A framework of deep learning networks provides expert-level accuracy for the detection and prognostication of pulmonary arterial hypertension.

AIMS: To test the hypothesis that deep learning (DL) networks reliably detect pulmonary arterial hypertension (PAH) and provide prognostic information. METHODS AND RESULTS: Consecutive patients with PAH, right ventricular (RV) dilation (without PAH), and normal controls were included. An ensemble of deep convolutional networks incorporating echocardiographic views and estimated RV systolic pressure (RVSP) was trained to detect (invasively confirmed) PAH. In addition, DL-networks were trained to segment cardiac chambers and extracted geometric information throughout the cardiac cycle. The ability of DL parameters to predict all-cause mortality was assessed using Cox-proportional hazard analyses. Overall, 450 PAH patients, 308 patients with RV dilatation (201 with tetralogy of Fallot and 107 with atrial septal defects) and 67 normal controls were included. The DL algorithm achieved an accuracy and sensitivity of detecting PAH on a per patient basis of 97.6 and 100%, respectively. On univariable analysis, automatically determined right atrial area, RV area, RV fractional area change, RV inflow diameter and left ventricular eccentricity index (P < 0.001 for all) were significantly related to mortality. On multivariable analysis DL-based RV fractional area change (P < 0.001) and right atrial area (P = 0.003) emerged as independent predictors of outcome. Statistically, DL parameters were non-inferior to measures obtained manually by expert echocardiographers in predicting prognosis. CONCLUSION: The study highlights the utility of DL algorithms in detecting PAH on routine echocardiograms irrespective of RV dilatation. The algorithms outperform conventional echocardiographic evaluation and provide prognostic information at expert-level. Therefore, DL methods may allow for improved screening and optimized management of PAH.

Utility of deep learning networks for the generation of artificial cardiac magnetic resonance images in congenital heart disease.

BACKGROUND: Deep learning algorithms are increasingly used for automatic medical imaging analysis and cardiac chamber segmentation. Especially in congenital heart disease, obtaining a sufficient number of training images and data anonymity issues remain of concern. METHODS: Progressive generative adversarial networks (PG-GAN) were trained on cardiac magnetic resonance imaging (MRI) frames from a nationwide prospective study to generate synthetic MRI frames. These synthetic frames were subsequently used to train segmentation networks (U-Net) and the quality of the synthetic training images, as well as the performance of the segmentation network was compared to U-Net-based solutions trained entirely on patient data. RESULTS: Cardiac MRI data from 303 patients with Tetralogy of Fallot were used for PG-GAN training. Using this model, we generated 100,000 synthetic images with a resolution of 256 × 256 pixels in 4-chamber and 2-chamber views. All synthetic samples were classified as anatomically plausible by human observers. The segmentation performance of the U-Net trained on data from 42 separate patients was statistically significantly better compared to the PG-GAN based training in an external dataset of 50 patients, however, the actual difference in segmentation quality was negligible (< 1% in absolute terms for all models). CONCLUSION: We demonstrate the utility of PG-GANs for generating large amounts of realistically looking cardiac MRI images even in rare cardiac conditions. The generated images are not subject to data anonymity and privacy concerns and can be shared freely between institutions. Training supervised deep learning segmentation networks on this synthetic data yielded similar results compared to direct training on original patient data.

Prebiotic effect during the first year of life in healthy infants fed formula containing GOS as the only prebiotic: a multicentre, randomised, double-blind and placebo-controlled trial.

PURPOSE: Currently, there is no consensus concerning the possible beneficial colonic and systemic effects of prebiotic-containing infant formula. This study assesses whether the feeding of a galactooligosaccharides (GOS)-containing infant formula (0.44 g/dl of GOS) and the subsequent feeding of a GOS-containing follow-on formula (0.50 g/dl of GOS) have a prebiotic effect on intestinal microbiota that helps to decrease infections and allergy manifestations in healthy infants during the first year of life. METHODS: A multicentre, randomised, double-blind and placebo-controlled trial was carried out on 365 healthy term infants enrolled before 8 weeks of age and randomly assigned to a formula with or without GOS, until 12 months of age. The incidence of infections and allergy manifestations, the antibiotics prescribed and faecal characteristics were recorded up to 12 months of age, while faecal samples were collected up to 4 months for the measurement of secretory immunoglobulin A, short-chain fatty acids and microbiota. RESULTS: A prebiotic effect on the faecal analysis was observed at 4 months of life. The GOS group showed a lower faecal pH (P = 0.019), a lower decreasing trend in secretory immunoglobulin A (P = 0.078), lower butyric acid concentration (P = 0.040) and an increase in Bifidobacterium counts (P = 0.010). Changes in faecal characteristics involved greater frequency (P < 0.001) and softer consistency (P < 0.05). The incidence of infections or allergic manifestations during the first year of life was similar in both groups, with no statistical differences (P > 0.05). CONCLUSIONS: The feeding of GOS-containing infant formula produced a definite prebiotic effect consisting of changes in faecal composition and microbiota, and in faecal consistency and the frequency of defaecation. No changes in the incidence of infection or allergic manifestation during the first year of life were observed.

Does ascorbic acid supplementation affect iron bioavailability in rats fed micronized dispersible ferric pyrophosphate fortified fruit juice?

BACKGROUND AND AIMS: Food iron (Fe) fortification is an adequate approach for preventing Fe-deficiency anemia. Poorly water-soluble Fe compounds have good sensory attributes but low bioavailability. The reduction of the particle size of Fe fortificants and the addition of ascorbic acid might increase the bioavailability of low-soluble compounds. The present work aims to compare the Fe absorption and bioavailability of micronized dispersible ferric pyrophosphate (MDFP) (poorly soluble) to ferrous sufate (FS) (highly soluble) added to a fruit juice in presence or absence of ascorbic acid (AA) by using the hemoglobin repletion assay in rats. METHODS: After a hemoglobin depletion period, four fruit juices comprised of (1) FS, (2) MDFP, (3) FS + AA, (4) MDFP + AA were produced and administered to a different group of rats (n = 18) over 21 days. During the repletion period, Fe balance, hemoglobin regeneration efficiency (HRE), relative bioavailability (RBV) and Fe tissue content were determined in the short, medium and long term. RESULTS: Fe absorption and bioavailability showed no significant differences between fortifying the fruit juice with FS or MDFP. The addition of AA to the juice enhanced Fe absorption during the long-term balance study within the same Fe source. HRE and Fe utilization increased after AA addition in both FS and MDFP groups in every period. CONCLUSION: Fe absorption and bioavailability from MDFP were comparable to FS added to a fruit juice in rats. Further, the addition of AA enhanced Fe absorption in the long term, as well as Fe bioavailability throughout the repletion period regardless of the Fe source employed.

Twenty years of pulmonary tuberculosis in children: what has changed?

OBJECTIVES: To compare the frequency, clinical and radiologic manifestations and source of infection of pulmonary tuberculosis in children treated in our hospital during two decades (1978 through 1987 and 1988 through 1997) and to evaluate the influence of the emergence of HIV infection (since 1985) and the effect of the discontinuation of Calmette-Guérin bacillus (BCG) vaccination (since 1987) on childhood tuberculosis. METHODS: We reviewed 324 children diagnosed with pulmonary tuberculosis in our hospital during the 20 years (1978 through 1997). The data from 2 decades, 1978 through 1987 and 1988 through 1997, were compared. BCG vaccination in Spain was discontinued in 1987, and HIV infection emerged significantly as a public health problem. RESULTS: An increase in the number of children with single hilar adenopathy was observed (32.2% in 1978 through 1987 vs. 43.4%, in 1988 through 1997, P < 0.05) in comparison with those with parenchymal involvement or a mixed pattern (62.4% in 1978 through 1987 vs. 45.7% in 1988 to 1997). Frequency in extrapulmonary manifestations in both periods was similar, with a nonsignificant trend toward a lower rate of tuberculous meningitis in the latter decade (10.4 vs. 5.6%, P = 0.07). We were able to identify an adult source case for 67.1% of the children (100 of 149) in the first decade vs.58.3% (102 of 175) in the second (P = NS); 10.8% of adult contacts but only 2.3% of children (all of them in the second period) were HIV-positive. CONCLUSIONS: Discontinuation of BCG vaccination and emergence of HIV infection have had little influence on childhood tuberculosis in our area.