Web-based newborn screening system for metabolic diseases: machine learning versus clinicians.

BACKGROUND: A hospital information system (HIS) that integrates screening data and interpretation of the data is routinely requested by hospitals and parents. However, the accuracy of disease classification may be low because of the disease characteristics and the analytes used for classification. OBJECTIVE: The objective of this study is to describe a system that enhanced the neonatal screening system of the Newborn Screening Center at the National Taiwan University Hospital. The system was designed and deployed according to a service-oriented architecture (SOA) framework under the Web services .NET environment. The system consists of sample collection, testing, diagnosis, evaluation, treatment, and follow-up services among collaborating hospitals. To improve the accuracy of newborn screening, machine learning and optimal feature selection mechanisms were investigated for screening newborns for inborn errors of metabolism. METHODS: The framework of the Newborn Screening Hospital Information System (NSHIS) used the embedded Health Level Seven (HL7) standards for data exchanges among heterogeneous platforms integrated by Web services in the C# language. In this study, machine learning classification was used to predict phenylketonuria (PKU), hypermethioninemia, and 3-methylcrotonyl-CoA-carboxylase (3-MCC) deficiency. The classification methods used 347,312 newborn dried blood samples collected at the Center between 2006 and 2011. Of these, 220 newborns had values over the diagnostic cutoffs (positive cases) and 1557 had values that were over the screening cutoffs but did not meet the diagnostic cutoffs (suspected cases). The original 35 analytes and the manifested features were ranked based on F score, then combinations of the top 20 ranked features were selected as input features to support vector machine (SVM) classifiers to obtain optimal feature sets. These feature sets were tested using 5-fold cross-validation and optimal models were generated. The datasets collected in year 2011 were used as predicting cases. RESULTS: The feature selection strategies were implemented and the optimal markers for PKU, hypermethioninemia, and 3-MCC deficiency were obtained. The results of the machine learning approach were compared with the cutoff scheme. The number of the false positive cases were reduced from 21 to 2 for PKU, from 30 to 10 for hypermethioninemia, and 209 to 46 for 3-MCC deficiency. CONCLUSIONS: This SOA Web service-based newborn screening system can accelerate screening procedures effectively and efficiently. An SVM learning methodology for PKU, hypermethioninemia, and 3-MCC deficiency metabolic diseases classification, including optimal feature selection strategies, is presented. By adopting the results of this study, the number of suspected cases could be reduced dramatically.

High-Performance Seizure Detection System Using a Wavelet-Approximate Entropy-fSVM Cascade With Clinical Validation.

The classification of electroencephalography (EEG) signals is one of the most important methods for seizure detection. However, verification of an atypical epileptic seizure often can only be done through long-term EEG monitoring for 24 hours or longer. Hence, automatic EEG signal analysis for clinical screening is necessary for the diagnosis of epilepsy. We propose an EEG analysis system of seizure detection, based on a cascade of wavelet-approximate entropy for feature selection, Fisher scores for adaptive feature selection, and support vector machine for feature classification. Performance of the system was tested on open source data, and the overall accuracy reached 99.97%. We further tested the performance of the system on clinical EEG obtained from a clinical EEG laboratory and bedside EEG recordings. The results showed an overall accuracy of 98.73% for routine EEG, and 94.32% for bedside EEG, which verified the high performance and usefulness of such a cascade system for seizure detection. Also, the prediction model, trained by routine EEG, can be successfully generalized to bedside EEG of independent patients.

Semantic similarity measures in the biomedical domain by leveraging a web search engine.

Various researches in web related semantic similarity measures have been deployed. However, measuring semantic similarity between two terms remains a challenging task. The traditional ontology-based methodologies have a limitation that both concepts must be resided in the same ontology tree(s). Unfortunately, in practice, the assumption is not always applicable. On the other hand, if the corpus is sufficiently adequate, the corpus-based methodologies can overcome the limitation. Now, the web is a continuous and enormous growth corpus. Therefore, a method of estimating semantic similarity is proposed via exploiting the page counts of two biomedical concepts returned by Google AJAX web search engine. The features are extracted as the co-occurrence patterns of two given terms P and Q, by querying P, Q, as well as P AND Q, and the web search hit counts of the defined lexico-syntactic patterns. These similarity scores of different patterns are evaluated, by adapting support vector machines for classification, to leverage the robustness of semantic similarity measures. Experimental results validating against two datasets: dataset 1 provided by A. Hliaoutakis; dataset 2 provided by T. Pedersen, are presented and discussed. In dataset 1, the proposed approach achieves the best correlation coefficient (0.802) under SNOMED-CT. In dataset 2, the proposed method obtains the best correlation coefficient (SNOMED-CT: 0.705; MeSH: 0.723) with physician scores comparing with measures of other methods. However, the correlation coefficients (SNOMED-CT: 0.496; MeSH: 0.539) with coder scores received opposite outcomes. In conclusion, the semantic similarity findings of the proposed method are close to those of physicians' ratings. Furthermore, the study provides a cornerstone investigation for extracting fully relevant information from digitizing, free-text medical records in the National Taiwan University Hospital database.

E-health and healthcare enterprise information system leveraging service-oriented architecture.

OBJECTIVE: To present the successful experiences of an integrated, collaborative, distributed, large-scale enterprise healthcare information system over a wired and wireless infrastructure in National Taiwan University Hospital (NTUH). In order to smoothly and sequentially transfer from the complex relations among the old (legacy) systems to the new-generation enterprise healthcare information system, we adopted the multitier framework based on service-oriented architecture to integrate the heterogeneous systems as well as to interoperate among many other components and multiple databases. We also present mechanisms of a logical layer reusability approach and data (message) exchange flow via Health Level 7 (HL7) middleware, DICOM standard, and the Integrating the Healthcare Enterprise workflow. The architecture and protocols of the NTUH enterprise healthcare information system, especially in the Inpatient Information System (IIS), are discussed in detail. IMPLEMENTATION: The NTUH Inpatient Healthcare Information System is designed and deployed on service-oriented architecture middleware frameworks. The mechanisms of integration as well as interoperability among the components and the multiple databases apply the HL7 standards for data exchanges, which are embedded in XML formats, and Microsoft .NET Web services to integrate heterogeneous platforms. MEASUREMENTS: The preliminary performance of the current operation IIS is evaluated and analyzed to verify the efficiency and effectiveness of the designed architecture; it shows reliability and robustness in the highly demanding traffic environment of NTUH. CONCLUSIONS: The newly developed NTUH IIS provides an open and flexible environment not only to share medical information easily among other branch hospitals, but also to reduce the cost of maintenance. The HL7 message standard is widely adopted to cover all data exchanges in the system. All services are independent modules that enable the system to be deployed and configured to the highest degree of flexibility. Furthermore, we can conclude that the multitier Inpatient Healthcare Information System has been designed successfully and in a collaborative manner, based on the index of performance evaluations, central processing unit, and memory utilizations.

Design ensemble machine learning model for breast cancer diagnosis.

In this paper, we classify the breast cancer of medical diagnostic data. Information gain has been adapted for feature selections. Neural fuzzy (NF), k-nearest neighbor (KNN), quadratic classifier (QC), each single model scheme as well as their associated, ensemble ones have been developed for classifications. In addition, a combined ensemble model with these three schemes has been constructed for further validations. The experimental results indicate that the ensemble learning performs better than individual single ones. Moreover, the combined ensemble model illustrates the highest accuracy of classifications for the breast cancer among all models.

Web-based healthcare hand drawing management system.

The paper addresses Medical Hand Drawing Management System architecture and implementation. In the system, we developed four modules: hand drawing management module; patient medical records query module; hand drawing editing and upload module; hand drawing query module. The system adapts windows-based applications and encompasses web pages by ASP.NET hosting mechanism under web services platforms. The hand drawings implemented as files are stored in a FTP server. The file names with associated data, e.g. patient identification, drawing physician, access rights, etc. are reposited in a database. The modules can be conveniently embedded, integrated into any system. Therefore, the system possesses the hand drawing features to support daily medical operations, effectively improve healthcare qualities as well. Moreover, the system includes the printing capability to achieve a complete, computerized medical document process. In summary, the system allows web-based applications to facilitate the graphic processes for healthcare operations.

Semantic similarity measure in biomedical domain leverage web search engine.

Semantic similarity measure plays an essential role in Information Retrieval and Natural Language Processing. In this paper we propose a page-count-based semantic similarity measure and apply it in biomedical domains. Previous researches in semantic web related applications have deployed various semantic similarity measures. Despite the usefulness of the measurements in those applications, measuring semantic similarity between two terms remains a challenge task. The proposed method exploits page counts returned by the Web Search Engine. We define various similarity scores for two given terms P and Q, using the page counts for querying P, Q and P AND Q. Moreover, we propose a novel approach to compute semantic similarity using lexico-syntactic patterns with page counts. These different similarity scores are integrated adapting support vector machines, to leverage the robustness of semantic similarity measures. Experimental results on two datasets achieve correlation coefficients of 0.798 on the dataset provided by A. Hliaoutakis, 0.705 on the dataset provide by T. Pedersen with physician scores and 0.496 on the dataset provided by T. Pedersen et al. with expert scores.

Design and implementation of web-based mobile electronic medication administration record.

Patients' safety is the most essential, critical issue, however, errors can hardly prevent, especially for human faults. In order to reduce the errors caused by human, we construct Electronic Health Records (EHR) in the Health Information System (HIS) to facilitate patients' safety and to improve the quality of medical care. During the medical care processing, all the tasks are based upon physicians' orders. In National Taiwan University Hospital (NTUH), the Electronic Health Record committee proposed a standard of order flows. There are objectives of the standard: first, to enhance medical procedures and enforce hospital policies; secondly, to improve the quality of medical care; third, to collect sufficient, adequate data for EHR in the near future. Among the proposed procedures, NTUH decides to establish a web-based mobile electronic medication administration record (ME-MAR) system. The system, build based on the service-oriented architecture (SOA) as well as embedded the HL7/XML standard, is installed in the Mobile Nursing Carts. It also implement accompany with the advanced techniques like Asynchronous JavaScript and XML (Ajax) or Web services to enhance the system usability. According to researches, it indicates that medication errors are highly proportion to total medical faults. Therefore, we expect the ME-MAR system can reduce medication errors. In addition, we evaluate ME-MAR can assist nurses or healthcare practitioners to administer, manage medication properly. This successful experience of developing the NTUH ME-MAR system can be easily applied to other related system. Meanwhile, the SOA architecture of the system can also be seamless integrated to NTUH or other HIS system.

A newborn screening system based on service-oriented architecture embedded support vector machine.

The clinical symptoms of metabolic disorders are rarely apparent during the neonatal period, and if they are not treated earlier, irreversible damages, such as mental retardation or even death, may occur. Therefore, the practice of newborn screening is essential to prevent permanent disabilities in newborns. In the paper, we design, implement a newborn screening system using Support Vector Machine (SVM) classifications. By evaluating metabolic substances data collected from tandem mass spectrometry (MS/MS), we can interpret and determine whether a newborn has a metabolic disorder. In addition, National Taiwan University Hospital Information System (NTUHIS) has been developed and implemented to integrate heterogeneous platforms, protocols, databases as well as applications. To expedite adapting the diversities, we deploy Service-Oriented Architecture (SOA) concepts to the newborn screening system based on web services. The system can be embedded seamlessly into NTUHIS.

Newborn screening healthcare information system based on service-oriented architecture.

In this paper, we established a newborn screening system under the HL7/Web Services frameworks. We rebuilt the NTUH Newborn Screening Laboratory's original standalone architecture, having various heterogeneous systems operating individually, and restructured it into a Service-Oriented Architecture (SOA), distributed platform for further integrity and enhancements of sample collections, testing, diagnoses, evaluations, treatments or follow-up services, screening database management, as well as collaboration, communication among hospitals; decision supports and improving screening accuracy over the Taiwan neonatal systems are also addressed. In addition, the new system not only integrates the newborn screening procedures among phlebotomy clinics, referral hospitals, as well as the newborn screening center in Taiwan, but also introduces new models of screening procedures for the associated, medical practitioners. Furthermore, it reduces the burden of manual operations, especially the reporting services, those were heavily dependent upon previously. The new system can accelerate the whole procedures effectively and efficiently. It improves the accuracy and the reliability of the screening by ensuring the quality control during the processing as well.

Application of portable CDA for secure clinical-document exchange.

Health Level Seven (HL7) organization published the Clinical Document Architecture (CDA) for exchanging documents among heterogeneous systems and improving medical quality based on the design method in CDA. In practice, although the HL7 organization tried to make medical messages exchangeable, it is still hard to exchange medical messages. There are many issues when two hospitals want to exchange clinical documents, such as patient privacy, network security, budget, and the strategies of the hospital. In this article, we propose a method for the exchange and sharing of clinical documents in an offline model based on the CDA-the Portable CDA. This allows the physician to retrieve the patient's medical record stored in a portal device, but not through the Internet in real time. The security and privacy of CDA data will also be considered.

A multi-voting enhancement for newborn screening healthcare information system.

The clinical symptoms of metabolic disorders during neonatal period are often not apparent. If not treated early, irreversible damages such as mental retardation may occur, even death. Therefore, practicing newborn screening is essential, imperative to prevent neonatal from these damages. In the paper, we establish a newborn screening model that utilizes Support Vector Machines (SVM) techniques and enhancements to evaluate, interpret the Methylmalonic Acidemia (MMA) metabolic disorders. The model encompasses the Feature Selections, Grid Search, Cross Validations as well as multi model Voting Mechanism. In the model, the predicting accuracy, sensitivity and specificity of MMA can be improved dramatically. The model will be able to apply to other metabolic diseases as well.