Purpose The nonalcoholic fatty liver disease is the most common liver abnormality. Up to date, liver biopsy is the reference standard for direct liver steatosis quantification in hepatic tissue samples. In this paper we propose a neural network-based approach for nonalcoholic fatty liver disease assessment in ultrasound. Methods We used the Inception-ResNet-v2 deep convolutional neural network pre-trained on the ImageNet dataset to extract high-level features in liver B-mode ultrasound image sequences. The steatosis level of each liver was graded by wedge biopsy. The proposed approach was compared with the hepatorenal index technique and the gray-level co-occurrence matrix algorithm. After the feature extraction, we applied the support vector machine algorithm to classify images containing fatty liver. Based on liver biopsy, the fatty liver was defined to have more than 5% of hepatocytes with steatosis. Next, we used the features and the Lasso regression method to assess the steatosis level. Results The area under the receiver operating characteristics curve obtained using the proposed approach was equal to 0.977, being higher than the one obtained with the hepatorenal index method, 0.959, and much higher than in the case of the gray-level co-occurrence matrix algorithm, 0.893. For regression the Spearman correlation coefficients between the steatosis level and the proposed approach, the hepatorenal index and the gray-level co-occurrence matrix algorithm were equal to 0.78, 0.80 and 0.39, respectively. Conclusions The proposed approach may help the sonographers automatically diagnose the amount of fat in the liver. The presented approach is efficient and in comparison with other methods does not require the sonographers to select the region of interest.
The present study investigates the influence of selected infant and maternal factors on the energy and macronutrient composition of mature human milk (HM). The study enrolled 77 mothers at 4–8 weeks postpartum. Each mother provided 1 sample of HM. Each extracted HM sample was formed by mixing four subsamples of HM, each of which were obtained in one predefined 6-h periods of the day. Among maternal factors, the analysis included: anthropometric data before and after pregnancy; weight gain in pregnancy; body composition, assessed using the Maltron BioScan 920-II to analyze bioimpedance; and dietary intake, assessed with three-day dietary records. Among the neonatal factors, birth weight and length, number of daily feedings and type of delivery were included. The composition of HM, including energy content, protein, fat and carbohydrate concentrations, was analyzed using the Miris human milk analyzer. Pearson’s and Spearman’s correlation coefficients and multivariable logistic regression models were used to analyze the association between the selected maternal and infant factors and HM milk composition. It was found that total protein content of HM was correlated with pre-pregnancy BMI (Spearman rho = 0.238; p = 0.037), current lean body mass (Spearman rho = −0.293, p = 0.01) and total water content (Spearman rho = −0.315, p = 0.005). Carbohydrates were the only macronutrients whose composition was significantly affected by the infant factors. It was reported that higher carbohydrate content was associated with male sex (OR = 4.52, p = 0.049). Our results show that maternal and infant factors, especially maternal pre-pregnancy and current nutritional status and infant sex, interact and affect HM composition, suggesting that macronutrient and energy content in HM may be determined in pregnancy and may have unique compositional profile for every mother–infant dyad.
1 February 2016 in Journal of the American Society of Echocardiography
Introduction: Right ventricular dysfunction (RVD) is an indicator of poor prognosis in normotensive patients with acute pulmonary embolism (APE). The aim of this study was to compare right ventricular (RV)/left ventricular (LV) ratio measured by echocardiography and multidetector computed...
1 November 2016 in Journal of the American Society of Echocardiography
Background:Patients with intermediate-risk acute pulmonary embolism (APE) are a heterogeneous group with an early mortality rate of 2–15%. The tricuspid annulus plane systolic excursion (TAPSE) and tricuspid regurgitation peak gradient (TRPG) can be used for risk stratification, so we analyzed the prognostic value of a new echo parameter (TRPG/TAPSE) for prediction of APE-related 30-day death or need for rescue thrombolysis in initially normotensive APE patients.Methods and Results:The study group consists of 400 non-high-risk APE patients (191 men, age: 63.1±18.9 years) who had undergone echocardiography within the first 24 h of admission. The TRPG/TAPSE parameter was calculated. The clinical endpoint (CE) was a combination of 30-day APE-related death and/or rescue thrombolysis. The CE occurred in 8 (2%) patients. All patients with TAPSE ≥20 mm (n=193, 48.2%) had a good prognosis. Among 206 patients with TAPSE 4.5 were 0.2 and 0.98, respectively. The CE was significantly more frequent in 19 (9.2%) patients with TRPG/TAPSE >4.5 than in 188 (90.8%) with TRPG/TAPSE ≤4.5 (4 (21.1%) vs. 4 (2.1%), P=0.0005). Among normotensive APE patients with TAPSE 4.5 was associated with 21.1% risk of APE-related death or rescue thrombolysis.Conclusions:TRPG/TAPSE, a novel echocardiographic parameter, may be useful for stepwise echocardiographic risk stratification in normotensive patients with APE, and it identifies patients with a poor prognosis.