Gustavo Santos-García

Prediction of postoperative morbidity after lung resection using an artificial neural network ensemble

OBJECTIVE To propose an ensemble model of artificial neural networks (ANNs) to predict cardio-respiratory morbidity after pulmonary resection for non-small cell lung cancer (NSCLC). METHODS Prospective clinical study was based on 489 NSCLC operated cases. An artificial neural network ensemble was developed using a training set of 348 patients undergoing lung resection between 1994 and 1999. Predictive variables used were: sex of the patient, age, body mass index, ischemic heart disease, cardiac arrhythmia, diabetes mellitus, induction chemotherapy, extent of resection, chest wall resection, perioperative blood transfusion, tumour staging, forced expiratory volume in 1s percent (FEV(1)%), and predicted postoperative FEV(1)% (ppoFEV(1)%). The analysed outcome was the occurrence of postoperative cardio-respiratory complications prospectively recorded and codified. The artificial neural network ensemble consists of 100 backpropagation networks combined via a simple averaging method. The probabilities of complication calculated by ensemble model were obtained to the actual occurrence of complications in 141 cases operated on between January 2000 and December 2001 and a receiver operating characteristic (ROC) curve for this method was constructed. RESULTS The prevalence of cardio-respiratory morbidity was 0.25 in the training and 0.30 in the validation series. The accuracy for morbidity prediction (area under the ROC curve) was 0.98 by the ensemble model. CONCLUSION In this series an artificial neural network ensemble offered a high performance to predict postoperative cardio-respiratory morbidity.

A New Criterion for Soft Set Based Decision Making Problems under Incomplete Information

We put forward a completely redesigned approach to soft set based decision making problems under incomplete information. An algorithmic solution is proposed and compared with previous approaches in the literature. The computational performance of our algorithm is critically analyzed by an experimental study.

Glaucoma Diagnosis: A Soft Set Based Decision Making Procedure

Glaucoma is one of the main causes of blindness in the world. Until it reaches an advanced stage, Glaucoma is asymptomatic, and an early diagnosis improves the quality of life of patients developing this illness. In this paper we put forward an algorithmic solution for the diagnosis of Glaucoma. We approach the problem through a hybrid model of fuzzy and soft set based decision making techniques. Automated combination and analysis of information from structural and functional diagnostic techniques are used in order to obtain an enhanced Glaucoma detection in the clinic.

Incomplete Soft Sets: New Solutions for Decision Making Problems

Alcantud and Santos-Garcia [2] revisit the soft set based decision making problem under incomplete information. Their solution relies on a classical Laplacian argument from probability theory. In view of the computational characteristics of such algorithm, we propose two related solutions that efficiently evaluate problems with many more incomplete data. A computational analysis assesses the performance of our algorithms and compares them with earlier solutions in the literature.

Using Artificial Neural Networks to Identify Glaucoma Stages

Glaucoma is one of the principal causes of blindness in the world1. It is an illness which has an asymptomatic form until advanced stages, thus early diagnosis represents an important objective to achieve with the aim that people who present Glaucoma maintain the best visual acuity throughout life, thereby improving their quality of life. An Artificial Neural Network (ANN) is proposed for the diagnosis of Glaucoma. Automated combination and analysis of information from structural and functional diagnostic techniques were performed to improve Glaucoma detection in the clinic. In our work we contribute the inclusion of Artificial Intelligence and neuronal networks in the diverse systems of clinical exploration and autoperimetry and laser polarimetry, with the objective of facilitating the adequate staging in a rapid and automatic way and thus to be able to act in the most adequate manner possible. Data from clinical examination, standard perimetry and analysis of the nerve fibers of the retina with scanning laser polarimetry (NFAII;GDx) were integrated in a system of Artificial Intelligence. Different tools in the diagnosis of Glaucoma by an automatic classification system were explained based on ANN. In the present work an analysis of 106 eyes, in accordance with the stage of glaucomatous illness was used to develop an ANN. Multilayer perceptron was provided with the Levenberg-Marquardt method. The learning was carried out with half of the data and with the training function of gradient descent w/momentum backpropagation and was checked by the diagnosis of a Glaucoma expert ophthalmologist. A correct classification of each eye in the corresponding stage of Glaucoma has been achieved. Specificity and sensitivity are 100%. This method provides an efficient and accurate tool for the diagnosis of Glaucoma in the stages of glaucomatous illness by means of AI techniques.

Identification of Glaucoma Stages with Artificial Neural Networks Using Retinal Nerve Fibre Layer Analysis and Visual Field Parameters

For the diagnosis of glaucoma, we propose a system of Artificial Intelligence that employs Artificial Neural Networks (ANN) and integrates the analysis of the nerve fibres of the retina from the study with scanning laser polarimetry (NFAII;GDx), perimetry and clinical data. The present work shows an analysis of 106 eyes of 53 patients, in accordance with the stage of glaucomatous illness in which each eye was found. The groups defined include stage 0, which corresponds to normal eyes; stage 1, for ocular hypertension; 2, for early glaucoma; 3, for established glaucoma; 4, for advanced glaucoma and 5, for terminal glaucoma. The developed ANN is a multilayer perceptron provided with the Levenberg-Marquardt method. The learning was carried out with half of the data and with the training function of gradient descent w/momentum backpropagation and was checked by the diagnosis of a glaucoma expert ophthalmologist. The other half of the data served to evaluate the model of the neuronal network. A 100% correct classification of each eye in the corresponding stage of glaucoma has been achieved. Specificity and sensitivity are 100%. This method provides an efficient and accurate tool for the diagnosis of glaucoma in the stages of glaucomatous illness by means of AI techniques.

Analysis of Cellular Proliferation and Survival Signaling by Using Two Ligand/Receptor Systems Modeled by Pathway Logic

Systems biology attempts to understand biological systems by their structure, dynamics, and control methods. Hepatocyte growth factor (HGF) and interleukin 6 (IL6) are two proteins involved in cellular signaling that bind specific cell surface receptors (HGFR and IL6R, respectively) in order to induce cellular proliferation in different cell types or cell contexts. In both cases, the signaling is initiated by binding the ligand (HGF or IL6) to the membrane-bound receptors (HGFR or IL6R) so as to trigger two cellular signaling paths that have several common elements. In this paper we discuss the processes by which an initial cell leads to cellular proliferation and/or survival signaling by using one of these two ligand/receptor systems analyzed by “rewriting logic” methodology. Rewriting logic procedures are suitable computational tools that handle these dynamic systems, and they can be applied to the study of specific biological pathways behavior. Pathway Logic (PL) constitutes a rewriting logic formalism that provides a knowledge base and development environment to carry out model checking, searches, and executions of signaling systems. Moreover, Pathway Logic Assistant (PLA) is a tool that helps us visualize, analyze and understand graphically cellular elements and their relations. We compare the models of HGF/HGFR and IL6/IL6R signaling pathways in order to investigate the relation between these processes and the way in which they induce cellular proliferation. In conclusion, our results illustrate the use of a logical system that explores complex and dynamic cellular signaling processes.

A Logic Computational Framework to Query Dynamics on Complex Biological Pathways

Biological pathways define complex interaction networks where multiple molecular elements work in a series of reactions to produce a response to different biomolecular signals. These biological systems are dynamic and we need mathematical methods that can analyze symbolic elements and complex interactions between them to produce adequate readouts of such systems. Rewriting logic procedures are adequate tools to handle dynamic systems which are applied to the study of specific biological pathways behaviour. Pathway Logic is a rewriting logic development applied to symbolic systems biology. Rewriting logic language Maude allows us to define transition rules and to set up queries about the flow in the biological system. In this paper we describe the use of Pathway Logic to model and analyze the dynamics in a well-known signaling transduction pathway: epidermal growth factor (EGF) pathway. We also use Pathway Logic Assistant (PLA) tool to browse and query this system.

Rewriting Logic Using Strategies for Neural Networks: An Implementation in Maude

A general neural network model for rewriting logic is proposed. This model, in the form of a feedforward multilayer net, is represented in rewriting logic along the lines of several models of parallelism and concurrency that have already been mapped into it. By combining both a right choice for the representation operations and the availability of strategies to guide the application of our rules, a new approach for the classical backpropagation learning algorithm is obtained. An example, the diagnosis of glaucoma by using campimetric fields and nerve fibres of the retina, is presented to illustrate the performance and applicability of the proposed model.

Expanded hesitant fuzzy sets and group decision making

We define expanded hesitant fuzzy sets, which incorporate all available information of the decision makers that provide the membership degrees that define a hesitant fuzzy set. We show how this notion relates to hesitant fuzzy set and extended hesitant fuzzy set. We define various scores for this setting, which generalize popular scores for hesitant fuzzy elements. Finally, a group decision making procedure is presented and illustrated with an example.