Using deep-learning in fetal ultrasound analysis for diagnosis of cystic hygroma in the first trimester.

OBJECTIVE: To develop and internally validate a deep-learning algorithm from fetal ultrasound images for the diagnosis of cystic hygromas in the first trimester. METHODS: All first trimester ultrasound scans with a diagnosis of a cystic hygroma between 11 and 14 weeks gestation at our tertiary care centre in Ontario, Canada were studied. Ultrasound scans with normal nuchal translucency were used as controls. The dataset was partitioned with 75% of images used for model training and 25% used for model validation. Images were analyzed using a DenseNet model and the accuracy of the trained model to correctly identify cases of cystic hygroma was assessed by calculating sensitivity, specificity, and the area under the receiver-operating characteristic (ROC) curve. Gradient class activation heat maps (Grad-CAM) were generated to assess model interpretability. RESULTS: The dataset included 289 sagittal fetal ultrasound images;129 cystic hygroma cases and 160 normal NT controls. Overall model accuracy was 93% (95% CI: 88-98%), sensitivity 92% (95% CI: 79-100%), specificity 94% (95% CI: 91-96%), and the area under the ROC curve 0.94 (95% CI: 0.89-1.0). Grad-CAM heat maps demonstrated that the model predictions were driven primarily by the fetal posterior cervical area. CONCLUSIONS: Our findings demonstrate that deep-learning algorithms can achieve high accuracy in diagnostic interpretation of cystic hygroma in the first trimester, validated against expert clinical assessment.

The impact of nurses' and physicians' shift change on obstetrical outcomes.

PURPOSE: To evaluate the effect around nurses' shift change and on-call physicians' shift change on obstetrical outcomes. METHODS: A retrospective study of women who had an attempt of labor in a single-medical center, January 2006-December 2017. Obstetrical outcomes were compared between the time around nurses' shift change (6:00-8:00, 14:00-16:00, and 22:00-00:00) to the rest of the day, and between the time around on-call physicians' shift change (6:00-8:00, 14:00-16:00) to the rest of the day. RESULTS: 32,861 women were included, 7826 deliveries occurred during nurses' shift-change, and 25,035 deliveries occurred during the rest of the day. The groups had similar general and obstetrical characteristics, with no statistical difference in cesarean delivery rate (10% vs. 9.8%, P = 0.45) (Table 1). Nurses' shift change had no measurable effect on obstetrical outcomes, including induction of labor, preterm labor, 5-min-Apgar score and cord pH value, except PPH which was less likely to occur during nurses' shift change period (3.8% vs. 4.4%, P = 0.045) (Table 2). From 32,861 deliveries, 5155 deliveries occurred during on-call physicians' shift-change, and 27,706 deliveries occurred during the rest of the day. Induction\augmentation of labor and epidural analgesia were less likely to happen during on-call physicians' shift change (34.4% vs. 38%, P < 0.0001, 59.6% vs. 61.8%, P = 0.003, respectively) (Table 3). The two groups had similar obstetrical outcomes, without statistical difference in cesarean delivery rate (10% vs. 9.8%, P = 0.63) (Table 4). Table 1 General and obstetric characteristics of women giving birth during the time of nurses shift change versus during the rest of the day Variable Change of nurses shifts (n = 7826) All other hours of the day (n = 25,035) P value Maternal age, y 30.3 ± 5.1 30.2 ± 5.2 0.09 Gestational age at birth (weeks) 39.7 ± 1.09 39.8 ± 1.10 0.55 Nulliparity 2077 (35%) 7067 (37%) 0.01 Induction\augmentation of labor 2905 (37) 9368 (38) 0.62 Epidural analgesia 4746 (61) 15,396 (62) 0.16 Neonatal birth weight, g 3340 ± 422 3330 ± 423 0.06 Data is presented as mean ± S.D or N (%) Table 2 Maternal and neonatal adverse outcomes of women giving birth during the time of nurses shift change versus during the rest of the day Variable (%) Change of nurses shifts (n = 7826) All other hours of the day (n = 25,035) P value Vacuum assisted delivery 615 (7.9) 2002 (8.0) 0.69 Cesarean delivery 788 (10) 2443 (9.8) 0.45 Postpartum hemorrhage 294 (3.8) 1089 (4.4) 0.045 Third- and fourth-degree perineal laceration 106 (1.4) 372 (1.5%) 0.51 5-min Apgar score < 7 39 (0.5) 139 (0.6) 0.65 Umbilical pH < 7.2 170 (23) 580 (23) 0.96 Prolonged second stage 190 (2.5) 559 (2.2) 0.22 Maternal and fetal composite adverse outcome* 1309 (16.7%) 4219 (16.9%) 1.00 Data is presented as N (%) *Maternal and fetal composite adverse outcome was defined as the presence of any of the following: vacuum delivery, CD, prolonged second stage, postpartum hemorrhage, third and fourth degree perineal laceration, 5-min Apgar score < 7 and umbilical cord pH < 7.2 Table 3 General and obstetric characteristics of women giving birth during the time of the on-call physicians shift change versus during the rest of the day Variable Change of physicians shifts (n = 5155) All other hours of the day (n = 27,706) P value Maternal age, years 30.3 ± 5.1 30.2 ± 5.2 0.38 Gestational age at birth (weeks) 39.8 ± 1.09 39.8 ± 1.10 0.95 Nulliparity (%) 1303 (33.4) 7841 (37) < 0.0001 Induction\augmentation of labor (%) 1769 (34.3) 10,504 (38) < 0.0001 Epidural analgesia (%) 3067 (59.6) 17,075 (61.8) 0.003 Neonatal birth weight (gr) 3345 ± 416 3330 ± 424 0.019 Data is presented as mean ± S.D or N (%) Table 4 Maternal and neonatal adverse outcomes of women giving birth during the time of physicians on-call shift change versus during the rest of the day Variable (%) Change of physicians shifts (n = 5155) All other hours of the day (n = 27,706) P value Vacuum assisted delivery 397 (7.7) 2220 (8.0) 0.45 Cesarean delivery 517 (10.0) 2714 (9.8) 0.63 Postpartum hemorrhage 209 (4.1) 1174 (4.3) 0.54 Third- and fourth-degree perineal laceration 67 (1.3) 411 (1.5) 0.31 5-min Apgar score < 7 22 (0.5) 156 (0.6) 0.30 Umbilical pH < 7.2 94 (20.3) 656 (23.3) 0.15 Prolonged second stage 127 (2.5%) 622 (2.3%) 0.36 Maternal and fetal composite adverse outcome* 852 (16.5%) 4676 (16.9%) 1.00 Data is presented as N (%) *Maternal and fetal composite adverse outcome was defined as the presence of any of the following: vacuum delivery, CD, prolonged second stage, postpartum hemorrhage, third and fourth degree perineal laceration, 5-min Apgar score < 7 and umbilical cord pH < 7.2 CONCLUSION: Nurses' shift change and on-call physicians' shift change does not appear to be associated with an increase in adverse maternal or neonatal outcomes.