RM21Blanket
Posts by Collection
contact
protocols
resources
SOFI-MISRGRU
Introduction
SOFI-MISGRU is a supervised deep learning model designed to accelerate second-order Super-resolution Optical Fluctuation Imaging (SOFI). Our model can reconstruct super-resolved SOFI images using just 20 frames while maintaining a two-fold improvement in spatial resolution. This reduces the usual requirement of hundreds of frames, enabling real-time temporal resolution of up to 4.85 fps for dynamic live-cell imaging. This project has been driven by Miyase Tekpinar and Jelle Komen.
SOFI-simulation
Introduction
We adapted The MATLAB based SOFI simulation tool and modified it to more accurately represent noise, background effects, and biological patterns (mitochondria, microtubules)[1]. To account for electronic noise, a gain and standard deviation map of the sCMOS camera used in our experimental setup was generated and incorporated into the simulation .
[1] Girsault, A. et al. Sofi simulation tool: a software package for simulating and testing super-resolution optical fluctuation imaging. PLoS One 11, e0161602 (2016).
Multiplane calibration
Introduction
Multiplane microscopy images several focal planes of a sample simultaneously by splitting the fluorescence emission onto different regions of one or more camera sensors. The raw frames therefore contain a tiled mosaic of planes that are shifted, slightly rotated and scaled with respect to each other, imaged at different effective transmission, and recorded in an order that is not known a priori. mp_proc is our Python toolbox that calibrates a multiplane setup from a bead z-stack and turns the raw acquisitions into ready-to-use, co-registered image stacks.