• DocumentCode
    2425066
  • Title

    Early Diagnosis and Its Benefits in Sepsis Blood Purification Treatment

  • Author

    Ghalwash, Mohamed ; Radosavljevic, Vladan ; Obradovic, Z.

  • Author_Institution
    Comput. & Inf. Sci., Temple Univ., Philadelphia, PA, USA
  • fYear
    2013
  • fDate
    9-11 Sept. 2013
  • Firstpage
    523
  • Lastpage
    528
  • Abstract
    Sepsis is a progressive medical condition characterized as an uncontrolled inflammatory response, which is the leading cause of death in non-coronary intensive care units in the United States. In sepsis treatment, accurate and timely diagnosis is essential for allowing physicians to design appropriate therapeutic strategies at early stages, when therapies are usually the most effective and the least costly. To make an adequate diagnosis, physicians usually rely on manual inspection of a large amount of complex, high-dimensional longitudinal data. We use our recently published data mining method for extracting patterns from such data and show that these patterns can be used to assist physicians in providing early diagnosis. In conducted experiments, we showed that combination of early diagnosis and blood purification therapy can rescue more patients (52%) than standard approach for blood purification therapy (32%). We also propose a hybrid therapy model that combines strengths of early and standard approaches and further improves the percentage of rescued patients. Finally, by correctly classifying 98% of patients who didn´t need treatment, MSD method provides opportunity to reduce the total cost of treatments.
  • Keywords
    blood; data mining; medical computing; patient care; patient diagnosis; patient treatment; MSD method; United States; blood purification therapy; data mining method; early diagnosis; high-dimensional longitudinal data; hybrid therapy model; noncoronary intensive care units; physicians; progressive medical condition; rescued patients; sepsis blood purification treatment; uncontrolled inflammatory response; Blood; Hidden Markov models; Mathematical model; Medical treatment; Standards; Time series analysis; Training;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Healthcare Informatics (ICHI), 2013 IEEE International Conference on
  • Conference_Location
    Philadelphia, PA
  • Type

    conf

  • DOI
    10.1109/ICHI.2013.81
  • Filename
    6680529