Automated anatomical demarcation using an active shape model for videofluoroscopic analysis in swallowing.

The current gold standard method in the clinical assessment of swallowing is the visual inspection of videofluoroscopic frames. Specific clinical measurements are estimated based on various anatomical and bolus positional information with respect to time (or frame number). However, due to the subjective nature of visual inspection clinicians face intra- and inter-observer repeatability issues and bias when making these estimations. The correct demarcations of reference lines highlighting the positions of important anatomical landmarks would serve as a visual aid and could also be used in conjunction with bolus detection methods to objectively determine these desirable measurements. In this paper, we introduce and test the reliability of applying a 16-point Active Shape Model as a deformable template to demarcate the boundaries of salient anatomical boundaries with minimal user input. A robust end and corner point detection algorithm is also used to provide image information for the suggested movement of the template during the fitting stage. Results show the model deformation constraints calculated from a training set of images are clinically coherent. The Euclidean distances between the fitted model points against their corresponding target points were measured. Test images were taken from two different data sets from frames acquired using two different videofluoroscopy units. Overall, fitting was found to be more reliable on the vertebrae and inferior points of the larynx compared to the superior laryngeal points and hyoid bone, with the model always fitting the C7 vertebra with discrepancies no higher than a distance of 23 pixels (3.2% of the image width, approximately 7.6mm).

Continuous and discrete time survival analysis: neural network approaches.

In this paper we describe and compare two neural network models aimed at survival analysis modeling, based on formulations in continuous and discrete time. Learning in both models is approached in a Bayesian inference framework. We test the models on a real survival analysis problem, and we show that both models exhibit good discrimination and calibration capabilities. The C index of discrimination varied from 0.8 (SE=0.093) at year 1, to 0.75 (SE=0.034) at year 7 for the continuous time model; from 0.81 (SE=0.07) at year 1, to 0.75 (SE=0.033) at year 7 for the discrete time model. For both models the calibration was good (p<0.05) up to 7 years.

Breast cancer predictions by neural networks analysis: a comparison with logistic regression.

This paper presents an exploratory fixed time study to identify the most significant covariates as a precursor to a longitudinal study of specific mortality, disease free survival and disease recurrences. The data comprise consecutive patients diagnosed with primary breast cancer and entered into the study from 1996 at a single French clinical center, Centre Léon Bérard, based in Lyon, where they received standard treatment. The methodology was to compare and contrast multi-layer perceptron neural networks (NN) with logistic regression (LR), to identify key covariates and their interactions and to compare the selected variables with those routinely used in clinical severity of illness indices for breast cancer. The Logistic regression in this work was chosen as an accepted standard for prediction by biostatisticians in order to evaluate the neural network. Only covariates available at the time of diagnosis and immediately following surgery were used. We used for comparison classification performance indices: AUROC (AREA Under Receiver-Operating Characteristics) curves, sensitivity, specificity, accuracy and positive predictive value for the two following events of interest: Specific Mortality and Disease Free Survival.