Selective targeting of diseased cells can increase therapeutic efficacy and limit off-target adverse effects. We developed a new tool to selectively perforate living cells with functionalized gold nanoparticles (AuNPs) and near-infrared (NIR) femtosecond (fs) laser. The receptor CD44 strongly expressed by cancer stem cells was used as a model for selective targeting. Citrate-capped AuNPs (100 nm in diameter) functionalized with 0.01 orthopyridyl-disulfide-poly(ethylene glycol) (5 kDa)-N-hydroxysuccinimide (OPSS-PEG-NHS) conjugated to monoclonal antibodies per nm(2) and 5 μM HS-PEG (5 kDa) were colloidally stable in cell culture medium containing serum proteins. These AuNPs attached mostly as single particles 115 times more to targeted CD44(+) MDA-MB-231 and CD44(+) ARPE-19 cells than to non-targeted CD44(-) 661W cells. Optimally functionalized AuNPs enhanced the fs laser (800 nm, 80-100 mJ cm(-2) at 250 Hz or 60-80 mJ cm(-2) at 500 Hz) to selectively perforate targeted cells without affecting surrounding non-targeted cells in co-culture. This novel highly versatile treatment paradigm can be adapted to target and perforate other cell populations by adapting to desired biomarkers. Since living biological tissues absorb energy very weakly in the NIR range, the developed non-invasive tool may provide a safe, cost-effective clinically relevant approach to ablate pathologically deregulated cells and limit complications associated with surgical interventions.

The estimation of musculoskeletal data such as muscle forces and joint torques could have a significant impact on patient care monitoring and medical robotics as well as on reducing healthcare and industrial costs by improving the treatment in the field of rehabilitation. Direct measurement of these data is now non-invasive, as they are computed from dedicated wireless on-body sensors, which can synchronously measure segment positions, muscle activation, external forces and allow to estimate muscle force and joint torques using musculoskeletal models. This paper presents a state-of-the-art survey reviewing both the most commonly used on-body sensors, over the last thirty years, to compute in-body data and the most popular optokinetic cameras. The results are presented and classified into tables which show the evolution of on-body sensors since the 1980's, but also the challenges that lie ahead, as very accurate sensors only accentuate the faults of an inaccurate musculoskeletal model. The survey results show that there is a lack of studies validating the different musculoskeletal models. In addition, current interfaces between hardware and software could be improved.

No abstract