9–12 September 2012, Copenhagen, Denmark Poster abstracts P29.13 A simple method for calculation of the fetal femur volume in 3D ultrasound C. Ioannou 1 , I. Sarris 1 , M. Yaqub 2 , R. Napolitano 1 , H. Nicholl 3 , J. A. Noble 2 , M. K. Javaid 4 , A. T. Papageorghiou 1,5 1 Nuffield Department of Obstetrics & Gynaecology, University of Oxford, Oxford, United Kingdom; 2 Institute of Biomedical Engineering, University of Oxford, Oxford, United Kingdom; 3 Department of Radiology, John Radcliffe Hospital, Oxford, United Kingdom; 4 Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Oxford, United Kingdom; 5 Oxford Maternal and Perinatal Health Institute, Green Templeton College, University of Oxford, Oxford, United Kingdom Objectives: Fetal femur volume (FV) measurement using multiplanar tracing is subject to considerable interobserver bias. Our objectives were to describe a simple mathematical model for FV calculation in 3D ultrasound; to establish its repeatability; and to validate it using postmortem computed tomography (CT). Methods: A method for FV was described, which consists of three linear measurements – femur length (FL), proximal metaphyseal diameter (PMD) and mid-shaft diameter (MSD) – and a volume equation. Intraobserver and interobserver repeatability were assessed for this mathematical model and for multiplanar tracing, using within-subject coefficients of variation (CV), intraclass correlation coefficients (ICC) and Bland-Altman plots. Validity was tested by comparing the ultrasound measurement against the true volume obtained from computed tomography (CT), in six fetuses following termination of pregnancy. Time taken for measurement was compared between the two methods. Results: Intraobserver agreement for the mathematical model (CV 10.5%, ICC 0.977) was better than multiplanar tracing (CV 13.2%, ICC 0.952). Interobserver agreement was substantially better for the proposed model (CV 16.8%, ICC 0.923) compared to multiplanar tracing (CV 34.4%, ICC 0.703). FV measured using the proposed ultrasound model was not significantly different when compared to CT (mean difference −0.01 ml, P = 0.739). The median time to measure a femur using the mathematical model was 1.3 min vs. 4.0 min for tracing (P < 0.001). Conclusions: We describe a simple method for FV calculation, which is suitable for clinical use. Compared with the existing tracing technique it is simpler, faster and associated with lower systematic interobserver error. It is accurate when compared to CT. P29.14 The fetal BMI in ultrasound A. Aabadli Kandri Rody 1,2 , L. Legendre 1 , S. Tribalat 1 1 Maternit´ e de Dreux, Dreux, France; 2 Universit´ e Ren´ e Descartes., Paris, France Objectives: To determine whether the fetal BMI is a variable during the pregnancy, and assess its clinical usefulness. Methods: This is a prospective study, including 485 ultrasound examinations from 16 to 39 weeks gestation, excluding all fetal skeletal malformations, and imprecise dating. Hadlock, Olivier & Pineau and Quetelet formulas were used for fetal Weight, stature, and BMI. The FBMI were compared to post natal premature BMI curves. The weight, stature; term and mode of delivery were collected. The women, who underwent elective Caesarean section before labor, or an instrumental vaginal delivery for fetal abnormalities, were excluded. Results: There is a strong correlation between the fetal BMI and the gestational age: (Pearson r 0, 8436, 95% confidence interval 0.8154 to 0.8678 R squared 0, 7117 P value < 0.0001). A first ultrasound fetal BMI percentile curve (10 ◦ , 50 ◦ , and 90 ◦ ) was outlined.133 pregnancy outcomes are known, among them 5 had a BMI above 95 ◦ in the third trimester prenatal ultrasound (Group B) and 128 a BMI under 95 ◦ (Group A). The comparison between the two groups shows for a difficult delivery (instrumental or cesarean section) a likelihood ratio: 2.88; a Relative risk: 2.226 (95% confidence interval: 0.7149 to 6.931), a low sensitivity (8% 95% confidence interval 0.009840 to 0.2603), a high specificity (97.22% 95% confidence interval: 0.9210 to 0.9942) but a P value (P: 0.2363). The smallness of group B probably explains the p value, the low sensitivity and the 95% confidences. Conclusions: This study has proven the direct correlation between the gestational age and the fetal BMI, from 16 weeks to the end of pregnancy; and the existence of a centile distribution. Therefore, we conclude, that some fetuses are too big for their stature, and others too thin for it. Are an excessive FBMI a sign of macrosomia (in some contexts hydrops), and too small an FBMI a sign of IUGR, even if the fetal estimated weights are normal or border line? This question remains open. P29.15 3/4D ultrasound and maternal-fetal bonding: second versus third trimester E. de Jong-Pleij 1 , L. S. Ribbert 1 , L. R. Pistorius 2 , E. Mulder 2 , E. Tromp 3 , C. M. Bilardo 4 1 Department of Obstetrics and Gynecology, St. Antonius Hospital, Utrecht, Netherlands; 2 Fetal Medicine Unit, University Medical Centre Utrecht, Utrecht, Netherlands; 3 Department of Statistics, St. Antonius Hospital, Utrecht, Netherlands; 4 Fetal Medicine Unit, University Medical Centre Groningen, Groningen, Netherlands Objectives: To compare the effect of 2 nd versus 3 rd trimester 3/4 dimensional ultrasound (3/4D US) on maternal-fetal bonding. Methods: 173 healthy Caucasian women completed the Mater- nal Antenatal Attachment Scale 1–2 weeks before (MAAS1) and 1–2 weeks after (MAAS2) a 2 nd or 3 rd trimester 3D/4D US exami- nation. A subset of the total score (Total) explores quality (Q) and another part the time (T) spent in attachment mode, i.e. being positively aware of the fetus. Visibility (V) was scored by the ultra- sonographer and recognition (R) and attractiveness (A) of the 3/4 D images were assessed by the mother according to a graded score from 1–4. A Voluson 730 Expert was used in both groups. Socio- demographic, obstetric and ultrasound characteristics were noted. T- or Chi-square-tests were used for statistics. Results: 107 2 nd trimester (T2) and 66 3 rd trimester (T3) women par- ticipated. Socio-demographic (age, education, smoking, living with partner), obstetric (planned pregnancy, assisted conception, prim- igravidity, first trimester screening) and ultrasound characteristics (BMI, amniotic fluid, placenta location) were not different between the groups (P > 0.05 for all). T2 had lower Total, T and Q MAAS1 scores than T3 (74.0 vs. 77.0 (P = 0.002); 24.9 vs. 27.1 (P < 0.001); 44.9 vs. 45.6 (n.s.). The MAAS 2 scores were not significantly dif- ferent (79.2, 27.4, 47.2 vs. 80.5, 28.6, 47.3, respectively). Increases in Total, T and Q scores were greater in T2 than in T3 (5.2 vs. 3.5 (P = 0.010); 2.5 vs. 1.5 (P = 0.017); 2.3 vs. 1.7 (n.s.)). The scores on V, R and A were similar in both groups (T2: 2.1, 1.7, 1.9, T3: 2.2, 1.7, 1.9). Only in T3 women, the increases in Total and T scores were significantly correlated with V (P = 0.003; P = 0.044) and R scores (P = 0.006, P = 0.001). Conclusions: Increase in maternal-fetal bonding following 3/4D US is greater in the 2 nd trimester. In the 3 rd trimester the effect on maternal-fetal bonding of 3D/4D US is stronger at better degrees of visibility and recognition. Ultrasound in Obstetrics & Gynecology 2012; 40 (Suppl. 1): 171–310 289