Nanoparticle analysis based on optical ion beam in nuclear imaging by deep learning architectures

Nanotechnology and photonics, two of the most promising fields of the 21st century, come together in nanophotonics. Its primary advantage is that it may be used to build a number of innovative features based on local electromagnetic interaction. Computed tomography (CT) imaging considerably aids in the identification, prognosis, and assessment of therapy response in kidney cancer. This paper presents a unique approach for nuclear image-based kidney tumour identification using deep learning-based segmentation and classification. In addition to CT-features-based gene mutation identification, segmentation-free volume estimate, autonomous kidney localization, and diagnosis of cancer, we developed deep learning approaches. The plan is to boost the contrast of the picture using a convolutional network-based active contour normalisation and then categorise the segmented image with a stacked Monte Carlo Markov encoder neural network. In the experimental study, we look at how different sets of nuclear images perform in terms of training accuracy, Jaccard index, root mean square error (RMSE), NSE, average precision, and F-measure. Training accuracy of 95% was achieved using the suggested method, along with a Jaccard index of 62%, RMSE of 58%, NSE of 61%, average precision of 65%, and F-measure of 71%.