Thèses sur le sujet « Medical knowledge engineering »

The increasingly high costs of health care in the U.S. have led the general public to search for different medical approaches. Since the 1990's, the use of Complementary and Alternative Medicine (CAM) has radically increased in the U.S. due to its approach to treat physical, mental, and emotional causes of illness. In 2009, the National Health Statistics reported the impact of CAM in the U.S. health care economy, with population expenditures of $14.8 billion out-of-pocket on natural Medicine and $12.4 billion out-of-pocket on visits to CAM providers as a complement to Western Medicine care. CAM interconnects human functions to reach a balanced state, whereas Western Medicine focuses on specialties and body systems. Both Western Medicine and CAM are unlimited sources of knowledge that follow different approaches but that have the common goal of improving patients' well-being. Identifying relationships between Alternative and Western Medicine can open a completely new approach for health care that can increase understanding of human medical conditions, and facilitate the development of new and more cost-effective treatments. However, the abundance and dissimilarity of CAM and Western Medicine data makes knowledge correlation and management an extremely challenging task. The objective of this research is to design the framework for a knowledge management system to organize, store, and manage the abundant data available for Western Medicine and CAM, and to establish key relationships between the two practices for an effective exploration of ideas and possible solutions for medical diagnosis. Three main challenges in the design of the proposed framework are addressed: data acquisition and modeling; data organization, storage and transfer; and information distribution for further generation and sharing of medical knowledge. A framework to relate the diagnosis process in Western Medicine and Traditional Chinese Medicine, as one of the various forms of CAM, is presented based on process-oriented analysis, hierarchical knowledge representation, relational database, and interactive interface for system utilization. The research is demonstrated using a case study on chronic prostatitis, and can be scalable to other medical conditions. The presented system for knowledge management is not intended to provide a definite solution for medical diagnosis, but to enable the exploration and discovery of knowledge for relational medical diagnosis. The results of this research will positively impact information distribution and knowledge generation via interactive medical knowledge systems, development of new skills for diagnosis and treatment, and a broader understanding of medical diseases and treatments.

The purpose of this thesis is to evaluate the medical data management expert system at the Pediatric Intensive Care Unit of the Montreal Children's Hospital. The objective of this study is to provide a systematic method to evaluate and, progressively improve the knowledge embedded in the medical expert system.
Following a literature survey on evaluation techniques and architecture of existing expert systems, an overview of the Patient Data Management System hardware and software components is presented. The design of the Expert Monitoring System is elaborated. Following its installation in the intensive Care Unit, the performance of the Expert Monitoring System is evaluated, operating on real vital sign data and corrections were formulated. A progressive evaluation technique, new methodology for evaluating an expert system knowledge-base is proposed for subsequent corrections and evaluations of the Expert Monitoring System.

Cancer is an ideal target for personal genomics-based medicine that uses high-throughput genome assays such as DNA sequencing, RNA sequencing, and expression analysis (collectively called omics); however, researchers and physicians are overwhelmed by the quantities of big data from these assays and cannot interpret this information accurately without specialized tools. To address this problem, I have created software methods and tools called OCCAM (OmiC data Cancer Analytic Model) and DIPSC (Differential Pathway Signature Correlation) for automatically extracting knowledge from this data and turning it into an actionable knowledge base called the activitome. An activitome signature measures a mutation's effect on the cellular molecular pathway. As well, activitome signatures can also be computed for clinical phenotypes. By comparing the vectors of activitome signatures of different mutations and clinical outcomes, intrinsic relationships between these events may be uncovered. OCCAM identifies activitome signatures that can be used to guide the development and application of therapies. DIPSC overcomes the confounding problem of correlating multiple activitome signatures from the same set of samples. In addition, to support the collection of this big data, I have developed MedBook, a federated distributed social network designed for a medical research and decision support system. OCCAM and DIPSC are two of the many apps that will operate inside of MedBook. MedBook extends the Galaxy system with a signature database, an end-user oriented application platform, a rich data medical knowledge-publishing model, and the Biomedical Evidence Graph (BMEG). The goal of MedBook is to improve the outcomes by learning from every patient.

Direct Volume Rendering (DVR) is a visualization technique that has proved to be a very powerful tool in many scientific visualization applications. Diagnostic medical imaging is one domain where DVR could provide clear benefits in terms of unprecedented possibilities for analysis of complex cases and highly efficient work flow for certain routine examinations. The full potential of DVR in the clinical environment has not been reached, however, primarily due to limitations in conventional DVR methods and tools. This thesis presents methods addressing four major challenges for DVR in clinical use. The foundation of all methods is to incorporate the domain knowledge of the medical professional in the technical solutions. The first challenge is the very large data sets routinely produced in medical imaging today. To this end a multiresolution DVR pipeline is proposed, which dynamically prioritizes data according to the actual impact in the rendered image to be reviewed. Using this prioritization the system can reduce the data requirements throughout the pipeline and provide high performance and visual quality in any environment. Another problem addressed is how to achieve simple yet powerful interactive tissue classification in DVR. The methods presented define additional attributes that effectively captures readily available medical knowledge. The task of tissue detection is also important to solve in order to improve efficiency and consistency of diagnostic image review. Histogram-based techniques that exploit spatial relations in the data to achieve accurate and robust tissue detection are presented in this thesis. The final challenge is uncertainty visualization, which is very pertinent in clinical work for patient safety reasons. An animation method has been developed that automatically conveys feasible alternative renderings. The basis of this method is a probabilistic interpretation of the visualization parameters. Several clinically relevant evaluations of the developed techniques have been performed demonstrating their usefulness. Although there is a clear focus on DVR and medical imaging, most of the methods provide similar benefits also for other visualization techniques and application domains.

Agile methods for software development have become popular, especially since the agile manifesto was written in 2001. Many positive effects have been found in organizations using agile methods, but also several dangers. Communication and collaboration in teams is often mentioned as something that works well with the methods, but interactions between teams are often lacking. Since agile teams are cross-functional and focused on products rather than specializations, knowledge becomes spread out in the organization. Within teams, different members often have different deep knowledge, but instead have a lot of knowledge about their products in common. This allows them to discuss knowledge related to the product well, but limits their possibility to discuss advanced topics and experiences regarding their deep knowledge or specialization within their team. These issues are important to consider when applying agile methods in organizations, and the research about the issues is quite thin.   In this research, I have taken an interpretive approach and carried out a case study at the development department of a medium-sized IT company providing large software systems for the healthcare industry, for which I have used the pseudonym MedTech. Three different teams have been studied through interviews with all members as well as observations of agile practices like daily meetings. Further, I have studied three different knowledge practices that MedTech uses to complement the creation and sharing of knowledge that happens in teams. These three complementary knowledge practices had different forms and handled knowledge in different ways. One was closely related to what literature often calls communities of practice, which are groups where members share an interest and interact to deepen their knowledge. In this practice at MedTech, meetings were used to discuss experiences and knowledge about topics within specific areas. Another was more focused on one-way communication through presentations and reading tips, spreading more basic knowledge to a wider audience. The third complementary knowledge practice let employees use 12 work hours every sixth week to do whatever they wanted that related to their knowledge, allowing them to e.g. explore new technologies and be creative or simply read up on some interesting topic.   My results show that agile teams support some sharing and creation of knowledge, especially through having members work closely to each other and share experiences, and through practising their skills in daily work, with help from each other when necessary. Like other research has shown, there was however a lack of practices for interactions between teams in the agile methods. Such interactions were crucial since I found teams to be comparable to theories about communities of knowing, where teams create strong perspectives, the sharing of which is important for utilization and creation of knowledge. The complementary knowledge practice that related to communities of practice was shown to be good for connecting employees with similar specializations, who would normally be separated in different teams. This allowed for creation and sharing of knowledge as individuals needed to explain their experiences and could combine knowledge from different members of the community. The other two complementary knowledge practices were shown to be good for increasing motivation to create and share knowledge, and showing that the organization valued the knowledge of individuals.

Thesis (M.S.)--Harvard University--Massachusetts Institute of Technology Division of Health Sciences and Technology, Program in Medical Engineering and Medical Physics, 1989.
Includes bibliographical references (leaves [165]-[172]).
by Henry C. Chueh.
M.S.

CETTE thèse est consacrée à la conception d'un système d'aide au diagnostic dédiéau muscle squelettique humain. Au cours du premier volet de ce manuscrit nousproposons une nouvelle représentation basée sur les modèles parcimonieux dans le cadrede la segmentation d'Images de Résonances Magnétiques (IRM) T1 du muscle squelettiquedu mollet. Notre méthode Sparse Shape Model/ Modèle de Formes Parcimonieux(MFP), apprend un modèle statistique de formes et de textures locales annoté et réussità en tirer une représentation réduite afin de reconstruire le mécanisme musculaire sur unexemple test. Dans la seconde partie du manuscrit, nous présentons une approche baséesur des ondelettes de diffusion pour la segmentation du muscle squelettique. Contrairementaux méthodes de l'état de l'art, notre approche au cours de la phase d'apprentissagepermet à optimiser les coefficients des ondelettes, ainsi que leur nombres et leur positions.Le modèle prend en charge aussi bien les hiérarchies dans l'espace de recherche,que l'encodage des dépendances géométriques complexes et photométriques de la structured'intérêt. Notre modélisation offre ainsi l'avantage de traiter des topologies arbitraires.L'évaluation expérimentale a été effectué sur un ensemble de mollets acquisespar un scanner IRM, ainsi qu'un ensemble d'images tomodensitométriques du ventriculegauche.

The technique of volume rendering can be a powerful tool when visualizing 3D medical data sets. Its characteristic of capturing 3D internal structures within a 2D rendered image makes it attractive in the analysis. However, the applications that implement this technique fail to reach out to most of the supposed end-users at the clinics and radiology departments of today. This is primarily due to problems centered on the design of the Transfer Function (TF), the tool that makes tissues visually appear in the rendered image. The interaction with the TF is too complex for a supposed end-user and its capability of separating tissues is often insufficient. This thesis presents methods for detecting the regions in the image volume where tissues are contained. The tissues that are of interest can furthermore be identified among these regions. This processing and classification is possible thanks to the use of a priori knowledge, i.e. what is known about the data set and its domain in advance. The identified regions can finally be visualized using tissue adapted TFs that can create cleaner renderings of tissues where a normal TF would fail to separate them. In addition an intuitive user control is presented that allows the user to easily interact with the detection and the visualization.

Includes vita.
Title from PDF of title page.
Document formatted into pages; contains 222 pages.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: This dissertation presents a knowledge-guided expert system that is capable of applying routinesfor multispectral analysis, (un)supervised clustering, and basic image processing to automatically detect and segment brain tissue abnormalities, and then label glioblastoma-multiforme brain tumors in magnetic resonance volumes of the human brain. The magnetic resonance images used here consist of three feature images (T1-weighted, proton density, T2-weighted) and the system is designed to be independent of a particular scanning protocol. Separate, but contiguous 2D slices in the transaxial plane form a brain volume. This allows complete tumor volumes to be measured and if repeat scans are taken over time, the system may be used to monitor tumor response to past treatments and aid in the planning of future treatment. Furthermore, once processing begins, the system is completely unsupervised, thus avoiding the problems of human variability found in supervised segmentation efforts.Each slice is initially segmented by an unsupervised fuzzy c-means algorithm. The segmented image, along with its respective cluster centers, is then analyzed by a rule-based expert system which iteratively locates tissues of interest based on the hierarchy of cluster centers in feature space. Model-based recognition techniques analyze tissues of interest by searching for expected characteristics and comparing those found with previously defined qualitative models. Normal/abnormal classification is performed through a default reasoning method: if a significant model deviation is found, the slice is considered abnormal. Otherwise, the slice is considered normal. Tumor segmentation in abnormal slices begins with multispectral histogram analysis and thresholding to separate suspected tumor from the rest of the intra-cranial region. The tumor is then refined with a variant of seed growing, followed by spatial component analysis and a final thresholding step to remove non-tumor pixels.The knowledge used in this system was extracted from general principles of magnetic resonance imaging, the distributions of individual voxels and cluster centers in feature space, and anatomical information. Knowledge is used both for single slice processing and information propagation between slices. A standard rule-based expert system shell (CLIPS) was modified to include the multispectral analysis, clustering, and image processing tools.A total of sixty-three volume data sets from eight patients and seventeen volunteers (four with and thirteen without gadolinium enhancement) were acquired from a single magnetic resonance imaging system with slightly varying scanning protocols were available for processing. All volumes were processed for normal/abnormal classification. Tumor segmentation was performed on the abnormal slices and the results were compared with a radiologist-labeled ground truth' tumor volume and tumor segmentations created by applying supervised k-nearest neighbors, a partially supervised variant of the fuzzy c-means clustering algorithm, and a commercially available seed growing package. The results of the developed automatic system generally correspond well to ground truth, both on a per slice basis and more importantly in tracking total tumor volume during treatment over time.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.

Cardiovascular disease (CVD) is the leading cause of morbidity, mortality, premature death and reduced quality of life for the citizens of the EU. It has been reported that CVD represents a major economic load on health care sys- tems in terms of hospitalizations, rehabilitation services, physician visits and medication. Data Mining techniques with clinical data has become an interesting tool to prevent, diagnose or treat CVD. In this thesis, Knowledge Dis- covery and Data Mining (KDD) was employed to analyse clinical and demographic data, which could be used to diagnose coronary artery disease (CAD). The exploratory data analysis (EDA) showed that female patients at an el- derly age with a higher level of cholesterol, maximum achieved heart rate and ST-depression are more prone to be diagnosed with heart disease. Furthermore, patients with atypical angina are more likely to be at an elderly age with a slightly higher level of cholesterol and maximum achieved heart rate than asymptotic chest pain patients. More- over, patients with exercise induced angina contained lower values of maximum achieved heart rate than those who do not experience it. We could verify that patients who experience exercise induced angina and asymptomatic chest pain are more likely to be diagnosed with heart disease. On the other hand, Logistic Regression, K-Nearest Neighbors, Support Vector Machines, Decision Tree, Bagging and Boosting methods were evaluated by adopting a stratified 10 fold cross-validation approach. The learning models provided an average of 78-83% F-score and a mean AUC of 85-88%. Among all the models, the highest score is given by Radial Basis Function Kernel Support Vector Machines (RBF-SVM), achieving 82.5% ± 4.7% of F-score and an AUC of 87.6% ± 5.8%. Our research con- firmed that data mining techniques can support physicians in their interpretations of heart disease diagnosis in addition to clinical and demographic characteristics of patients.

This thesis looks at two sides of the same coin: how to support the production and future development at a specialist medical technology department at Danderyd Hospital. The two sides are; a pilot study of a product management system (PDM) and an interview based study on the characteristics of the silent knowledge of the technicians. The department (National respiratory centre, NRC) is facing retirement of several key employees.

The technical study shows that the success of an implementation is largely dependent on the users’ prior knowledge and use of a 3D Computer aided design system (CAD).The system itself is shown to fulfill the Lifecycle requirement of tracking the products (mostly tracheostomy tubes) but without a CAD centered workflow, some substantial education and preferably some new recruits, an implementation of the PDM system will fail. The author recommends development of the current “low-tech” system of MS Excel and Access rather than redistribute the dependency from technician towards a complex, commercial software and its vendor.

The analysis of the technicians’ silent knowledge with the newly developed method, epithet for silent knowledge (ETK), shows that the longer employment time:

• the more differentiated technicians become in describing their work,
• practical knowledge are regarded higher and
• the social and collective problem solving factors of the work becomes more important.

Typically, it is shown that a new employee should preferably enjoy problem solving, being pragmatic and social as well as having some prior education or work experience in a CAD and/or a PDM system.

Denna essä syftar till att ur filosofiska och idéhistoriska perspektiv belysa utmaningar och möjligheter med artificiell intelligens (AI) och digitalisering inom hälso- och sjukvården, med fokus på läkekonst, kropp, hälsa och ovisshet. Essän undersöker hur automatisering och digitala vårdformer omformar läkekonstens grund, nämligen mötet mellan patienten och läkaren. Genom en fenomenologisk kritik av AI och teknikens väsen, belyses skillnaden mellan människan och maskinen och hur den levda erfarenheten är situerad, förkroppsligad, fylld av mening och delad med andra. Essän utforskar hur situationsunik kunskap som praktisk klokhet, fronesis, samt ett reflekterande förnuft, intellectus,kan hantera den ovisshet som är inbäddad i det allmänmedicinska mötet. Essän belyser även hur digitalisering och AI passar väl med pågående marknadsanpassning av sjukvården, där homo economicus och homo digitalis båda omformar kropp och hälsa till mätbara resurser och data. Avslutningsvis lyfts etiska dilemman kring AI och digitalisering, samt vikten av praktisk och existentiell kunskap som förutsättningar för utvecklandet och designen av en teknik som syftar främja det mänskligt goda.
This essay aims to illuminate challenges and opportunities with artificial intelligence (AI) and digitalization in health care, focusing on the art of medicine, body, health and uncertainty. The theoretical framework is mainly within the fields of phenomenology and philosophical hermeneutics. The essay explores how automatization and digital health care are transforming the essence of medicine: the patient – physician encounter. By a phenomenological critique of AI and the essence of technology, the essay highlights the difference between machines and humans and how lived experience is situated, embodied, filled with meaning and shared with others. The essay explores how situational knowledge such as practical wisdom, phronesis, and reflective understanding, intellectus, can deal with the uncertainty that is embedded in the medical encounter in primary health care. The essay also highlights how digitalization and AI fit well with current market adaptation of health care, where homo economicus and homo digitalis both transform body and health into measurable resources and data. Finally, ethical dilemmas of AI and digitalization are highlighted, as well as the importance of practical and existential knowledge as preconditions for the development and design of a technology that aims to promote the human good.

Bayesian networks offer an extremely flexible environment for knowledge representation, so that they are often claimed to be the best statistical tools to support medical diagnosis. An Acyclic Directed Graph encodes causal relationships and provides a factorization of the joint probability distribution according to conditional independence properties. Although knowledge for the specification of the Acyclic Directed Graph is easilyretrievable, the information useful to develop the quantitative part of the network is typically scattered and varying in quality. For instance, medical literature seldom covers all the aspects of interest and clinical data are typically sparse. This is why the most relevant applications of Bayesian networks to medical diagnosis are entirely built from expert knowledge. However, when this is the case, the accuracy of the quantitative part remains arguable, since the the elicited information is unlikely to be fully trustworthy. In this thesis, the quantitative part of a Bayesian network for the diagnosis of cardiopulmonary diseases is estimated by combining elicitation from medical experts and clinical data. An original elicitation framework is developed to accurately quantify expert uncertainty on parameters, then prior distributions are updated in the light of data by means of Markov Chain Monte Carlo methods. The framework includes several generalizations of the Noisy-Or model and a Generalized Beta regression which are exploited to avoid polytomization of continuous variables. Parsimony in the number of parameters is dramatically improved with respect to the traditional framework, while a rescaling procedure based on a distinction between normal and pathological values makes parameters meaningful for medical experts. As such, the framework allows to either incorporate the sample size of clinical studies into the prior distributions or, when no study provides sufficiently detailed information, to compute expert uncertainty on assessments as if they were based on a virtual experiment. By taking advantage of two different sources of knowledge, the consistency between the model and data is readily checked by inspecting prior-to-posterior divergence. This enables a proper refinement of the Bayesian network in a cyclic-iterated fashion, possibly questioning the specification of the Acyclic Directed Graph, a feature which is beyond the capability of applications entirely built from either expert knowledge or data.

This presentation was one of several presentations delivered at the First International Open Access Day event held on October 14, 2008 at UBC. In support of the open access movement, the UBC Library joined with SPARC, PLoS (Public Library of Science), and Students for FreeCulture along with 65 other institutions in celebration of this worldwide event.