No abstract

No abstract

were oligohydramnios, bilateral hydronephrosis, bilateral ureteral dilatation, megacystis, bladder thickening and urinoma. The nomogram was configured as an electronic calculator with the baseline 6% PUV incidence among male fetus with moderate to severe hydronephrosis. The pooled diagnostic odds-ratio generated for each diagnostic index was used as coefficient factor for the nomogram configuration. Using our institutional prenatal consultation database (March 2020-present) validation of the configured nomogram in contrast to the keyhole sign was performed.RESULTS: Based on 63 antenatal consults for male infants with moderate to severe-hydronephrosis, the keyhole sign has a specificity of 100% (95% CI 93.8-100%) and sensitivity of 60% (95% CI 14.7-94.7), while the nomogram has specificity of 96.55 (5% CI 88.1-99.6%) and sensitivity of 80 (95% CI 28.4-99.5%). The keyhole sign had 2 false negatives and the nomogram had 2 false positives. The ROC, showed that the nomogram had a superior AUC than the keyhole sign (0.978 vs 0.800, respectively, Figure ). The suggested cut-off using Youdin's index for the nomogram was 95% probability to prevent false positives. The calculated NNS was 1.6 for the Keyhole sign and 1 for the nomogram.CONCLUSIONS: This pilot validation study determined that the Toronto antenatal ultrasound indices nomogram calculator for PUV has similar, if not better, diagnostic accuracy than the keyhole sign. Given the long-term implication of missing a PUV diagnosis, the nomogram can be an adjunctive diagnostic tool to trigger additional post-natal screening for patients do not present with classic keyhole sign but have high index of suspicion.

RESULTS: We were able to predict normal/abnormal renal function in US images with an AUROC 0f 0.776. Including both sagittal and transverse views in the model improved our performance (Table 1). To improve the interpretability of our predictions, we generated heat maps to view areas of interest in US images that our classifier deemed most indicative for predicting function abnormalities.CONCLUSIONS: Prediction of normal or abnormal function based on US images alone appears to be feasible without featureengineering or clinical/patient variables. This technology may allow for closer monitoring of infants and reduce exposure to invasive testing patients receive to assess function.