Cell quantification in digital contrast microscopy images with convolutional neural networks algorithm

Possible applications of this microscopy include evaluations of cell morphology, cell death, nuclear morphology, internalization of membrane proteins and others2. Working with microscopy images also requires the management and interpretation of terabyte scale data of images generated by analysis algorithms, which requires increasingly robust and sophisticated solutions. Digital phase contrast/brightfield image processingThe Highyield digital contrast experiments are more demanding than fluorescence image processing, presenting several difficulties. Full size imageConvolutional neural networkThe use of a Convolutional Neural Network (CNN) has performance advantages in some ML problems. This paper aims to evaluate the accuracy of predictive models from ML algorithms in the task of identifying the number of cells present in digital contrast microscopy images.