Neural Networks and Survival Analysis Robust Methodology for the Discrimination of Brain Tumors from in vivo Magnetic Resonance Spectra Student name:Ms Yuen Yue Barbara Lee (Ph.D Project) Student name: M s Helen W o n g (Ph.D Project) Supervisor: Professor P.J.G. Lisboa School of Computing and Mathematical Sciences Liverpool John Moores University, Liverpool L3 3AF, UK Abstract Estimate the survival function and predicting the event occurrence has been the center of interest in medical statistic area. Currently, a set of 1,616 breast cancer patients was given by Manchester Christie hospital. Here we investigate different methods, which able to estimate the survival rate of patients after surgery for 5 years accurately, in a way to improve their quality of life and also detect the important factors, which may affect the prognosis accuracy. Cox regression, known as the proportional hazard model, is the most conventional and widely used method for censored survival data. Censorship is a feature of survival data, the endpoint of individuals are not the event of interest. Neural network model has been considered as an alternative model to handle survival data. It has been used for regression and classification problem in several areas, such as engineering, biomedicine, etc. Breast cancer is one of the deadly diseases for women from the last century. In this study, we demonstrated how the neural networks handle censorship, overcome over-fitting problem and avoid significant bias that introduced when using Bayesian framework in stewed data condition. Patients were grouped into different prognostic groups using prognostic indexes and estimate their survival rate for each groups. The result of neural network model was comparable to the Cox regression. Abstract Magnetic resonance spectroscopy has become an important clinical tool, serving as a quantitative indicator of the chemical composition of tissue, whether sampled in vitro or in vivo. In particular, 1H-MRS provides detailed biochemical information of tissues of interest and is increasingly used in tumors diagnosis and grading, where a number of automated decision support have shown promising results. Statistical pattern recognition (PR) techniques have an important role in the allocation of spectra to diagnostic classes because they utilise the multivariate nature of spectra However, the very l a r g e number of spectral components combined with significant noise effects, especially when acquired in vivo, pose substantial difficulties for their statistical analysis, and severely limit to generalise to future data the conclusion from studies using limited number of samples. Linear statistical models commonly be applied after the dimensionality of the spectrum has been reduced to a fraction of the number of samples available, and the reproducibility of the results obtained is strongly dependent on how this parsimony is achieved. Previous studies have passed on clinically identified variables of interest, typically area ratios or attempted to identify the most predictive frequency components. We propose a systematic variable selection methodology based on a well established statistical technique know as the bootstrap to select subset of variable for linear discriminant analysis (LDA) to classify different tissue types to class. This is complemented by the further use of the bootstrap to provide robust estimates of the accuracy of the discriminant models when they are applied to future data. The multivariate structure of MRS also makes it very difficult to visualise the separation between different tissues types. A 2dimensional map of the class membership of the spectra also has considerable potential as a graphical aid for interactive decision support for differential diagnosis by clinicians. Therefore, we propose the application of a planar representation of the spectra, which preserves their relationship to the multivariate class means and variances. In this way, the discriminatory power for different selections of frequency components can be visually demonstrated. Finally, a reject option is introduced into the graphical display with the aim of minimising the risks of misclassification by the explicit visualisation of the overlap between spectra from tumor.

/lp/association-for-computing-machinery/neural-networks-and-survival-analysis-abstract-only-ZLro3WbZY6