Deep Learning for the Detection, Localization, and Characterization of Focal Liver Lesions on Abdominal US Images.

Purpose: To train and assess the performance of a deep learning-based network designed to detect, localize, and characterize focal liver lesions (FLLs) in the liver parenchyma on abdominal US images. Materials and Methods: In this retrospective, multicenter, institutional review board-approved study, two object detectors, Faster region-based convolutional neural network (Faster R-CNN) and Detection Transformer (DETR), were fine-tuned on a dataset of 1026 patients (n = 2551 B-mode abdominal US images obtained between 2014 and 2018). Performance of the networks was analyzed on a test set of 48 additional patients (n = 155 B-mode abdominal US images obtained in 2019) and compared with the performance of three caregivers (one nonexpert and two experts) blinded to the clinical history. The sign test was used to compare accuracy, specificity, sensitivity, and positive predictive value among all raters. Results: DETR achieved a specificity of 90% (95% CI: 75, 100) and a sensitivity of 97% (95% CI: 97, 97) for the detection of FLLs. The performance of DETR met or exceeded that of the three caregivers for this task. DETR correctly localized 80% of the lesions, and it achieved a specificity of 81% (95% CI: 67, 91) and a sensitivity of 82% (95% CI: 62, 100) for FLL characterization (benign vs malignant) among lesions localized by all raters. The performance of DETR met or exceeded that of two experts and Faster R-CNN for these tasks. Conclusion: DETR demonstrated high specificity for detection, localization, and characterization of FLLs on abdominal US images. Supplemental material is available for this article. RSNA, 2022Keywords: Computer-aided Diagnosis (CAD), Ultrasound, Abdomen/GI, Liver, Tissue Characterization, Supervised Learning, Transfer Learning, Convolutional Neural Network (CNN).

Hepatic Production of Fibroblast Growth Factor 23 in Autosomal Dominant Polycystic Kidney Disease.

Context: The bone-derived hormone fibroblast growth factor (FGF) 23 controls phosphate homeostasis and urinary phosphate excretion. FGF23 plasma levels increase in the early stage of renal insufficiency to prevent hyperphosphatemia. Recent evidence suggests that this increase has effects on cardiac and immune cells that compromise patients' health. Patients with autosomal dominant polycystic kidney disease (ADPKD) have been reported to have higher FGF23 concentrations than other patients with similar renal function. The significance of this finding has remained unknown. Methods and Results: Analyzing the FGF23 plasma levels in 434 patients with ADPKD and 355 control subjects with a measured glomerular filtration rate (mGFR) between 60 and 120 mL/min per 1.73 m2, we confirmed that patients with ADPKD had higher FGF23 plasma concentrations than controls. Remarkably, this difference did not translate into renal phosphate leakage. Using different assays for FGF23, we found that this discrepancy was explained by a predominant increase in the cleaved C-terminal fragment of FGF23, which lacks phosphaturic activity. We found that FGF23 plasma concentration independently correlated with the severity of cystic liver disease in ADPKD. We observed that, in contrast to control liver tissues, the cystic liver from patients with ADPKD markedly expressed FGF23 messenger RNA and protein. In line with this finding, the surgical reduction of polycystic liver mass was associated with a decrease in FGF23 plasma levels independently of any modification in mGFR, phosphate, or iron status. Conclusion: Our findings demonstrate that severely polycystic livers produce FGF23 and increase levels of circulating FGF23 in patients with ADPKD.

Use of computed tomography assessed kidney length to predict split renal GFR in living kidney donors.

OBJECTIVES: Screening of living kidney donors may require scintigraphy to split glomerular filtration rate (GFR). To determine the usefulness of computed tomography (CT) to split GFR, we compared scintigraphy-split GFR to CT-split GFR. We evaluated CT-split GFR as a screening test to detect scintigraphy-split GFR lower than 40 mL/min/1.73 m2/kidney. METHODS: This was a monocentric retrospective study on 346 potential living donors who had GFR measurement, renal scintigraphy, and CT. We predicted GFR for each kidney by splitting GFR using the following formula: Volume-split GFR for a given kidney = measured GFR*[volume of this kidney/(volume of this kidney + volume of the opposite kidney)]. The same formula was used for length-split GFR. We compared length- and volume-split GFR to scintigraphy-split GFR at donation and with a 4-year follow-up. RESULTS: A better correlation was observed between length-split GFR and scintigraphy-split GFR (r = 0.92) than between volume-split GFR and scintigraphy-split GFR (r = 0.89). A length-split GFR threshold of 45 mL/min/1.73 m2/kidney had a sensitivity of 100 % and a specificity of 75 % to detect scintigraphy-split GFR less than 40 mL/min/1.73 m2/kidney. Both techniques with their respective thresholds detected living donors with similar eGFR evolution during follow-up. CONCLUSION: Length-split GFR can be used to detect patients requiring scintigraphy. KEY POINTS: • Excellent correlation between kidney length and scintigraphy predicted GFR • Kidney length screening detects all donors with GFR lower than 40 mL/min/1.73 m 2 • Kidney length screening can replace scintigraphy screening.

Prostate cancer: diagnostic performance of real-time shear-wave elastography.

PURPOSE: To prospectively evaluate the performance of real-time ultrasonographic (US) shear-wave elastography (SWE) in the diagnosis of peripheral zone prostate cancer in patients with high and/or increasing prostate-specific antigen levels and/or abnormal digital rectal examination results. MATERIALS AND METHODS: After signing an informed consent form, men referred for transrectal prostate biopsy were enrolled in this prospective HIPAA-compliant two-center study, which was conducted with institutional review board approval. Transrectal US SWE of the prostate was performed after a conventional transrectal US examination and immediately before US-guided 12-core sextant biopsy. For each sextant, the maximum SWE value was measured and matched to the pathologic results of that sextant biopsy. The diagnostic performance of SWE was assessed at both patient and sextant levels. The elasticity value maximizing the Youden index was used to derive sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). RESULTS: The elasticity values were matched to pathologic results for a total of 1040 peripheral zone sextants in 184 men. One hundred twenty-nine positive biopsy findings (size, ≥3 mm; Gleason score, ≥6) were identified in 68 patients. The sextant-level sensitivity, specificity, PPV, NPV, and area under the receiver operating characteristic curve for SWE with a cutoff of 35 kPa for differentiating benign from malignant lesions were 96% (95% confidence interval [CI]: 95%, 97%), 85% (95% CI: 83%, 87%), 48% (95% CI: 46%, 50%), 99% (95% CI: 98%, 100%), and 95% (95% CI: 93%, 97%), respectively. CONCLUSION: Use of a 35-kPa threshold at SWE may provide additional information for the detection and biopsy guidance of prostate cancer, enabling a substantial reduction in the number of biopsies while ensuring that few peripheral zone adenocarcinomas are missed.