Automatic platelets counter for supporting dengue case detection in primary health care in indonesia

Stud Health Technol Inform. 2013:192:585-8.

Abstract

Dengue fever is a major problem in many developing countries, including Indonesia. Laboratory examination is used to diagnose dengue infection and to monitor disease progression. Hematology tests, such as platelet count, are also used for timely recognition of the development of severe dengue. In primary health care centers platelet counting is typically performed manually, which is labor intensive and requires an experienced laboratory technician. To address this challenge, we have developed an automatic platelet counter for primary health care and resource-poor settings. The technology is based on a conventional microscope equipped with a digital camera linked to a personal computer, which can capture and analyze microscope images of blood samples. To evaluate the accuracy of the technology, it was compared to platelet counts performed manual by an experienced laboratory technician. Statistical analysis shows no difference between the techniques with a kappa coefficient of 0.6. This method is judged to have great potential as a tool to help primary health centers and other facilities with limited resources to deal with the burden of dengue.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Artificial Intelligence
  • Blood Platelets / pathology*
  • Cells, Cultured
  • Dengue / blood*
  • Dengue / diagnosis*
  • Dengue / pathology
  • Diagnosis, Computer-Assisted / instrumentation
  • Diagnosis, Computer-Assisted / methods*
  • Equipment Design
  • Equipment Failure Analysis
  • Humans
  • Image Interpretation, Computer-Assisted / instrumentation
  • Image Interpretation, Computer-Assisted / methods*
  • Indonesia
  • Microscopy / instrumentation
  • Microscopy / methods*
  • Pattern Recognition, Automated / methods
  • Platelet Count / instrumentation
  • Platelet Count / methods*
  • Primary Health Care / methods
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Software