or viscous meconium is the hypothesized cause in CF and bowel atresia [7, 8]. The link between placental dysfunction and consequent FGR and hyperechogenic fetal bowel has been suggested to be the consequence of hemodynamic redistribution in favor of the brain and at the expenses of the musculoskeletal and splanchnic systems, leading to mesenteric ischemia [8, 9].

Assessment of FEB is traditionally subjective and is defined as a bowel of similar or greater echogenicity than the surrounding bone, mainly the iliac wing. Hyperechogenicity of the fetal bowel can be either diffuse or focal, uniform over a well-defined area, and is located primarily in the lower fetal abdomen and pelvis [10]. Transducer frequency and acoustic gain settings can influence the diagnosis as well as the BMI of the patient [11]. Grading systems are based on degrees that can be inferred by the comparison with echogenicity of the surrounding structures, such as the fetal liver and bone [12, 13].

However, an objective method to estimate the degree of the bowel echogenicity and to predict the postnatal outcome is currently lacking. The aim of this study was to investigate the use of histogram analysis as a possible objective method to classify the ultrasonographic appearance of FEB and its ability to predict the outcome.

This was a retrospective, single-center study. The data are a collection of 6-year cases referred for FEB to the Prenatal Diagnostics Department (PND) of the University Medical Centre Groningen (UMCG) between 2015 and 2021. All cases were selected by running a query in the Astraia imaging software program, version 1.27.2, Copyright 2000–2021, used to store all ultrasound scan data in our hospital. All US assessments were performed transabdominally. All examinations were performed using either Voluson E8 or E10 (General Electric, GE healthcare, Milwaukee, WI, USA).

All the ultrasound investigations of patients referred for FEB were performed by a fetal medicine specialist. The ultrasound scans of the control group were performed by expert sonographers in our center. Fetal bowel echogenicity was assessed according to Slotnick’s criteria [12]. The overall ultrasound time gain setting was reduced, and only cases whose bowel echodensity was similar to that of the bone was classified as FEB. Subjective grade assessment was recorded.

All grades of FEB were included in the study. Associated findings such as bowel dilatations, gallbladder abnormalities, calcifications, extra-abdominal abnormalities, FGR, and a decrease in amniotic fluid volume or the presence of intra-amniotic bleeding were recorded. A third trimester scan was repeated at around 28 weeks of gestation to monitor sonographic evolution of fetal bowel echogenicity and the appearance of the fetal bowel.

According to the local protocol, each patient was counseled by the fetal medicine specialist and a geneticist. Additional investigations were offered including amniocentesis for karyotyping, parental DNA CF testing, and maternal serological infection screening for toxoplasmosis, rubella, cytomegalovirus, herpes simplex virus, and parvovirus B19.

Outcomes considered favorable were term deliveries of a newborn with normal clinical examinations and meconium elimination. Adverse outcomes were all cases of termination of pregnancy, IUFD, anatomical defects or genetic disorders (CF, chromosomal anomalies), and surgical procedures performed within the first weeks of life as a result of intestinal pathology. FGR was defined according to the International Federation of Gynecology and Obstetrics (FIGO) guidelines [14].

The digital images were processed using the ImageJ software (National Institutes of Health, Bethesda, MD, USA). A squared region of interest of 10 × 10 pixels was placed on three regions: the most echogenic parts of the bowel, the fetal liver parenchyma, and the fetal bone, avoiding regions of acoustic shadow. The iliac wing and femur were used as reference for bones. By computerized analysis, the mean pixel intensity of the bowel, the liver, and the bone were calculated for each patient, and the ratio between these parameters was used to overcome gain and setting variations, the use of different probes, and inter-patient variability, as an indicator of the echogenicity. Only good quality images were used for the histogram evaluation, where the bowel, liver, and bone were displayed in the same image (shown in Fig. 1).

Fig. 1.

Digital image processed using the ImageJ software. A coronal view of the fetal abdomen was used, where all the interested structures are present. A square ROI of 10 × 10 pixels was used to take three different measurements of the bowel, the liver, and the iliac bone. The pink box shows where the ROIs were placed. Below: The three histograms created based on the ROIs, from left to the right, respectively, the bone, bowel, and liver analysis. ROI, region of interest.

Fig. 1.

Digital image processed using the ImageJ software. A coronal view of the fetal abdomen was used, where all the interested structures are present. A square ROI of 10 × 10 pixels was used to take three different measurements of the bowel, the liver, and the iliac bone. The pink box shows where the ROIs were placed. Below: The three histograms created based on the ROIs, from left to the right, respectively, the bone, bowel, and liver analysis. ROI, region of interest.

Close modal

All the patients with associated structural anomalies were excluded as well as patients with incomplete data, poor quality of US images, and patients lost to follow-up. We only included fetuses with isolated echogenic bowel at the mid-second trimester (18–22 weeks of gestation) with or without the presence of other soft markers. Fetal bowel echogenicity in the group of FEB-diagnosed fetuses (216 patients) was compared with two control groups: (1) a matched control group of fetuses from healthy, uncomplicated, singleton pregnancies (200 patients) and (2) a group including patients referred to our PND department for FEB in which the diagnosis was not confirmed by the specialist, FEB false-positive (FEBfp, 47 patients).

Statistical analysis was performed in collaboration with the Group of Medical Imaging Informatics at the UMCG using Python (version 3.7.9). To assess the difference between bowel/bone and bowel/liver ratios from the FEB group and the control groups, the Mann-Whitney U test was used. A logistic regression model and independent receiver operating characteristic analysis were used to find a cutoff for both bowel/bone and bowel/liver ratios to identify FEB patients. Within the FEB group patients, the one-way ANOVA test was used to determine if there was a difference between the grades of echogenicity subjectively assessed by the operators. We performed a Mann-Whitney U test to assess the difference between patients with normal and adverse outcomes. A p value <0.05 was considered to indicate statistical significance.

In a period of 6 years, a total of 598 patients were referred from peripheral ultrasound screening centers to the Prenatal Diagnosis Unit of the UMCG with a suspicion of FEB. In 76 patients (12.7%), the diagnosis was not confirmed by a specialist; 47 of these patients were in the mid-second trimester (FEBfp).

In 522 cases, the diagnosis was confirmed: in the first trimester in 8 patients (1.5%), in the second trimester in 493 patients (94.5%), and in the third trimester in 21 patients (4%). The largest group of FEB suspicions arose (N = 493) at the mid-second trimester anatomical US screening. Complete pregnancy and neonatal outcome and follow-up data were available in 216 patients (41.4%). The present study only refers to this group of patients with FEB (shown in Fig. 2).

Fig. 2.

Flowchart of patient’s inclusion for histogram analysis.

Fig. 2.

Flowchart of patient’s inclusion for histogram analysis.

Close modal

The mean maternal age at the time of diagnosis of FEB was 30.5 years (+/− 5.1 SD), and the mean BMI was 20.5 (+/−9.9 SD). Subjective assessment of the echogenicity of FEB was assigned as follows: grade 1 – 93 cases (43%), grade 2 – 109 cases (50.5%), and grade 3 – 14 cases (6.5%). Hyperechogenicity of the bowel resolved in the 3rd trimester in 173 cases (80%). Out of 216 patients, 181 newborns had an uneventful outcome (84.3%). Adverse neonatal outcomes and US findings are summarized in Table 1. Adverse pregnancy outcomes occurred in 34 cases (15.7%) and included 4 terminations of pregnancy (1.8%) and 5 (2.3%) IUFD. There were 3 cases of trisomy 21 (1.4%), all showing other associated US markers and 2 cases (0.92%) of genetic aberrations diagnosed by chromosomal microarray in fetuses with isolated FEB (one was a duplication 1q21.1, described in the literature associated with neurological sequelae; the other was a ALG6-CDG type1c, described as a congenital disorder of glycosylation). A total of 7 mothers had proven congenital infections (3.2%): 2 cases of CMV, 2 cases of toxoplasmosis, one parvovirus B19, one case of syphilis, and one of HSV-1.

Table 1.

Characteristics and sonographic findings at the moment of the diagnosis of FEB and the outcome of the study group

Outcome (N = 216)N total% totalFEB isolated, n (%)FEB + other soft markers, n (%)FEB + dilated bowel, n (%)TOP, n (%)IUD, n (%)Preterm delivery, n (%)
No anomalies 182 83.8 139 (76.7) 42 (23.3) 33 (18.2) − − 9 (5) 
Trisomy 21 1.4 − 3 (100) − − − 
Abnormal CMA 0.9 1 (50) 1 (50) − − 1 (50) 1 (50) 
CDG type1c − − − − 
Duplic. 1q21.1 − − − − 
CF 0.9 2 (100) − − 2 (100) − − 
TORCH infections 3.2       
 CMV  − − − − − 
 Toxo  − − − − − 
 Syphilis  − − − − − 
 Parvo B19  − − − − 
 HSV-1  − − − − 
FGR 17 7.8 9 (60) 6 (40) − − 4 (23.5) 5 (33.3) 
Intestinal anomalies 1.4 3 (100) − − − − 2 (66.6) 
Outcome (N = 216)N total% totalFEB isolated, n (%)FEB + other soft markers, n (%)FEB + dilated bowel, n (%)TOP, n (%)IUD, n (%)Preterm delivery, n (%)
No anomalies 182 83.8 139 (76.7) 42 (23.3) 33 (18.2) − − 9 (5) 
Trisomy 21 1.4 − 3 (100) − − − 
Abnormal CMA 0.9 1 (50) 1 (50) − − 1 (50) 1 (50) 
CDG type1c − − − − 
Duplic. 1q21.1 − − − − 
CF 0.9 2 (100) − − 2 (100) − − 
TORCH infections 3.2       
 CMV  − − − − − 
 Toxo  − − − − − 
 Syphilis  − − − − − 
 Parvo B19  − − − − 
 HSV-1  − − − − 
FGR 17 7.8 9 (60) 6 (40) − − 4 (23.5) 5 (33.3) 
Intestinal anomalies 1.4 3 (100) − − − − 2 (66.6) 

TOP, termination of pregnancy; CMA, chromosomal microarray.

The diagnosis of CF was confirmed in 2 fetuses (0.9%), both with isolated FEB. In 3 cases (1.4%), intestinal abnormalities were diagnosed at birth. These were all 3 cases of small bowel atresia without US signs of bowel dilation at the mid-second trimester scan (when the FEB was observed), showing signs of intestinal obstruction at the third trimester scan.

In 17 cases (7.8%), the fetuses developed FGR: in 12, growth restriction was diagnosed early in pregnancy (early FGR). Among these cases, there were 4 IUFD cases, and 5 were delivered preterm on fetal indication. In the group of patients with an adverse outcome, 10/34 patients (29.4%) had a persistent FEB in the third trimester.

The mean of the bowel/bone ratio was 0.88 (+/− 0.5 SD) in FEB patients, 0.63 (+/− 0.2 SD) in the healthy control group, and 0.67 (+/− 0.19) in the FEBfp group. The mean of the bowel/liver ratio was 3.42 (+/− 2.0 SD) in the FEB group, 1.65 (+/− 0.6 SD) in the healthy control group, and 1.84 (+/−0.68) in the FEBfp group.

Both bowel/bone and bowel/liver ratios were significantly different in the comparison between FEB and the healthy control group (p < 0.001) (shown in Fig. 3, 4). For the diagnosis of FEB according to the receiver operating characteristic curve, the optimal probability cutoff for the bowel/bone ratio was 0.731 (AUC = 0.840, sensitivity: 81%, specificity: 73%, accuracy: 77%), while the optimal probability cutoff for the bowel/liver ratio was 2.134 (AUC = 0.927, sensitivity: 78%, specificity: 89%, accuracy: 83%). The use of both bowel/bone and bowel/liver ratios gave a better AUC (AUC = 0.95, sensitivity: 88%, specificity: 91%, accuracy: 89%) (shown in Fig. 5). We also found a statistically significant difference between both bowel/bone and bowel/liver ratios between the FEB group and the FEBfp group (p < 0.001). Using both cutoffs on the FEBfb group, 10/47 patients (21.3%) would be classified as FEB with an accuracy of 79%.

Fig. 3.

Boxplot of the bowel/bone ratio for the FEB, the control group and the FEBfp group.

Fig. 3.

Boxplot of the bowel/bone ratio for the FEB, the control group and the FEBfp group.

Close modal
Fig. 4.

Boxplot of bowel/liver ratio for the FEB, the control group, and the FEBfp group.

Fig. 4.

Boxplot of bowel/liver ratio for the FEB, the control group, and the FEBfp group.

Close modal
Fig. 5.

ROC curve analysis of the bowel/bone ratio, the bowel/liver ratio, and both ratios to predict the FEB. ROC, receiver operating characteristic.

Fig. 5.

ROC curve analysis of the bowel/bone ratio, the bowel/liver ratio, and both ratios to predict the FEB. ROC, receiver operating characteristic.

Close modal

For groups classified subjectively as grades 1, 2, or 3, there was no significant difference in the bowel/bone and bowel/liver ratios among the three groups (shown in Fig. 6). In the real FEB group, there was no significant difference between the bowel/liver (p = 0.28) and the bowel/bone (p = 0.76) ratios in patients with a normal or adverse outcome (shown in Fig. 7).

Fig. 6.

Scatterplot of histogram analysis of subjective evaluation (grading) of FEB (grades 1, 2, or 3, considered as mild, moderate, or severe).

Fig. 6.

Scatterplot of histogram analysis of subjective evaluation (grading) of FEB (grades 1, 2, or 3, considered as mild, moderate, or severe).

Close modal
Fig. 7.

Scatterplot of histogram analysis of different outcomes in the group of patients with confirmed FEB.

Fig. 7.

Scatterplot of histogram analysis of different outcomes in the group of patients with confirmed FEB.

Close modal

Our study was designed to provide an objective instrument to establish and to quantify the real echogenicity of the fetal bowel. FEB is a fairly common finding at routine ultrasound screening in the second trimester, and this sonomarker entails a rather intensive and costly follow-up as it may be associated with chromosomal, genetic, infectious, placental, and other intestinal pathological causes. Not only the woman is offered a number of investigation and follow-up scans, in case no immediate cause is found, but also the psychological burden posed by this diagnosis is considerable. There is an urgent need to refine the diagnosis in order to reduce the incidence of false-positive cases and restrict the panel of diagnostic test to the truly high-risk cases.

The classical definition of bowel hyperechogenicity as being similar to that of the bone is susceptible to be subjective and operator-dependent. Surprisingly, of the two ratios defined by histogram analysis, the bowel/bone ratio was a weaker classifier than the bowel/liver ratio, whereas when combining both ratios, the sensitivity and specificity of the prediction increased. We provided two cutoffs to discriminate between real and normal echogenicity of the fetal bowel.

Using histogram analysis as an objective assessment method could reduce the number of referrals avoiding costs for follow-up tests along with women’s anxiety. The diagnosis of FEB is usually made in the second trimester, and it has significant implications for prenatal management and counseling. In our study, isolated FEB had a relatively good prognosis with an 84.3% chance of a favorable outcome. However, among the isolated FEB cases, 13.4% still had an adverse outcome, not related to chromosomal abnormalities. This is in agreement with earlier data from our center by Buiter et al. [2]. Histogram analysis and both bowel/bone and bowel/liver ratios can be used to assess the objective echogenicity, especially in cases where different operators disagree in their subjective judgment.

Grading systems are still being used in clinical practice, but recent studies confirm that the grade itself should no longer be considered as a prognostic factor [1]. This was confirmed in our study. To correctly classify the degree of FEB can be challenging as it is also known that the diagnosis of FEB is highly dependent on the ultrasound equipment used and the sonographer’s experience. In our study, there was no difference between subjectively assigned grades 1, 2, and 3, confirming the high inter-operator variability in assigning a specific grade. Considering the group of patients with confirmed FEB, no significance was found for the use of the ratios as a predictive factor for the adverse outcome. Next to histogram analysis, the presence of other US findings such as dilated bowel was the most important variable associated with a pathological outcome, especially in cases of anatomical bowel anomalies. In fetuses with intestinal atresia, this was only visible in the third trimester, confirming the importance of follow-up scans.

In summary, the use of histogram analysis of fetal bowel can distinguish between cases with echogenic bowel and false-positive or normal cases. We provided two different cutoff values for the bowel/bone and the bowel/liver ratios with a high sensitivity and specificity as the objective method to diagnose echogenic bowel in the second trimester. Between the two ratios, the bowel/liver ratio resulted to be the most accurate to identify cases with FEB. Histogram analysis alone was not of value for the prediction of pathological outcomes in fetuses with FEB as additional findings such as a dilated bowel and/or the presence of other soft markers are better predictors of the adverse outcome than the severity of echogenicity.

This was a retrospective study, making us dependent on completeness and accuracy of data collection. A considerable number of cases had to be excluded because of incomplete follow-up, leading to a relatively small number of pathological cases. A multicenter study could be undertaken to increase the study group and the cases with an adverse outcome to further validate these results.

The study protocol was reviewed by the Medical Ethics Review Board of the University Medical Center Groningen (METc UMCG) and approved as non-WMO (METc number 2021/399). The need for informed consent was waived by the METc considering the retrospective nature of the study.

All the authors declare that they have no conflict of interest.

This study received no specific funding.

Silvia Spinnato designed the study and was responsible for the data acquisition and for writing the manuscript. Alessia De Biase contributed to statistical analysis. Ayten Elvan-Taşpınar and Caterina M. Bilardo contributed to design of the study, supervision of the preparation of the manuscript, and editing.

Data are not publicly available due to ethical reasons. Further inquiries can be directed to the corresponding author.

1.
Ronin C, Mace P, Stenard F, Loundou A, Capelle M, Mortier I, et al. Antenatal prognostic factor of fetal echogenic bowel. Eur J Obstet Gynecol Reprod Biol. 2017;212:166–70.
2.
Buiter HD, Holswilder-Olde Scholtenhuis MAG, Bouman K, van Baren R, Bilardo CM, Bos AF. Outcome of infants presenting with echogenic bowel in the second trimester of pregnancy. Arch Dis Child Fetal Neonatal Ed. 2013;98(3):F256–9.
3.
D’Amico A, Buca D, Rizzo G, Khalil A, Silvi C, Makatsariya A, et al. Outcome of fetal echogenic bowel: a systematic review and meta-analysis. Prenat Diagn. 2021;41(4):391–9.
4.
Masini G, Maggio L, Marchi L, Cavalli I, Ledda C, Trotta M, et al. Isolated fetal echogenic bowel in a retrospective cohort: the role of infection screening. Eur J Obstet Gynecol Reprod Biol. 2018;231:136–41.
5.
Ruiz MJ, Thatch KA, Fisher JC, Simpson LL, Cowles RA. Neonatal outcomes associated with intestinal abnormalities diagnosed by fetal ultrasound. J Pediatr Surg. 2009;44(1):71–4; discussion 74–5.
6.
Goetzinger KR, Cahill AG, Macones GA, Odibo AO. Echogenic bowel on second-trimester ultrasonography. Obstet Gynecol. 2011;117(6):1341–8.
7.
Sepulveda W, Sebire NJ. Fetal echogenic bowel: a complex scenario. Ultrasound Obstet Gynecol. 2000;16(6):510–4.
8.
Nyberg DA, Dubinsky T, Resta RG, Mahony BS, Hickok DE, Luthy DA. Echogenic fetal bowel during the second trimester: clinical importance. Radiology. 1993;188(2):527–31.
9.
De Oronzo MA. Hyperechogenic fetal bowel: an ultrasonographic marker for adverse fetal and neonatal outcome?J Prenat Med. 2011;5(1):9–13.
10.
Copel J, Tutschek B, Platt L, D’Alton ME, Feltovich H, Gratacos E. Obstetric imaging: fetal diagnosis and care, 2017th ed. Elsevier.
11.
Lee HJ, Cho JY. Quantitative assessment of fetal bowel echogenicity: comparison of harmonic, compound, and fundamental sonographic images. J Clin Ultrasound. 2003;31(6):302–7.
12.
Slotnick RN, Abuhamad AZ. Prognostic implications of fetal echogenic bowel. Lancet. 1996;347(8994):85–7.
13.
Hill LM, Fries J, Hecker J, Grzybek P. Second-trimester echogenic small bowel: an increased risk for adverse perinatal outcome. Prenat Diagn. 1994;14(9):845–50.
14.
Melamed N, Baschat A, Yinon Y, Athanasiadis A, Mecacci F, Figueras F, et al. FIGO (International Federation of Gynecology and Obstetrics) initiative on fetal growth: best practice advice for screening, diagnosis, and management of fetal growth restriction. Int J Gynaecol Obstet. 2021;152(Suppl 1):3–57.