Model-based image reconstruction for super-resolution photo-thermal imaging.

Super-Resolution (SR) far-field optical microscopy has revolutionized the quantitative inspection of biological samples by shedding light onto details of the cellular architecture that were previously inaccessible by diffraction-limited imaging. Similarly, SR methods can be exploited to broaden the applicability of imaging modalities operating beyond the visible range. Here we focus on label-free photo-activated thermal imaging and describe the conceptualization and experimental validation of a novel SR imaging approach in the far-infrared. We exploit the photo-thermal effect primed by the absorption of raster-scanned focused visible laser light and the detection of the resulting temperature variations by a microbolometer thermal camera. Based on a full-wave modeling of heat diffusion, we reconstruct the distribution of photo-thermal endogenous entities at tunable sub-diffraction $<5 \mu$m spatial resolution. By significantly enhancing the mm-sized conventional resolution of infrared thermography, our results expand the applicability of quantitative photo-thermal imaging to the micro-scale characterization of the homogeneity, morphology and functional state of biological samples.