Sample size considerations in soft tissue biomechanics

Biomechanical experiments help link tissue morphology with load-deformation characteristics. A tissue-dependent minimum sample number is indispensable to obtain accurate material properties. Stress-strain properties were retrieved from human dura mater and scalp skin, exemplifying two distinct soft tissues. Minimum sample sizes necessary for a stable estimation of material properties were obtained in a simulation study. One-thousand random samples were sequentially drawn for calculating the point at which a majority of the estimators settled within a corridor of stability at given tolerance levels around a 'complete' reference for the mean, median and coefficient of variation. Stable estimations of means and medians can be achieved below sample sizes of 30 at a ± 20%-tolerance within 80%-conformity for scalp skin and dura. Lower tolerance levels or higher conformity dramatically increase the required sample size. Conformity was barely ever reached for the coefficient of variation. The parameter type appears decisive for achieving conformity. STATEMENT OF SIGNIFICANCE: Biomechanical trials utilizing human tissues are needed to obtain material properties for surgical repair, tissue engineering and modeling purposes. Linking tissue mechanics with morphology helps elucidate form-function relationships, the 'morpho-mechanical link'. For material properties to be accurate, it is vital to examine a minimum number of samples. This number may vary between tissues, and the effects of intrinsic tissue characteristics on data accuracy are unclear to date. This study used data obtained from human dura and skin to compute minimum sample sizes required for estimating material properties at a stable level. It was shown that stable estimations are possible at a ± 20%-tolerance within 80%-conformity below sample sizes of 30. Higher accuracy warrants much higher sample sizes for most material properties.

Separating the impact of gradual lockdown measures on air pollutants from seasonal variability

Analysis of near-surface measurements at several measuring points in Graz, Austria, reveals the impact of restrictive measures during the COVID-19 pandemic on the emission of atmospheric pollutants. We quantify the effects at traffic hotspots, industrial and residential areas. Using historical data collected over several years, we are able to account for meteorological and seasonal confounders. Our analysis is based on daily means as well as intraday pollution level curves. Nitrogen dioxide (NO₂) has decreased drastically while the levels of particulate matter PM₁₀ and carbon monoxide (CO) mostly exhibit little change. Traffic data shows that the decrease in traffic frequency is parallel to the decline in the levels of NO₂ and NO.