Deep learning imaging features derived from kidney ultrasounds predict chronic kidney disease progression in children with posterior urethral valves.

BACKGROUND: We sought to use deep learning to extract anatomic features from postnatal kidney ultrasounds and evaluate their performance in predicting the risk and timing of chronic kidney disease (CKD) progression for boys with posterior urethral valves (PUV). We hypothesized that these features would predict CKD progression better than clinical characteristics such as nadir creatinine alone. METHODS: We performed a retrospective cohort study of boys with PUV treated at two pediatric health systems from 1990 to 2021. Features of kidneys were extracted from initial postnatal kidney ultrasound images using a deep learning model. Three time-to-event prediction models were built using random survival forests. The Imaging Model included deep learning imaging features, the Clinical Model included clinical data, and the Ensemble Model combined imaging features and clinical data. Separate models were built to include time-dependent clinical data that were available at 6 months, 1 year, 3 years, and 5 years. RESULTS: Two-hundred and twenty-five patients were included in the analysis. All models performed well with C-indices of 0.7 or greater. The Clinical Model outperformed the Imaging Model at all time points with nadir creatinine driving the performance of the Clinical Model. Combining the 6-month Imaging Model (C-index 0.7; 95% confidence interval [CI] 0.6, 0.79) with the 6-month Clinical Model (C-index 0.79; 95% CI 0.71, 0.86) resulted in a 6-month Ensemble Model that performed better (C-index 0.82; 95% CI 0.77, 0.88) than either model alone. CONCLUSIONS: Deep learning imaging features extracted from initial postnatal kidney ultrasounds may improve early prediction of CKD progression among children with PUV. A higher resolution version of the Graphical abstract is available as Supplementary information.

Occult Hepatocellular Carcinoma Associated With Transjugular Intrahepatic Portosystemic Shunts in Liver Transplant Recipients.

Transplant eligibility for hepatocellular carcinoma (HCC) is determined by the imaging identification of tumor burden within the Milan criteria. Transjugular intrahepatic portosystemic shunt(s) (TIPS) reduce portal hypertension but may impact HCC visualization. It was hypothesized that the presence of pretransplant TIPS would correlate with occult HCC and reduced survival. A single-center, retrospective, case control study was performed among liver transplant recipients with HCC (2000-2017). The primary endpoint was occult disease on explant pathology. Backward stepwise logistic regression was performed. The secondary endpoints disease-free survival (DFS) and overall survival (OS) were evaluated with Kaplan-Meier curves and Cox regression analysis. Of 640 patients, 40 had TIPS and more frequently exhibited occult disease (80.0% versus 43.1%; P < 0.001; odds ratio [OR], 4.16; P < 0.001). Portal vein thrombosis (PVT) similarly correlated with occult disease (OR, 1.97; P = 0.02). Explant tumor burden was equivalent between TIPS subgroups; accordingly, TIPS status was not independently associated with reduced DFS or OS. However, exceeding the Milan criteria was associated with reduced DFS (hazard ratio, 3.21; P = 0.001), and TIPS status in patients with a single suspected lesion (n = 316) independently correlated with explant tumor burdens beyond these criteria (OR, 13.47; P = 0.001). TIPS on pretransplant imaging are associated with occult HCC on explant pathology. Comparable occult disease findings in patients with PVT suggest that the mechanism may involve altered hepatic perfusion, obscuring imaging diagnosis. TIPS are not independently associated with reduced DFS or OS but are associated with exceeding the Milan criteria for patients with a single suspected lesion. The presence of TIPS may necessitate a higher index of suspicion for occult HCC.