Malaria parasite detection and cell counting for human and mouse using thin blood smear microscopy

Area of Science:

• Medical Diagnostics
• Computational Biology
• Parasitology

Background:

• Malaria remains a significant global health challenge, with inadequate diagnostics hindering effective control efforts.
• Automated systems offer potential for faster and more reliable malaria screening.
• Accurate detection and quantification of infected red blood cells (RBCs) are crucial for diagnosis and monitoring.

Purpose of the Study:

• To develop and evaluate an automated system for detecting and segmenting RBCs and identifying malaria-infected cells in stained blood smears.
• To assess the system's performance using image analysis and machine learning techniques.
• To establish a reliable method for determining parasitemia in both human and mouse blood samples.

Main Methods:

• Development of an automated system employing image analysis and machine learning for malaria diagnosis.
• Utilizing a cell extraction method for segmenting red blood cells, including overlapping cells.
• Feature extraction combining RGB color and texture characteristics for improved cell identification.

Main Results:

• The automated system accurately detects and segments red blood cells and identifies infected cells in Wright-Giemsa stained thin blood smears.
• A combination of RGB color and texture features demonstrated superior performance compared to other feature sets.
• The system achieved an absolute error of 1.18% for human cell parasite counts and showed strong correlation with expert and flow cytometry counts for mouse cells.

Conclusions:

• The developed automated system provides a fast and reliable method for malaria diagnosis.
• This system is the first to successfully analyze both human and mouse blood smears for malaria detection.
• The findings highlight the potential of image analysis and machine learning in advancing malaria diagnostics.