Search

Artificial Intelligence is providing astonishing results, with medicine being one of its favourite playgrounds. Machine Learning and, in particular, Deep Neural Networks are behind this revolution. Among the most challenging targets of interest in medicine are cancer diagnosis and therapies but, to start this revolution, software tools need to be adapted to cover the new requirements. In this sense, learning tools are becoming a commodity but, to be able to assist doctors on a daily basis, it is essential to fully understand how models can be interpreted. In this survey, we analyse current machine learning models and other in-silico tools as applied to medicine—specifically, to cancer research—and we discuss their interpretability, performance and the input data they are fed with. Artificial neural networks (ANN), logistic regression (LR) and support vector machines (SVM) have been observed to be the preferred models. In addition, convolutional neural networks (CNNs), supported by the rapid development of graphic processing units (GPUs) and high-performance computing (HPC) infrastructures, are gaining importance when image processing is feasible. However, the interpretability of machine learning predictions so that doctors can understand them, trust them and gain useful insights for the clinical practice is still rarely considered, which is a factor that needs to be improved to enhance doctors' predictive capacity and achieve individualised therapies in the near future. [ABSTRACT FROM AUTHOR]

• Proposed a deep learning model to detect cervical cancer. • Proposed a cell merger model to classify in a similar way to real smear examination. • Classification of squamous cells without prior image preprocessing. • Classification of squamous cells in HSIL, LSIL, ASCUS and normal. • Detection of cervical squamous atypiae with 88.8% accuracy in classification. Early detection of cancer is important to improve survival and reduce associated morbility. Nowadays, there is no automatic classification process with enough accuracy to be recommended to its use in population cervical cancer screening. In most automatic medical image classifications, these images are clean in background and without overlap between elements, which means that these images do not reflect reality and the model cannot be applied to directly obtained images from medical samples. The objectives of this study are to design and implement a Cell Merger Approach to improve the efficiency and realism of the PAP-smears classification model, by allowing overlapping and folding of different cells, to design and implement a Convolutional Neural Network for PAP-smears image classification, and to optimize and integrate the cell fusion approach with the neural network building a feasible, reliable and highly accurate system for cervical smears classification. The carried out experiments have validated both the CNN and the proposed Cell Merger Approach with very interesting results. The most outstanding results show that the Convolutional Neural Network models together with the Cell Merger Approach have a classification accuracy of 88.8% with a standard deviation of 1%, obtaining a sensitivity and specificity of 0.92 and 0.83 respectively. This classification level depicts a robust and accurate model that is comparable to an expert pathologist competencies. [ABSTRACT FROM AUTHOR]