Caracterización de la variabilidad del ritmo cardiaco mediante modelos ocultos de Markov

• Autores: Manuel Eduardo Palacios Muñoz
• Directores de la Tesis: Montserrat Vallverdú Ferrer (dir. tes.)
• Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2007
• Idioma: español
• Tribunal Calificador de la Tesis: Pere Caminal Magrans (presid.), Francesc Claria Sancho (secret.), Miquel A. Garcia Gonzalez (voc.), Luca Mainardi (voc.), Pedro Gomis Roman (voc.)
• Materias:
  • Matemáticas
    • Investigación operativa
      • Programación no lineal
    • Estadística
      • Teoría estocástica y análisis de series temporales
  • Ciencias médicas
    • Medicina interna
      • Cardiología
• Texto completo no disponible (Saber más ...)
• Resumen

  • This PhD thesis has focused on the propose, design and evaluation a new methodology based on hidden Markov models (HMM) to characterize and to identify aspects of the complex autonomic communication present in the rhythms of the autonomous nervous system (ANS), i,e., the autonomic information flow (AIF) by the study of the RR series. The proposed methodology has been applied to the RR series obtained by linear interpolation of the RR tachogram, resampled and filtered in the frequency bands VLF (0.003-0.04 Hz), LF (0.04-0.15 Hz) and HF (0.15-0.45 Hz): HF RR , LF RR , VLF RR . Also the RR series have been considered without filtering RES RR .

    In chapter 1, the problems of the analysis of the heart rate variability are introduced together with an explanation of the organization of this PhD thesis. In chapter 2, the methods of analysis of the heart rate variability are discussed, together with the technical characteristics and the groups of patients of the databases used in this thesis, IDEAL (Intercity Digital ECG Alliance, Rochester University, USA) and NIC (National Institute of Cardiology of Warsaw, Poland). Also, physiologies of two types of Cardiomyopathy investigated in this thesis are explained. In chapter 3, the mathematical theory of the hidden Markov models and the HMM learning algorithms are detailed, and architectures of the models used in this thesis are described. In chapter 4, a new methodology of analysis of the heart rate variability has been identified by means of HMM, obtained by the transformation of the RR series in sequences of words provided by a nonlinear methodology of symbolic dynamics. The new methodology was applied to the IDEAL database. Finally, the results of the quantitative evaluation of the new proposed methodology are discussed. In chapter 5, an analysis based on nonlinear methodology of mutual information and a new methodology based on the final structure of the hidden Markov models are proposed. The RR series filtered according to