Frei zugänglich
Refine
Year of publication
Document Type
- Master's Thesis (19)
- Bachelor Thesis (4)
- Preprint (2)
- Diploma Thesis (1)
Language
- English (26) (remove)
Has Fulltext
- yes (26)
Is part of the Bibliography
- no (26)
Keywords
- Neuronales Netz (3)
- Application (2)
- Bildgebung (2)
- MRI (2)
- Maschinelles Lernen (2)
- Simulation (2)
- Software (2)
- stroke (2)
- web application (2)
- 3D Slicer (1)
Institute
- Informatik (13)
- Medizinische Informatik (12)
The e-commerce turnover has a constant growth rate of about 10%. An additional increase
in complexity and traffic spikes clarify the need for a scalable software architecture to prevent
a potential technical debt, higher financial cost, longer maintenance, or a reduced reliability.
Due to the fact, that existing approaches like the Palladio Approach require a high modelling
overhead and the importance of dropping this overhead was identified this master thesis is
focused on the modelling and simulation of e-commerce web application architectures using
a high-level approach to provide a faster, but possibly more inaccurate prediction of the
scalability.
This is done by the usage of the Design Science Research Process as a frame, a scientific
literature review for use of the existing knowledge base and the Conical Methodology for the
artefact creation. The artefact is a graphical model which is evaluated using a simulation
developed with Python and its framework SimPy. For model creation and evaluation a total
of twelve papers investigating the scalability of e-commerce web application architectures is
split into a test and train group. The training group and parts of the scientific research are
used to identify the components load balancer, application server, web tier, ERP system,
legacy system and database as well as some general characteristics that need to be considered.
The components with the most modelling variables are the application server and web
tier with a total of thirteen, while the ERP and legacy system only required five.
The model is evaluated using a total of three papers from the test group, where an average
throughput error of 5.78% and a response time error of 46.55% or 26.46% was identified. An
additional evaluation based on two non-e-commerce architectures shows the usability of the
model for other types of architectures. Even though the average error gives the impression,
that the model is not providing a good estimation, the graphical results show, that the model
and its simulation can be used to provide a faster scalability prediction. The model is least
accurate for the prediction of the situation, where the response time increases exponentially,
as this is the point, where variables, only accountable for some percentage and thus ignored
for the model, have the highest influence.
Future research can be found in the extension of the model by either adding or investigating
additional components, adding features ignored within this work or applying it to other
types of web application architectures. Additionally, both the low-level and the high-level
approaches can be brought together to combine the advantages from both approaches.
There are many drug interactions and to know every single interaction is impossible. In Uganda, a country located in East Africa, patients often do not get a patient information leflaet when a physician prescribes drugs because they only get the drugs without packaging and information inside. Even in developed countries many poeple die because of drug interactions.
This work aims at developing a clinical decision support system for different kinds of drug interactions: 1) drug-drug interaction,
2) drug-food interactions,
3) drug-condition
interactions and
4) drug-disease interactions.
This system must be integrated into an
existing hospital information system called electronic Health Management Information System (eHMIS).
In the first part of this thesis different kinds of clinical decision support systems are described to find out which one is the best for eHMIS. The two different types are knowledge-based and non knowledge-based systems. The second part of this thesis, the data base of eHMIS is extended to have a full
knowledge base for the new module which contains drug-drug interactions, drug-food interactions, drug-disease interactions as well as drug-condition interactions. Therefore new tables were created and filled with data of several data bases with drug interactions. The last part is about designing the clinical decision support system for drug interactions
with the knowledge base of eHMIS, including the implementation considering the integration into the existing system. To know how health professionals in Uganda work
with an electronical health system as well as their other work ows was important. The system now runs in a hospital in Kampala, the capital of Uganda and in a health center level three in Mifumi, a village located in the east of the country.
Background: Stroke rehabilitation is a complex process that requires collaboration between stroke patients
and various health professionals. One important component of the rehabilitation is to set goals collaboratively with health professionals. The goal setting process can be time-consuming. In many cases, it is complicated for the patient and difficult to track for the health professionals. A simple user interface that supports patients, their family members and health professionals can help both sides to make the goal setting and attainment process easier.
Objectives: The aim is to design and develop a software for the goal attainment process of stroke patients with milder disabilities that facilitates goal setting process and the traceability of the goal progress for patients and health professionals.
Methods: Based on previous evaluated results, the web interface was developed and improved. Using this knowledge, a goal setting interface was added. To analyze the the goal setting process, goal attainment scaling (GAS) was included as well as parts of the International Classification of Functioning, Disability and Health (ICF) core set for stroke. The results were discussed afterwards in focus groups and evaluated based on two stroke patients, one family member and health professionals.
Results: We developed an interactive prototype, that can aid the rehabilitation at home by inserting
problems with ICF codes and different kinds of goals, creating new activities and tracing goal progress by reviewing the different goals. With the help of the GAS the outcome of the patient’s goals are visualized by a line chart presenting the positive or negative outcomes of the stroke rehabilitation.
Conclusion: The interactive prototype showed that it can support stroke patients during their rehabilitation
at home. A usability test indicated that the goal setting and attainment process was perceived as useful for patients and their family members. Small improvements have to be made to simplify use and error handling. For health professionals, the prototype could also simplify the documentation process by using ICF in the prototype, and also improving collaboration when using the tool for coordination.
Medical imaging produces many images every day in clinical routine. Keeping up with the
daily image analysis task and this vast amount of data is quite a challenge for radiologists.
However, these analysis tasks can be automated with well-proven automatic segmentation
methods. Segmentation reviewing of an expert is necessary because learningbased
automatic segmentation methods may not perform well on exceptional image
data. Creating valid segmentations by reviewing them also improve the learning-based
methods.
Combining established standards with modern technologies creates a flexible environment
to efficiently evaluate multiple segmentation algorithm outputs based on different metrics
and visualizations and report these analysis results back to a clinical system environment.
The presented software system can inspect such quantitative results in a fast and intuitive
way, potentially improving the daily repetitive segmentation review and rework of a
research radiologist. The presented system is designed to be integrated into a virtual
distributed computing environment with other systems and analysis methods. Critical
factors for this particular environment are the handling of many patient data and routine
automated analysis with state of the art technology.
First experiments show that the time to review automatic segmentation results can be
roughly divided in half while the confidence of the radiologist is enhanced. The system
is also able to highlight individual slices which are essential for the expert’s review
decision. For this highlighting, different metric scores are compared and evaluated.
An architectural concept for implementing the socio-technical workflow of Digital Pathology in Chile
(2014)
Virtual Microscopy opens up the possibility to remotely access high quality images at large scales for scientific research, education, and clinical application. For clinical diagnostics, Digital Pathology (DP) presents a novel opportunity to reduce variability [Bauer et al., 2013] due to the reproducible access to Whole Slide Imaging, quantitative parameters (e.g. HER2 stained membrane) [Al-Janabi et al., 2012], second opinion and Quality Assurance [Ho et al., 2013]. Despite of the mentioned advantages, the challenge remains to incorporate DP into the pathologists workflow within a heterogeneous environment of systems and infrastructures [Stathonikos
et al., 2013]. Different issues must be solved in order to optimize the impact of DP in the daily clinical practice [Daniel et al., 2012] [Ho et al., 2006]. The integration needs precise planning and comprehensive evaluation for adopting this technology
[Stathonikos et al., 2013]. This thesis will focus on an organizational development approach based on a Socio-Technical System (STS). The socio-technical approach covers: (i) the technical issue: tissue-scanner, NDP.view, NDP.serve, analysis software, and (ii) the social issue: pathologists, technicians. In order to improve the integration, a joint optimization (of i and ii) is necessary. The developed STS approach will optimize the integration of DP towards improved workflows in clinical environments. The improved workflows will reduce the pathologists turnaround time, improve the certainty of the diagnostics, and provide a more effective patient care within the covered institutions. An overt multi-site Participatory Observation, Questionnaires, and Business Process Modelling Notation will be used to analyse the existing pathological workflows. Based on this, the system will be modelled with the 3lgm2 Toolkit [Winter et al., 2007] under consideration of various technical subsystems that are present in the clinical environment. Afterwards, the interfaces between subsystems and its possible interoperabilities will be evaluated, taking into account the different existing standards and guidelines for image processing and management, as well as business processes in DP. In order to analyse the existing preconditions a questionnaire will be evaluated to establish a robust and valid view. In addition, the overt participatory observation will support this elevation, giving a deeper insight on the social part. This observation also covers the technical side including the whole pathological process. The socio technical model will then reveal measurable potential for optimization with incorporated DP (e.g. higher throughput for slides). The organizational development approach consists of a Socio-Technical System based on overt multi-site participatory observations, questionnaires, business process modelling and 3LGM2, will optimize the use of Digital Pathology in the daily clinical practice and raise the acceptance to incorporate integrate the new technology within the dayly workflow through the user centred process of incorporation.
• Perform and evaluate a questionnaire and a participant observation of pathologists work days in private & public institutions
• Create and evaluate a 3lgm2 model
• Model the current pathological process (viewpoint of pathologist & technical assistant) & perform and evaluate a contextual inquiry to elevate the pathologists requirements & expectations towards the system
• Test the future WF according the model parameters.
This project will detect unsuspected interrelations and interdependencies within the socio- technical workflow with a pathology laboratory. The observation will reveal the action conformity as well as the environment in which the process has to be embedded. Furthermore it will establish an optimized workflow for a specific clinical environment to prepare the implementation of DP. Additionally it will be possible to
quantify digitized images in order to improve decision making and lastly to improve patient care. In the future it will be possible to extend automated image analysis in order to support clinical decision support. Depending on acceptance, this can lead towards an automated clinical decision support for cases with low complexity.
In this bachelor thesis, different models for predicting the influenza virus are
examined in more detail.
The focus is on epidemiological compartmental models, as well as on different
Machine Learning approaches.
In particular, the basics chapter presents the SIR model and its various extensions.
Furthermore, Deep Learning and Social Network approaches are
investigated and the applied methods of a selected article are analysed in more
detail.
The practical part of this work consists in the implementation of a Multiple
Linear Regression model and an Artificial Neural Network. For the development
of both models the programming language Python was chosen using the
Deep Learning Framework Keras.
Tests were performed with real data from the Réseau Sentinelles, a French
organisation for monitoring national health.
The results of the tests show that the Neural Network is able to make better
predictions than the Multiple Linear Regression model.
The discussion shows ideas for improving influenza prediction including the
establishment of a worldwide collaboration between the surveillance centres as
well as the consolidation of historical data with real-time social media data.
Therefore, this work consists of a state-of-the art of models regarding the
spread of influenza virus, the development and comparison of several models
programmed in Python, evaluated on real data.
In this thesis a software system is proposed that provides transparent access to dynamically processed data using a synthetic filesystem for the data transfer as well as interaction with the processing pipeline. Within this context the architecture for such a software solution has been designed and implemented. Using this implementation various profiling measurements have been acquired in order to evaluate the applicability in different data processing scenarios. Usability aspects, considering the interaction with the processing pipeline, have been examined as well. The implemented software is able to generate the processing result on-the-fly without modification of the original input data. Access to the output data is provided by means of a common filesystem interface without the need of implementing yet another communication protocol. Within the processing pipeline the data can be accessed and modified independently from the actual input and output encoding. Currently the data can be modified using a C/C++, GLSL or Java front end. Profiling data has shown that the overhead induced by the filesystem is negligible for most usage patterns and is only critical for realtime processing with a high data throughput e. g. video processing at or above 30 frames per second where typically no file operations are involved.
Access, Handling and Visualization Tools for Multiple Data Types for Breast Cancer Decision Support
(2011)
Breast cancer is the most commonly diagnosed cancer among U.S women, besides skin cancer. More than 1 in 4 cancers among women are breast cancer. And though death rates have been decreasing since 1990, about 40,170 women in the U.S. were expected to die in 2009 from breast cancer. The progress of molecular profiling, in the last decade has revolutionized the understanding of cancer, but also introduced more complexity with new data such as gene expression, copy number variation, mutations and DNA methylation. These new data open up the possibility of differential diagnosis, much more precise prognosis as well as prediction of therapy response than any of the diagnostic tools that are available in the current practice. Additionally, epidemiological databases store clinically relevant information on hundreds of thousands of patients. However, with the abundance of all this information, clinicians will need new tools to access and visualize such data and use the information gained to treat new patients. The general problem will be to access, filter and analyze the data and then visualize them in a clinical context. This data ranges from clinico-pathological information, to molecular profiles from highthroughput genomic measurements and imaging data. Furthermore, data from patient populations is aggregated on epidemiological level and can be found under numerous clinical studies.
Background: An important factor in approaching the challenges of chronic diseases, requiring long-term management and high costs, is the active participation of the patient in the care process. Objectives: Facing the problem of lacking patient-tailored, comprehensive health management software, the aim of this thesis is to generate ideas for a graphical user interface (GUI) to support stroke patients in the management of their individual care process. The objectives are to prototype a GUI for a patient e-service and to evaluate its usefulness and usability with stroke patients. Methods: A scenario-based, user-centered design method was used to envision ideas for the user interface. Static prototypes were realized with the tool Pencil and for the implementation of a dynamic prototype web programming techniques were used. For the evaluation of the prototypes the methods of focus group discussion and cooperative evaluation were applied. Results: The situation of a representative stroke patient and his interaction with the e-service were described in scenarios. Graphical user interfaces of the involved system views were derived from the scenarios and illustrated with static wireframe prototypes. A welcome screen, a care process timeline overview, and a diary with data sharing functionality were designed. The diary functionality was further examined by implementing a prototypical web application. During the evaluation, feedback for further improvements was gathered, and assumptions about the user information and functionality needs could be verified. Conclusion: The developed prototypes represent a suitable graphical user interface and visualizations to support stroke patients in the management of their care process. An overview of appointments on the welcome screen, a diary to document and monitor health, a timeline overview of all time-related health information and a selected sharing functionality were found to be important features of a personal health system for stroke patients.
Implementation of an interactive pattern mining framework on electronic health record datasets
(2019)
Large collections of electronic patient records contain a broad range of clinical information highly relevant for data analysis. However, they are maintained primarily for patient administration, and automated methods are required to extract valuable knowledge for predictive, preventive, personalized and participatory medicine. Sequential pattern mining is a fundamental task in data mining which can be used to find statistically relevant, non-trivial temporal dependencies of events such as disease comorbidities. This works objective is to use this mining technique to identify disease associations based on ICD-9-CM codes data of the entire Taiwanese population obtained from Taiwan’s National Health Insurance Research Database.
This thesis reports the development and implementation of the Disease Pattern Miner – a pattern mining framework in a medical domain. The framework was designed as a Web application which can be used to run several state-of-the-art sequence mining algorithms on electronic health records, collect and filter the results to reduce the number of patterns to a meaningful size, and visualize the disease associations as an interactive model in a specific population group. This may be crucial to discover new disease associations and offer novel insights to explain disease pathogenesis. A structured evaluation of the data and models are required before medical data-scientist may use this application as a tool for further research to get a better understanding of disease comorbidities.