Multiscale Convolutional Neural Networks for Vision–Based Classification of Cells

Abstract

Abstract. We present a Multiscale Convolutional Neural Network (MCNN) approach for vision-based classification of cells. Based on several deep Convolutional Neural Networks (CNN) acting at different resolutions, the proposed architecture avoid the classical handcrafted features extraction step, by processing features extraction and classification as a whole. The proposed approach gives better classification rates than classical state-of-the-art methods allowing a safer Computer-Aided Diagnosis of pleural cancer.

Text Classification Model Based on Document Matrix Convolutional Neural Networks

“…DM-CNN is a special and effective multi-layer neural network model, and is outstanding in pattern recognition tasks, and do not require any particular feature extractor selection. The proposed architecture relies on several deep neural networks of alternating convolution layers and subsampling layers [21]. The features detected by the first layer in the DM-CNN can be identified and interpreted easily, and the later layers detected more abstract features.…”

“…The CNN models used to process images usually have multiple layers and sub-sampling layers [21,23,24], and most of the CNN models that deal with the text can use very few layers and sub-sampling layers to achieve good performance [25], as long as there are required word embeddings. This is probably because the pixel of a image is a low-level representation, and the word embedding itself is a more advanced abstract representation, the amount of information carried is much larger than the pixel.…”

Training Multiscale-CNN for Large Microscopy Image Classification in One Hour

“…One novel approach is to use wide networks, which explicitly model various levels of coarseness. In these topologies, several copies of the input image are downsampled and used to train separate, parallel convolutional layers, which are eventually concatenated together to form a single feature vector that is passed on to fully-connected layers (e.g., see Buyssens et al [19]). A recent application of this idea to HCS is the Multiscale Convolutional Neural Network (M-CNN) architecture [16], which has been shown to be generally applicable to multiple microscopy datasets, in particular for identifying the effect of compound treatment.…”

Machine learning approach for homolog chromosome classification

“…In the field of optics and vessel segmentations, cervical Cervigram images . More over for cancer analysis, pleural cancer and epithelial‐2 cell images were discussed . Various deep learning application in the field of medical image analysis is discussed .…”

“…19 More over for cancer analysis, 20 pleural cancer and epithelial-2 cell images were discussed. 21 Various deep learning application in the field of medical image analysis is discussed. [22][23][24][25][26] The existing works in the field of chromosome image analysis is discussed in a review.…”