Search

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, 2022 Keywords: Computer-aided Diagnosis (CAD), Ultrasound, Abdomen/GI, Liver, Tissue Characterization, Supervised Learning, Transfer Learning, Convolutional Neural Network (CNN)

Contrast enhanced ultrasound (CEUS) takes advantage of the nonlinear behaviour of injected microbubbles. If these contrast techniques yield good specificity between bubbles and tissues, they suffer some drawbacks, inherently linked to their dependence on nonlinear content. In recent years, plane-wave ultrasound reached frame rates of up to 20 000 fps. In this study we propose a linear technique for CEUS that takes advantage of these very high frame rates to separate bubbles from tissue without requiring nonlinearities. Data-driven spatiotemporal filtering operations are used to separate different features in the image on the basis of coherence both in space and time. Such filter recently proved to improve Doppler sensitivity (Demene et al 2015 IEEE Trans. Med. Imaging 34 2271-85). In contrast with bubbles, even slow moving ones, tissues are highly coherent both in space and time. Therefore, singular value decomposition (SVD) seems to be a powerful tool for the separation of contrast agents and tissues. In this paper, we apply SVD processing to linear ultrafast ultrasound images for CEUS Doppler. The contrast levels reached by this technique were compared to those of a nonlinear gold standard sequence (PMPI Doppler) through a flow phantom study. The SVD technique reached contrast-to-tissue ratios (CTR) up to 10 dB higher in vitro, and proved to be robust in terms of probe motion and slow flow. A trial was also conducted on a transplanted human kidney, already imaged by means of power Doppler (Claudon et al 1999 Am. J. Roentgenol. 173 41-6) and microbubbles (Kay et al 2009 Clin. Radiol. 64 1081-7). Contrast levels yielded by the SVD technique measured up to 13 dB higher than those of PMPI Doppler.

Background and objectives The predictors of long–term renal function in living kidney donors are currently discussed. Our objectives were to describe the predictors of functional gain of the remaining kidney after kidney donation. We hypothesized that GFR of the remaining kidney divided by volume of this kidney (rk-GFR/vol) would reflect the density of functional nephrons and be inversely associated with functional gain of the remaining kidney. Design, setting, participants, & measurements We conducted a prospective monocentric study including 63 living donors (26 men; 50.3±11.8 years old) who had been evaluated for 51Cr-EDTA and measured GFR, split renal function by scintigraphy before donation (between 2004 and 2009), and measured GFR at 5.7±0.5 years after donation. For 52 donors, volume of the remaining kidney (measured and estimated with the ellipsoid formula using renal computed tomography scannography) was determined before donation. We tested our hypothesis in an external validation cohort of 39 living donors (13 men; 51.0±9.4 years old) from another single center during the same time period. Results For the main cohort, the mean measured GFR was 97.6±13.0 ml/min per 1.73 m2 before donation and 63.8±9.4 ml/min per 1.73 m2 at 5 years. Functional gain averaged 16.2±7.2 ml/min per 1.73 m2 (+35.3%±16.7%). Multivariate analysis showed that age, body mass index, and rk-GFR/vol at donation were negatively correlated with functional gain and had strong predictive power of the 5-year functional gain (adjusted 5-year functional gain for age: −0.4 [95% confidence interval (95% CI), −0.5 to −0.1]; body mass index: −0.3 [95% CI, −0.6 to −0.1]; rk-GFR/vol: −55.1 [95% CI, −92.3 to −17.9]). We tested this model in the external validation cohort (adjusted 5-year functional gain for age: −0.1 [95% CI, −0.5 to 0.3]; body mass index: −0.9 [95% CI, −1.8 to −0.1]; rk-GFR/vol: −97.6 [95% CI, −137.5 to −57.6]) and confirmed that rk-GFR/vol was inversely associated with 5-year functional gain. Conclusions For given age and body mass index, the long–term functional gain of the remaining kidney is inversely associated with the new variable rk-GFR/vol at donation.

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.

Resume Objectif Etudier la correlation et la concordance entre les volumes renaux calcules en echographie (Vol3DUS) et le volume de reference mesure au scanner (VolTDM) dans la polykystose renale autosomique dominante debutante (PKRAD). Patients et methodes Etude prospective de la correlation et la concordance du volume renal chez 24 patients PKRAD sans insuffisance renale (48 reins analyses separement), avec estimation de Vol3DUS par la formule de l’ellipsoide selon trois modalites differentes et mesure du VolTDM par contourage manuel. Calculs des coefficients de correlation ( r ) et des coefficients de correlation intra-classe (ICC) avec leur intervalle de confiance a 95 %. Resultats Le volume echographique etait fortement correle au volume mesure au scanner en utilisant le plus grand diametre transverse dans le plan axial ( r = 0,83) avec un Vol3DUS moyen = 692 ± 348 mL [180;2069]. La mesure echographique la plus reproductible etait la hauteur. Lorsque le volume renal depassait 800 mL, l’echographie sous-estimait la mesure. Cependant, l’erreur mediane du Vol3DUS etait de −57,5 mL [−1090;183] et 85 % des Vol3DUS calcules s’ecartaient de plus de 5 % de la mesure de reference. Conclusion L’echographie permettait de mesurer le volume renal avec une excellente correlation au volume calcule en scanner. Neanmoins, l’ecart median par rapport au scanner est trop important pour detecter les faibles variations annuelles du volume renal dans la PKRAD debutante et en faire un marqueur du developpement de cette maladie.

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.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).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.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.

Resume Le cancer de la prostate est le premier cancer en incidence et la troisieme deces liee aux cancers chez les hommes, la premiere etant celui du poumon. Le cancer de la prostate est un cancer difficile a detecter et mal traite malgre l’utilisation plus frequente des tests biologiques (antigene prostatique specifique [PSA]), le developpement de nouvelles modalites qui permettent une meilleure visualisation de la prostate et l’utilisation d’outils invasifs comme la biopsie. L’elastographie ultrasonore prostatique est une nouvelle modalite d’imagerie qui permet de cartographier les differentes rigidites des tissus prostatiques. En effet, les tissus cancereux de la prostate sont souvent plus durs que les tissus sains (caracteristique recherchee par le toucher rectal). Cette technique d’imagerie a ete developpee suivant deux principes differents ; le premier est base sur le regime quasi statique et le second est base sur la technique par ondes de cisaillement en regime transitoire. L’elastographie fournit donc une nouvelle information au clinicien, la rigidite des tissus, qui permettra sans doute d’ameliorer la detection et le guidage de la biopsie du cancer prostatique. En effet, l’elastographie, et plus particulierement l’elastographie par onde de cisaillement (EOC), offre une sensibilite elevee pour la detection du cancer prostatique et presente egalement une tres forte valeur predictive negative, assurant que tres peu de cancers sont omis dans la zone peripherique de la prostate. L’elastographie devrait devenir une modalite supplementaire de l’examen de la prostate en complement de l’imagerie ultrasonore endorectale classique et de l’IRM. Cependant, cette technique necessite une formation specifique afin que l’utilisateur se familiarise avec les caracteristiques et les limites de la technique utilisee pour effectuer le diagnostic.

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. 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. 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. Length-split GFR can be used to detect patients requiring scintigraphy. • 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.

Objectifs Evaluer les performances diagnostiques de l’elastographie par ondes de cisaillement ou ShearWave (SWE) et de l’IRM MultiParametrique (IRM-MP) pour le cancer de la prostate. Methodes Soixante-six hommes adresses pour biopsies prostatiques echo-guidees ont beneficie d’une IRM-MP 1,5 T (T2, diffusion, sequence dynamique injectee) et un examen ultrasonore par SWE. Deux radiologues ont analyse en aveugle l’elasticite et l’IRM-MP pour chaque sextant, qui ont ensuite ete comparees aux resultats pathologiques (score de Gleason (SG) et taille). Deux populations de cancer ont ete considerees, PCA1 (taille ≥ 3 mm et SG ≥ 6) et PCA2 (taille ≥ 9 mm et SG ≥ 7). Pour l’IRM-MP, deux scenarios ont ete envisages : IRM-MP1, test considere positif si au moins l’une des sequences IRM est positive ; IRM-MP2, test positif si au moins deux des trois sequences doivent etre positives. Au total, 748 sextants ont ete analyses. Resultats Les resultats sont presentes dans le Tableau 1 . Conclusion L’IRM-MP presente de bonnes performances pour les foyers cancereux centimetriques et agressifs et le SWE offre une meilleure efficacite diagnostique pour tous les cancers prostatiques.

Purpose To investigate the correlation and concordance between the ellipsoid volume calculated by ultrasonography measurements (Vol3DUS) and the reference kidney volume measured by CT (VolTDM) in early autosomal dominant polycystic kidney disease (ADPKD). Materials and methods Prospective study of the correlation and concordance of renal volumes in 24 patients with early ADPKD (48 kidneys analysed separately), with calculation of Vol3DUS using the formula for an ellipsoid in three different manners and VolTDM measurement by manual contouring. Calculations of correlation coefficients (r) and coefficients of intra-class correlation (ICC) with confidence intervals at 95%. Results The US volume was strongly correlated with the CT volume by using the maximum width in a transverse section (r = 0.83) with a mean Vol3DUS = 692 ± 348 ml [180; 2069]. The most reproducible ultrasonography measurement was the height. When the kidney volume exceeded 800 ml, US underestimated the volume. However, the median error was −57.5 ml [−1090; 183] and 85% of the Vol3DUS calculated differed by more than 5% from the reference measurement. Conclusion The correlation between the US calculated volumes and the CT volumes was strong. However, the median error with ellipsoid US volume was too high to detect a small renal variation in early ADPKD.