Deep Learning Based Image Enhancement using Super Resolution

With the ever-growing field of microscopy there is pretty much a necessity of high - resolution microscopic images. A microscope may have powerful magnifying lenses, but if the resolution is poor, the magnified image is just blur and no useful insights can be gained from such images. Traditional techniques like Structured Illumination Microscopy (SIM) are not feasible enough for proper use and current solutions based on deep learning assume that the input image is noise free. Based on our research and existing applications related to deep learning-based image enhancement, our proposed solution of deep learning based General Adversarial Network (GAN), will help jointly denoise and super-resolved microscopy images. Thus, this project has competitive applications in different research areas including biomedical microscopy, medical diagnosis, astronomical research, surveillance or investigation, etc., and many other areas as well.

Can Microwave Ovens Reduce Immunocytochemical Labeling Times?

In recent years the use of microwave ovens in biomedical microscopy laboratories has contributed to reducing the times of fixation and resin embedding. Reports of the use of microwaves for histochemsitry and immunocytochemistry led us to investigate the possible use of a microwave oven to reduce immunocytochemical labeling protocols.The application of specific antibodies to thawed cryosections of aldehyde-fixed material is becoming more accessible to research and service laboratories. These detection methods, routinely performed in our laboratory, were used to study the effect of microwaves on labeling protocols using affinity purified, polyclonal antibodies and protein A-gold.Cells containing 3-(2,4-dinitroanilino)-3-arnino-N-methyldipropylamine (DAMP), a compound which accumulates in low pH compartments, were aldehyde-fixed, cryosectioned and then labeled with rabbit antibodies to dinitrophenol (which bind to DAMP) and 10nm protein-A gold. Regular sequential labeling protocols were compared with protocols using a microwave oven operating at 100% power, where the antibody incubation and washing times were reduced. The effect of microwaves on the labeling efficiency was investigated using simple quantitative methods. The protocol which produced reduced incubation times with no loss of labeling efficiency was then applied to sections in the absence of microwaves. The effect of reducing the final methyl cellulose-uranyl acetate contrasting step was also investigated.