A deep neural network approach to predicting clinical outcomes of neuroblastoma patients

Abstract

BackgroundThe availability of high-throughput omics datasets from large patient cohorts has allowed the development of methods that aim at predicting patient clinical outcomes, such as survival and disease recurrence. Such methods are also important to better understand the biological mechanisms underlying disease etiology and development, as well as treatment responses. Recently, different predictive models, relying on distinct algorithms (including Support Vector Machines and Random Forests) have been investigated. In this context, deep learning strategies are of special interest due to their demonstrated superior performance over a wide range of problems and datasets. One of the main challenges of such strategies is the “small n large p” problem. Indeed, omics datasets typically consist of small numbers of samples and large numbers of features relative to typical deep learning datasets. Neural networks usually tackle this problem through feature selection or by including additional constraints during the learning process.MethodsWe propose to tackle this problem with a novel strategy that relies on a graph-based method for feature extraction, coupled with a deep neural network for clinical outcome prediction. The omics data are first represented as graphs whose nodes represent patients, and edges represent correlations between the patients’ omics profiles. Topological features, such as centralities, are then extracted from these graphs for every node. Lastly, these features are used as input to train and test various classifiers.ResultsWe apply this strategy to four neuroblastoma datasets and observe that models based on neural networks are more accurate than state of the art models (DNN: 85%-87%, SVM/RF: 75%-82%). We explore how different parameters and configurations are selected in order to overcome the effects of the small data problem as well as the curse of dimensionality.ConclusionsOur results indicate that the deep neural networks capture complex features in the data that help predicting patient clinical outcomes.

Status and Recommendations of Technological and Data-Driven Innovations in Cancer Care: Focus Group Study

“…Typically, they are based on deep networks that can directly learn the hidden characteristics of the data from different sources. As such, we recommend, for instance, the use of deep neural networks to extract features from genomic and clinical data [ 77 ], convolutional neural networks to extract features from pathology images [ 78 ], and recurrent neural networks for text and medical records data [ 79 ].…”

Machine learning analysis to predict the need for ankle foot orthosis in patients with stroke

“…The deep neural network model may be more appropriate for predicting clinical outcomes 31 . Multiple layers of complex networks may be efficient for representing the complex characteristics of the clinical outcomes in a stroke patient 13 .…”

Neuroblastoma GD2 Expression and Computational Analysis of Aptamer-Based Bioaffinity Targeting

“…Feature extraction is a method used to represent the state of an image as a simple vector where each vector component defines a specific measurable attribute of the image. The main aim of feature extraction for GD2-stained tissues biopsies or cell culture WSIs is to provide a means for quantifying stained cells with the potential to map their quantities to disease state including, but not limited to stage [94][95][96][97], gene expression profile [98], GD2 concentration, etc. Due to the high dimensionality of the image data (typically 100,000 × 100,000 pixels [99,100]), performing feature extraction is a necessary step for many biological image processing techniques as these features could correspond directly to disease state.…”