Democritus University of Thrace (DUTH)
Υποστήριξη αποφάσεων για ιατρικές εφαρμογές με τεχνικές νευρωνικών δικτύων, στατιστικής ανάλυσης δεδομένων και γενετικών αλγορίθμων
Abstract
dc:descriptionIn the context of this dissertation the development of computational algorithms for the diagnosis and prognosis of diseases is achieved, in order to provide a useful tool for physicians. Additionally, different pruning techniques of the prognostic factors are investigated, such as Principal Component Analysis and Partial Least Squares Regression in order to keep only the useful information. This is the first time -according to the literature- that these methods are applied in childhood asthma and non-small cell lung cancer.Prediction is one of the most interesting areas where intelligent systems are utilized. In particular, prediction is an attempt to accurately forecast the evolution or outcome of a specific situation, using as input information a concrete set of variables that describe this situation. In medicine, the valid and effective interpretation of medical data, the correct and early diagnosis along with a documented prognostic evaluation of the clinical data are very important parameters for a better management of the disease.Prediction is a very difficult task because the expert human can hardly process the huge amount of data. The prediction accuracy is of major importance in the case of the asthma disease. The early identification of patients at risk for asthma persistence beyond the age of five may lead to better treatment opportunities and disease long term outcomes. The long-term solution to the asthma epidemic is believed to be prevention, and not treatment of the established disease. The most cases of asthma begin during the first years of life and thus the early determination which of the young children will have asthma later in their life is an important priority for public health.The same happens in the case of non-small cell lung cancer. The progress of survival of patients with non-small cell lung cancer is small despite the huge advances in oncology over the past decade. The group of patients with early lung cancer shows much better survival rates compared to that of advanced non-small cell lung cancer. Therefore, it is very important to find new biomarkers that will determine the prognosis of these patients.
Degree
thesis:*- Grantor dc:publisher
- Democritus University of Thrace (DUTH)
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Chatzimichail, Eleni
- Χατζημιχαήλ, Ελένη
Subjects
dc:subject × 20- Διάγνωση
- Πρόβλεψη
- Άσθμα
- Μη-μικροκυτταρικός καρκίνος των πνευμόνων
- Τεχνικές μείωσης διαστάσεων
- Τεχνητά νευρωνικά δίκτυα
- Γενετικοί αλγόριθμοι
- Μηχανές διανυσματικής υποστήριξης
- Diagnosis
- Prediction
- Asthma
- Non-small cell lung cancer
- Dimension reduction techniques
- Artificial neural networks
- Genetic algorithms
- Support vector machines
- Επιστήμες Μηχανικού και Τεχνολογία
- Βιοϊατρική Μηχανική
- Engineering and Technology
- Medical Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/34710