Catálogo de publicaciones - libros

In the current paper, the Promedas model for internal medicine, developed by our team, is introduced. The model is based on up-to-date medical knowledge and consists of approximately 2000 diagnoses, 1000 findings and 8600 connections between diagnoses and findings, covering a large part of internal medicine. We show that Belief Propagation (BP) can be successfully applied as approximate inference algorithm in the Promedas network. In some cases, however, we find errors that are too large for this application. We apply a recently developed method that improves the BP results by means of a loop expansion scheme. This method, termed Loop Corrected (LC) BP, is able to improve the marginal probabilities significantly, leaving a remaining error which is acceptable for the purpose of medical diagnosis.

In this paper we describe a module that allows to collect (a) motivations for non-compliance to guidelines, (b) motivations for poor data entry into the electronic patient record, and (c) comments on medical aspects of guideline recommendations, on their formalisation into computerised rules, and on the guideline integration into the computerised clinical chart. We organised a well-structured taxonomy of non-compliance motivations in such a way that the main hierarchical levels correspond to different medical or technical roles suitable for feedback managing. We analysed about 400 consecutive cases of patients with ischemic stroke. About 40 non-compliances, as well as several incomplete data forms have been identified and motivated.

The degree of fulfillment of clinical guidelines is considered a key factor when evaluating the quality of a clinical service. Guidelines can be seen as processes describing the sequence of activities to be done. Consequently, workflow formalisms seem to be a valid approach to model the flow of actions in the guideline and their temporal aspects. The application of a guideline to a specific patient (guideline instance) can be modeled by means of a workflow case. The best (worst) application of a guideline, represented as a reference workflow case, can be used to evaluate the quality of the service, by comparing the optimal case with specific patient instances. On the other hand, the correct application of a guideline to a patient involves the fulfillment of the guideline temporal constraints. Thus, the evaluation of the temporal similarity degree between different workflow cases is a key aspect in evaluating health care quality. In this work, we represent a portion of the stroke guideline using a temporal workflow schema and we propose a method to evaluate the temporal similarity between workflow cases. Our proposal, based on temporal constraint networks, consists of a linear combination of functions to differentiate intra-task and inter-task temporal distances.

Careflow systems implement workflow concepts in the clinical domain in order to administer, support and monitor the execution of health care services performed by different health care professionals and structures. In this work we focus on the monitoring aspects and propose a solution for the conformance verification of careflow process executions.

Given a careflow model, we have defined an algorithm capable of translating it to a formal language based on computational logic and abductive logic programming in particular. The main advantage of this formalism lies in its operational proof-theoretic counterpart, which is able to verify the conformance of a given careflow process execution (in the form of an event log) w.r.t. the model.

The feasibility of the approach has been tested on a case study related to the careflow process described in the cervical cancer screening protocol.

In medicine, prognosis is the task of predicting the probable course and outcome of a disease. Questions like, is a patient going to improve?, what is his/her chance of recovery?, and how likely a relapse is? are common and they rely on the concept of state. The feasible states of a disease define a partial order structure with extreme states those of ’cure’ and ’death’; improving, recovering, and survival meaning particular transitions between states of the partial order. In spite of this, it is not usual in medicine to find an explicit representation either of the states or of the states partial order for many diseases. On the contrary, the variables (e.g. signs and symptoms) related to a disease and their normality and abnormality values are broadly agreed. Here, an inductive algorithm is introduced that generates partial orders from a data matrix containing information about the patient-professional encounters, and the normality functions of each one of these disease variables.

Allergic diseases are increasing all over the world. Therefore, the risk assessment of allergy in newborns is a key issue for prevention purposes. The risk can be assessed at the birth by combining information about familiarity with results of blood examination. Then, the individual must be monitored, particularly in the fist months of life, in order to better define the type of allergy and the risk. The monitoring is carried on by different professionals (agents), therefore the communication and collaboration between these agents must be supported in order to obtain the best treatment strategy for the baby. This paper presents a new project which allows the cooperation between the agents involved in the risk assessment of allergy in newborn babies, and presents the main technologies which will be used to develop it.