Predicting the probability of survival in intensive care unit patients from a small number of variables and training examples

Survival in the Intensive Care Unit: A prognosis model based on Bayesian classifiers

We develop a predictive prognosis model to support medical experts in their clinical decision-making process in Intensive Care Units (ICUs) (a) to enhance early mortality prediction, (b) to make more efficient medical decisions about patients at higher risk, and (c) to evaluate the effectiveness of new treatments or detect changes in clinical practice. It is a machine learning hierarchical model based on Bayesian classifiers built from some recorded features of a real-world ICU cohort, to bring about the assessment of the risk of mortality, also predicting destination at ICU discharge if the patient survives, or the cause of death otherwise, constructed as an ensemble of five base Bayesian classifiers by using the average ensemble criterion with weights, and we name it the Ensemble Weighted Average (EWA). We compare EWA against other state-of-the-art machine learning predictive models. Our results show that EWA outperforms its competitors, presenting in addition the advantage over the ensemble using the majority vote criterion of allowing to associate a confidence level to the provided predictions. We also prove the convenience of locally recalibrate from data the standard model used to predict the mortality risk based on the APACHE II score, although as a predictive model it is weaker than the other.

AUTOMATIC DETECTION OF COLORECTAL POLYPS IN STATIC IMAGES

Colorectal cancer continues to be one of the leading causes of mortality worldwide. Scanning using colorectal endoscopy is a useful and common method in clinical examinations. However, the scanning and polyps detections are performed by physicians. Failures to detect polyps might be caused due to lack of experience or knowledge. The purpose of this paper is to discover a scheme able to distinguish polyps from normal tissue in static images off-line.Texture features are studied for the discrimination between polyps and normal tissue. Two useful and simple features are proposed. The student's t-test is applied in selecting useful features to reduce the computation time. The support vector machine is used as a classifier to identify the position of polyps. A study on the numbers in the training patterns is done in order to select an optimal ratio between the polyps and non-polyps sub-images. Seventy-four colonoscopic images are collected to test the system. Half are used as training images and half for testing. The experimental result shows the system can identify all polyps if the colonoscopic images contain a single polyp. The sensitivity is 86.2% and the false-positive rate is 1.26 marks per image.