Paulo Félix

Addressing the flaws of current critical alarms: a fuzzy constraint satisfaction approach

OBJECTIVES Threshold alarms, the support supplied by commercial monitoring devices to supervise the signs that pathologies produce over physiological variables, generate a large amount of false positives, owing to the high number of artifacts in monitoring signals, and they are not capable of satisfactorily representing and identifying all monitoring criteria used by healthcare staff. The lack of an adequate support for monitoring the evolution of physical variables prevents the suitable exploitation of the information obtained when monitoring critical patients. This work proposes a solution for designing intelligent alarms capable of addressing the flaws and limitations of threshold alarms. MATERIALS AND METHODS The solution proposed is based on the multivariable fuzzy temporal profile (MFTP) model, a formal model for describing certain monitoring criteria as a set of morphologies defined over the temporal evolution of the patients physiological variables, and a set of relations between them. The MFTP model represents these morphologies through a network of fuzzy constraints between a set of points in the evolution of the variables which the physician considers especially relevant. We also provide a knowledge acquisition tool, TRACE, with which clinical staff can design and edit alarms based on the MFTP model. RESULTS Sixteen alarms were designed using the MFTP model; these were capable of supervising monitoring criteria that could be satisfactorily supervised with commercial monitoring devices. The alarms were validated over a total of 196h of recordings of physiological variables from 78 different patients admitted to an intensive care unit. Of the 912 alarm triggerings, only 7% were false positives. A study of the usability of the tool TRACE was also carried out. After a brief training seminar, five physicians and four nurses designed a number of alarms with this tool. They were then asked to fill in the standard System Usability Scale test. The average score was 68.2. CONCLUSION The proposal presented herein for describing monitoring criteria, comprising the MFTP model and TRACE, permits the supervision of monitoring criteria that cannot be represented by means of thresholds, and makes it possible to construct alarms that give a rate of false positives far below that for threshold alarms.

Detecting Sleep Apnea by Heart Rate Variability Analysis: Assessing the Validity of Databases and Algorithms

Obstructive sleep apnea (OSA) is a serious disorder caused by intermittent airway obstruction which may have dangerous impact on daily living activities. Heart rate variability (HRV) analysis could be used for diagnosing OSA, since this disease affects HRV during sleep. In order to validate different algorithms developed for detecting OSA employing HRV analysis, several public or proprietary data collections have been employed for different research groups. However, for validation purposes, it is obvious and evident the lack of a common standard database, worldwide recognized and accepted by the scientific community. In this paper, different algorithms employing HRV analysis were applied over diverse public and proprietary databases for detecting OSA, and the outcomes were validated in terms of a statistical analysis. Results indicate that the use of a specific database may strongly affect the performance of the algorithms, due to differences in methodologies of processing. Our results suggest that researchers must strongly take into consideration the database used when quoting their results, since selected cases are highly database dependent and would bias conclusions.

Fuzzy constraint networks for signal pattern recognition

This paper deals with representation and reasoning on information concerning the evolution of a physical parameter by means of a model based on the Fuzzy Constraint Satisfaction Problem formalism, and with which it is possible to define what we call Fuzzy Temporal Profiles (FTP). Based on fundamentally linguistic information, this model allows the integration of knowledge on the evolution of a set of parameters into a knowledge representation scheme in which time plays a fundamental role.The FTP model describes the behaviour of a physical parameter on the basis of a set of signal events, and which allows the evolution of the parameter between each pair of events to be modelled as signal episodes. Given the fundamentally linguistic nature of the information represented, the consistency analysis of this information is an essential task. Nevertheless, the obtention of the minimal representation of the network that defines an FTP is an NP-hard problem. In spite of this, we supply algorithms guaranteeing local levels of consistency that allow to correct a large proportion of the errors committed by a human expert in the linguistic description of the profile. Furthermore, we propose a new topology whose consistency can be guaranteed in polynomial time. We also study the applicability of this model in the recognition of signal patterns.

Algorithms for the analysis of polysomnographic recordings with customizable criteria

The diagnosis of Sleep Apnoea-Hypopnoea Syndrome requires the visual inspection of a recording containing a large number of physiological parameters of the patient - the polysomnogram. The purpose of this visual inspection is the identification and characterization of different types of pathological events that occur over these parameters. These events are defined by a set of morphological criteria. Based on these criteria, commercial tools have been developed to support clinicians in the task of visually inspecting polysomnograms. This article argues that the standard morphological criteria are just guiding recommendations that experienced physicians often adapt to each specific diagnostic context. Thus, tools that support the analysis of polysomnograms ideally should use flexible criteria that could be easily customizable by clinicians. In this paper, we propose algorithms capable of identifying pathological events relevant in the diagnosis of SAHS using custom criteria that are acquired directly from the clinician. These algorithms take advantage of the Fuzzy Set Theory to capture and handle the vagueness and uncertainty that are characteristics of medical knowledge. Knowledge acquisition using the traditional linguistic approach of the Fuzzy Sets Theory is supported by a desktop tool. However, the authors feel that some of the criteria that need to be acquired are more visual in nature than linguistic. An alternative mechanism for the visual acquisition of these criteria is proposed. Finally, when presenting the pathological events that have been identified, the tool uses several visual metaphors designed to simplify visual inspection of the polysomnogram. We have validated our proposal over 69h of polysomnographic recordings arising from 12 patients that were subjected to a sleep study. 95.7% of the events identified were correct detections. The rate of false negatives was 1.6%.

Linguistic representation of fuzzy temporal profiles

We take as a setting point the Fuzzy Temporal Profile (FTP) model, which is used to describe the temporal evolution of a certain physical parameter. This model defines the approximation of an evolution curve by means of a set of linear sections between a series of significant points (X0, X1, …, XN), and in which each section is defined by means of an imprecise constraint on the duration, the increase in value and the slope between the points connected by the section. Following the FTP model we present a language for the representation of trends, the semantics of which is projected onto a fuzzy definition of each one of the constraints of a profile of interest.

An Evaluation of Indexes as Support Tools in the Diagnosis of Sleep Apnea

This article evaluates several indexes as support tools to diagnose patients with Sleep Apnea-Hypopnea Syndrome (SAHS). Some of these indexes, such as the Apnea-Hypopnea Index, have been standardized and studied in depth in the literature. Other indexes are used extensively in the reports that commercial polysomnographs generate. However, they have not been studied in detail and clinicians have no standardized guidelines for interpreting them. Examples are the mean and maximum duration of apneas and hypopneas. Finally, several novel indexes proposed by the authors are also evaluated. To evaluate the indexes, we have used a database of 274 patients who have undergone a polysomnographic test. Several feature selection techniques were used to assess the capability of each index to discriminate between healthy and SAHS patients. The capability of the indexes for diagnosing the patients was analyzed by using decision trees which were trained using each index individually, and all the indexes together. Our results suggest that some indexes which are often present in the reports of commercial polysomnographs provide little or no information. On the other hand, other indexes that are usually not considered have a great capability to discern between SAHS and control patients.

A low cost screening test for Obstructive Sleep Apnea that can be performed at the patient's home

This paper presents a new screening test for Obstructive Sleep Apnea-Hypopnea (OSAH) Syndrome based on a PDA platform capable of recording the patients electrocardiogram during his/her nightly rest. OSAH patients present alterations in certain bands of the heart rate variability power spectrum. The more serious the disorder afflicting the patient it is, the greater these alterations are. An algorithm running on the PDA identifies the heart beats, and it calculates the RR intervals time series. Then the algorithm calculates an index that quantifies the magnitude of the heart rate variability power spectrum alterations, and therefore the severity of the patients condition. The reliability of this index has been validated using the Apnea-ECG Database. After the first night of use of the device at the patients home, the PDA communicates the test results to the hospital via the Internet using a WiFi connection, or through GPRS/UMTS. In the hospital, the physicians can decide whether the results are conclusive, or whether it is preferable to repeat the test. Sometimes it may be necessary to repeat the test just because one of the sensors of the electrocardiograph was disconnected during the night. This decision is communicated through the PDA to the patient, who then knows whether he/she should return the device, or if he/she should use it again the following night. The end result is a cheap and reliable OSAH screening test that neither involves a long waiting list nor requires an overnight stay in the hospital for a polysomnographic test.

Fuzzy structural algorithms to identify and characterize apnea and hypopnea episodes

We present a new automated method to identify apneas and hypopneas and to relate them to the drops in blood oxyhemoglobin saturation that they produce. The algorithm takes advantage of the fuzzy set theory for the representation and manipulation of the vagueness of the medical knowledge on which it is based. Its structural nature allows us to easily modify the morphological detection criteria, and to perform a detailed characterization of the identified events. Based on this proposal, a tool for screening polysomnographic records has been developed. The tool allows the physician to modify the morphological criteria that define apneas, hypopneas, and desaturations as well as to inspect the descriptors that make up their characterization. Using this tool we have analyzed five polysomnographic recordings obtaining an overall correct detection rate of 95%. The false negative rate was 6.6% and the false positive rate was 4.3%.

An open platform for the protocolization of home medical supervision

This paper describes SERVANDO, a distributed open platform that deals with a series of recurrent problems in current telemedicine systems, particularly: (1) the scheduling of the different medical actions that should be executed, organized in a personalized agenda generated from a follow-up protocol; (2) functionality encapsulation and reuse in a set of services; (3) communications between the home of the patient and the hospital, through a flexible scheme for bidirectional message exchange; or (4) the management of the events generated during the monitoring. Supervision of patients is carried out through last generation smartphones. SERVANDO provides comprehensive facilities for generic telemedicine applications development, adaptable according to the disease and the particular characteristics of the patient. At the moment, with validation purposes, a follow-up protocol for the supervision of patients with chronic obstructive pulmonary disease has been implemented.

Discovering metric temporal constraint networks on temporal databases

OBJECTIVE In this paper, we propose the ASTPminer algorithm for mining collections of time-stamped sequences to discover frequent temporal patterns, as represented in the simple temporal problem (STP) formalism: a representation of temporal knowledge as a set of event types and a set of metric temporal constraints among them. To focus the mining process, some initial knowledge can be provided by the user, also expressed as an STP, that acts as a seed pattern for the searching procedure. In this manner, the mining algorithm will search for those frequent temporal patterns consistent with the initial knowledge. BACKGROUND Health organisations demand, for multiple areas of activity, new computational tools that will obtain new knowledge from huge collections of data. Temporal data mining has arisen as an active research field that provides new algorithms for discovering new temporal knowledge. An important point in defining different proposals is the expressiveness of the resulting temporal knowledge, which is commonly found in the bibliography in a qualitative form. METHODOLOGY ASTPminer develops an Apriori-like strategy in an iterative algorithm where, as a result of each iteration i, a set of frequent temporal patterns of size i is found that incorporates three distinctive mechanisms: (1) use of a clustering procedure over distributions of temporal distances between events to recognise similar occurrences as temporal patterns; (2) consistency checking of every combination of temporal patterns, which ensures the soundness of the resultant patterns; and (3) use of seed patterns to allow the user to drive the mining process. RESULTS To validate our proposal, several experiments were conducted over a database of time-stamped sequences obtained from polysomnography tests in patients with sleep apnea-hypopnea syndrome. ASTPminer was able to extract well-known temporal patterns corresponding to different manifestations of the syndrome. Furthermore, the use of seed patterns resulted in a reduction in the size of the search space, which reduced the number of possible patterns from 2.1×10⁷ to 1219 and reduced the number of frequent patterns found from 1167 to 340, thereby increasing the efficiency of the mining algorithm. CONCLUSIONS A temporal data mining technique for discovering frequent temporal patterns in collections of time-stamped event sequences is presented. The resulting patterns describe different and distinguishable temporal arrangements among sets of event types in terms of repetitive appearance and similarity of the dispositions between the same events. ASTPminer allows users to participate in the mining process by introducing domain knowledge in the form of a temporal pattern using the STP formalism. This knowledge constrains the search to patterns consistent with the provided pattern and improves the performance of the procedure.