Henry Völzke

The Predictive Value of Different Measures of Obesity for Incident Cardiovascular Events and Mortality

CONTEXT To date, it is unclear which measure of obesity is the most appropriate for risk stratification. OBJECTIVE The aim of the study was to compare the associations of various measures of obesity with incident cardiovascular events and mortality. DESIGN AND SETTING We analyzed two German cohort studies, the DETECT study and SHIP, including primary care and general population. PARTICIPANTS A total of 6355 (mean follow-up, 3.3 yr) and 4297 (mean follow-up, 8.5 yr) individuals participated in DETECT and SHIP, respectively. INTERVENTIONS We measured body mass index (BMI), waist circumference (WC), waist-to-height ratio (WHtR), and waist-to-hip ratio (WHR) and assessed cardiovascular and all-cause mortality and the composite endpoint of incident stroke, myocardial infarction, or cardiovascular death. RESULTS In both studies, we found a positive association of the composite endpoint with WHtR but not with BMI. There was no heterogeneity among studies. The relative risks in the highest versus the lowest sex- and age-specific quartile of WHtR, WC, WHR, and BMI after adjustment for multiple confounders were as follows in the pooled data: cardiovascular mortality, 2.75 (95% confidence interval, 1.31-5.77), 1.74 (0.84-3.6), 1.71 (0.91-3.22), and 0.74 (0.35-1.57), respectively; all-cause mortality, 1.86 (1.25-2.76), 1.62 (1.22-2.38), 1.36 (0.93-1.69), and 0.77 (0.53-1.13), respectively; and composite endpoint, 2.16 (1.39-3.35), 1.59 (1.04-2.44), 1.49 (1.07-2.07), and 0.57 (0.37-0.89), respectively. Separate analyses of sex and age groups yielded comparable results. Receiver operating characteristics analysis yielded the highest areas under the curve for WHtR for predicting these endpoints. CONCLUSIONS WHtR represents the best predictor of cardiovascular risk and mortality, followed by WC and WHR. Our results discourage the use of the BMI.

Lung Segmentation of MR Images: A Review

Magnetic resonance imaging (MRI) is a non-radiation based examination method, which gains an increasing popularity in research and clinical settings. Manual analysis of large data volumes is a very time-consuming and tedious process. Therefore, automatic analysis methods are required. This paper reviews different methods that have been recently proposed for automatic and semi-automatic lung segmentation from magnetic resonance imaging data. These techniques include thresholding, region growing, morphological operations, active contours, level sets, and neural networks. We also discuss the methodologies that have been utilized for performance and accuracy evaluation of each method.

Risk Factor Prevalence and Their Relative Influence on Fatty Liver and Gallstone Disease: A Cross-Ethnic Study Comparing Two High-Risk Populations from Chile and Northeast Germany

Objective: Fatty Liver Disease (FLD) and Gallstone disease (GSD) are among the most common gastrointestinal disorders worldwide. Based on data of two ethnically distant populations from Santiago de Chile and Northeast Germany sharing remarkably high prevalence for both diseases, we aimed to estimate (1) the standardised prevalence of FLD and GSD, (2) the population-specific prevalence of known risk factors and (3) the relative influence of each risk factor for the occurrence of the respective outcome. Design: We used data from two population-based studies from Chile (n=742) and Germany (n=3632). Regression methods were applied to assess the prevalence of risk factors at a specific age. To quantify the relative contribution of the risk factors to the occurrence of FLD and GSD, we used hierarchical variation partitioning. Results: The standardised prevalence of FLD ranged from 20.7% in Chilean men to 37.9% in German men and the standardised prevalence of GSD from 14.5% in German men to 40.5% in Chilean women. Risk profiles differed considerably between populations and sexes. Overall, the contributions of the risk factors to the occurrence of FLD and GSD coincided widely. For FLD, age and BMI were the predominant factors, followed by serum triglyceride levels and the Single-Nucleotide Polymorphism (SNP) rs738409 C>G of the PNPLA3 gene. In GSD age was the predominant determinant, in women additionally BMI. Alcohol consumption played a marked role in Chilean men, being positively associated with GSD, similarly the use of contraceptives in Chilean women. The relative contribution of the lithogenic SNP rs11887543 was low in both populations, with marginal effect in Chileans. Conclusion: Despite marked differences in risk profiles, relative influences of risk factors coincided widely between study populations. Knowledge about importance of risk factors for fatty liver and gallstone disease may guide therapy and advises for lifestyle changes.

Low-Circulating Homoarginine is Associated with Dilatation and Decreased Function of the Left Ventricle in the General Population

Low homoarginine is an independent marker of mortality in heart failure patients and incident cardiovascular events. Whether homoarginine is related with healthier cardiac structure and function is currently unclear. We used data of the population-based “Study of Health in Pomerania” (SHIP-Trend) to assess this relation. Homoarginine was measured in serum using liquid chromatography-tandem mass spectrometry. Linear regression models assessed the relation between homoarginine and several structural as well as functional parameters and N-terminal pro B-type natriuretic peptide (NTproBNP). All models were adjusted for age, sex, body mass index, and renal function. A total of 3113 subjects (median age 48 (25th percentile 37 to 75th percentile 60) years, 46% male) were included. A standard deviation decrease in homoarginine was associated with a larger left ventricular diastolic diameter (0.3; 95%-confidence interval (CI): 0.2 to 0.5 mm; p < 0.001), left ventricular systolic diameter (0.38; 95%-CI: −0.22 to 0.54 mm; p < 0.001) as well as a less relative wall thickness (–0.003 95%-CI: −0.006 to −0.0008; p = 0.01), left ventricular ejection fraction (–0.47; 95%-CI: –0.79 to −0.15%; p < 0.01) and fractional shortening (−0.35; 95%-CI: −0.62 to 0.07%; p = 0.01). Low homoarginine was also related to higher NTproBNP (−0.02 95%-CI: −0.034 to −0.009 log pg/mL; p < 0.01). Lower serum homoarginine is associated with dilatation of the heart and decreased function. Prospective clinical studies should assess if homoarginine supplementation improves cardiac health in subjects with low serum concentrations.

Segmentierung von Brustvolumina in Magnetresonanztomographiedaten unter der Verwendung von Deep Learning

Kurzfassung. Die Segmentierung von Hintergrund und Brustgewebe ist ein wichtiger Teil der Auswertung von Magnetresonanztomographie-Daten der Brust. Normalerweise wird diese von Arzten manuell durchgefuhrt. In dieser Arbeit wurde die Segmentierung hingegen mit einer U-net Architektur realisiert. Dabei wurden zwei Netzwerke trainiert und anschliesend auf ein unbekanntes Testset, bestehend aus 8 Probandinnen, angewendet. Die so berechneten Segmentierungen wurden dann mit von Arzten manuell vorgenommenen verglichen. Das erste U-net nutzt keine weitere Vorverarbeitungsmethode und erreicht einen DSC von 0.91±0.09 (Mittelwert ± Standardabweichung). Beim zweiten Netzwerk wurde der N4ITK Bias Correction Algorithmus als Vorverarbeitungsmethode verwendet. Die Masken fur N4ITK konnen sehr grob sein und daher in einer spateren Anwendung von einem Arzt schnell erstellt werden. In dieser Konstellation wurde bei der Segmentierung des Testsets ein DSC von 0.98±0.05 erreicht. Die Segmentierungen benotigen daruber hinaus nach Anfertigung der Masken fur den Vorverarbeitungsalgorithmus 14s. Die Methode hat somit das Potential, Anwendung in der medizinischen Diagnostik zu finden.

Effiziente Segmentierung trachealer Strukturen in MRI-Aufnahmen

Kurzfassung. Die Segmentierung verschiedener Strukturen im Korper ist eine der grundlegenden Operationen in der medizinischen Bildverarbeitung. In dieser Arbeit werden auf Machine Learning basierende Methoden zur Segmentierung medizinischer Bilder untersucht. Das Ziel ist es, in MRI-Scans die Trachea zu segmentieren. Jedoch soll in dieser Arbeit speziell die Effizienz der Algorithmen im Vordergrund stehen. Die verwendeten Ansatze basierten auf einer Deep Learning Architektur, welche zunachst individuell optimiert wird. Es konnte ein maximaler DICE-Koeffizient von (94.4±2.1)% erzielt werden. Zusatzlich kann festgestellt werden, dass die Segmentierung sehr effizient geschieht. Die Segmentierung von einmen Datensatz aus 40 Schichten dauert dabei weniger als eine Sekunde, wobei bei bisherigen Methoden es uber eine Minute benotigte.

Inanspruchnahme von Krebs-früherkennungsuntersuchungen und Einnahme von Sexualhormonen bei Frauen — Ergebnisse der Study of Health in Pomerania (SHIP)

Hintergrund: In Deutschland besteht nach wie vor ein Mangel an evidenzbasierten Informationen auf dem Gebiet der Frauengesundheitsforschung in der Humanmedizin. Die vorliegende Untersuchung analysiert wichtige Charakteristika der Gesundheitsversorgung von Frauen anhand der Daten des Gesundheitssurveys „Study of Health in Pomerania“ (SHIP). Die dargestellten Ergebnisse beziehen sich auf die Inanspruchnahme gynakologischer Behandlungen und Beratungen, die Teilnahme an Krebsfruherkennungsuntersuchungen, die Lebenszeitpravalenz von gynakologischen Operationen einschlieslich Brustchirurgie und der Einnahme von Sexualhormonen.