Early identification of dropouts during the special forces selection program.

Recruits are exposed to high levels of psychological and physical stress during the special forces selection period, resulting in dropout rates of up to 80%. To identify who likely drops out, we assessed a group of 249 recruits, every week of the selection program, on their self-efficacy, motivation, experienced psychological and physical stress, and recovery. Using linear regression as well as state-of-the-art machine learning techniques, we aimed to build a model that could meaningfully predict dropout while remaining interpretable.

Deep learning on CT scans to predict checkpoint inhibitor treatment outcomes in advanced melanoma.

Immune checkpoint inhibitor (ICI) treatment has proven successful for advanced melanoma, but is associated with potentially severe toxicity and high costs. Accurate biomarkers for response are lacking. The present work is the first to investigate the value of deep learning on CT imaging of metastatic lesions for predicting ICI treatment outcomes in advanced melanoma.

Flow-driven pattern formation during coacervation of xanthan gum with a cationic surfactant.

We experimentally demonstrate that the coacervation of a biopolymer can trigger a hydrodynamic instability when a coacervate is formed upon injection of a xanthan gum dispersion into a cationic surfactant (CTAB) solution. The local increase of the viscosity due to the coacervate formation induces a viscous fingering instability. Three characteristic displacement regimes were observed: a viscous fingering dominated regime, a buoyancy-controlled "volcano" regime and a "fan"-like regime determined by the coacervate membrane dynamics.

Control of Chemical Waves by Fluid Stretching and Compression.

Oscillatory kinetics coupled to diffusion can produce traveling waves as observed in physical, chemical, and biological systems. We show experimentally that the properties of such waves can be controlled by fluid stretching and compression in a hyperbolic flow. Localized packet waves consisting in a train of parallel waves can develop due to a balance between diffusive broadening and advective compression along the unstable manifold.