Enhanced Critical Congenital Cardiac Disease Screening by Combining Interpretable Machine Learning Algorithms.

Critical Congenital Heart Disease (CCHD) screening that only uses oxygen saturation (SpO2), measured by pulse oximetry, fails to detect an estimated 900 US newborns annually. The addition of other pulse oximetry features such as perfusion index (PIx), heart rate, pulse delay and photoplethysmography characteristics may improve detection of CCHD, especially those with systemic blood flow obstruction such as Coarctation of the Aorta (CoA). To comprehensively study the most relevant features associated with CCHD, we investigated interpretable machine learning (ML) algorithms by using Recursive Feature Elimination (RFE) to identify an optimal subset of features. We then incorporated the trained ML models into the current SpO2-alone screening algorithm. Our proposed enhanced CCHD screening system, which adds the ML model, improved sensitivity by approximately 10 percentage points compared to the current standard SpO2-alone method with minimal to no impact on specificity.Clinical relevance- This establishes proof of concept for a ML algorithm that combines pulse oximetry features to improve detection of CCHD with little impact on false positive rate.

Tc-99m mebrofenin hepatobiliary scan in obstructive hepatobiliary disease: determining causes with early and late delayed imaging.

Hepatobiliary radionuclide imaging is typically performed to detect cholecystitis. Infrequently, imaging reveals an obstructive pattern. Although delayed hepatobiliary imaging is commonly used to differentiate between intrahepatic (IH) and extrahepatic (EH) obstruction in the newborn; there is room to clarify the use of delayed imaging in the adult population. A retrospective review was performed of adult patients demonstrating a complete obstructive pattern on initial Tc-99m mebrofenin hepatobiliary imaging. Delayed imaging was divided into early delayed (ED) (<10 h) and late delayed (LD) (≥10 h) imaging. Two physicians qualified the presence of intestinal radiotracer (negative, low to high) on delayed images. Determination of EH or IH pathology was obtained from chart review. A total of 24 patients demonstrated an obstructive pattern using delayed Tc-99m mebrofenin hepatobiliary imaging, with delayed imaging ranging from 4 to 30 h. EH pathologies (choledocholithiasis, stricture, other) represented 63% of cases (n = 15), IH pathologies (cirrhosis, hepatitis, other) represented 33% cases (n = 8) and 1 case was indeterminate. 67% of EH cases showed intestinal activity on delayed imaging (67% on ED and 67% on LD imaging), whereas 63% of IH cases showed intestinal activity on delayed imaging (67% on ED imaging and 60% on LD imaging). The presence of intestinal activity on the both the early and delayed images did not differentiate between the IH and EH pathology groups. Subdividing the groups into ED imaging and LD imaging was also not predictive of determining location of obstructive pattern on the initial 1 h of imaging. This data suggests that delayed hepatobiliary scintigraphy has little or no role in determining the cause of obstructive pathology.