Detecting Birefringence of Hemozoin to Diagnose Malaria
Sheel M. Shah,1 Andrew M. Davis,1 Naman K. Shah,2 Steven R. Meshnick 2,3
1 Department s of Physics, 2 Epidemiology, 3 Microbiology & Immunology, University of North Carolina, Chapel Hill, NC, USA
ABSTRACT RESULTS CONCLUSIONS
• Malaria infects roughly half a billion people every year, • Sensitivity of signal and ease of detection is greatly
resulting in over a million deaths annually. increased with birefringence imaging
• As a host of inexpensive drug therapies exist to treat malaria, •Based on the increased signal sensitivity of using
the key to survival is early diagnosis and treatment 7. birefringent hemozoin6, images produced via hemozoin
• Early diagnosis might be possible using the birefringence birefringence could enable an automated, high-throughput,
property of hemozoin, a metabolic byproduct of the parasite. low cost diagnosis device.
• We used a collimated polarized beam incident on an infected • Diagnosis can be automated by translating a thick smear
sample to detect hemozoin with a polarizer completely out of sample using stepper motors and imaging random fields.
phase with the polarization of the beam before the CCD. Figure 1: Birefringent hemozoin superimposed on the same field as seen with ordinary
• The increased sensitivity of the signal and ease of detection FUTURE DIRECTIONS
lends itself to automated algorithms for diagnosis. Figure 2 shows an image acquired over two channels. The
• Develop a full working prototype with images being
red channel carried the transmitted light image and the green
exported to image analysis software loaded onto blackfin tm
channel carried the polarized light image. The images were
(Analog Devices) Processor.
superimposed using Adobe Photoshoptm .The hemozoin
BACKGROUND crystals can be clearly seen as the green artifacts on the red
•Determine correlation of the parasitemia level to hemozoin
background. Machine detection of the presence of the
• Current Diagnosis Techniques – Rapid Diagnostic Tests density in the blood smear
parasite in blood samples is greatly improved using
(RDT) utilize antigen detection, but have variable sensitivity,
birefringence detection. We successfully imaged hemozoin
low shelf life, and high cost. Microscopy is limited by the •Test the prototype with infected whole blood in the field.
in samples with parasitemia as low as 0.1%.
need for trained personnel to perform the test.
• There is a need to develop an automated, high-throughput,
accurate, and inexpensive device that will diagnose malaria Annie Purfield and Jaina Patel for slide preparation and
rapidly and efficiently. maintaining parasite cultures. Dr. Micheal Chua and
Dr. Bob Bagnell for guidance in the microscopy
METHODS facilities. Dr. Robert Dennis of TESLa Laboratories for
providing engineer space and materials. Dr. Jungsang
• P.falciparum (3D7) parasites from in vitro culture were Kim for general guidance with the project and feedback
reconstituted with packed RBC’s. 30 μL was dried on a slide of methods
as a thick smear and methanol fixed for 30 seconds.
•Images were taken using a Nikon TE2000 Inverted
microscope with a 20x objective and a 10x eye piece. REFERENCES
Images were acquired using the DVC-1412 monochrome
1. Scott C, van Zyl D, Ho E, Meyersfeld D, Ruvio L, Mendelow B,
• The collimated beam was polarized uniaxially and a Coetzer T, 2003. Automated Detection of Malaria-Associated
polarizer was inserted before the CCD in an orientation Intraleucocytic Haemozoin by Cell-Dyn CD4000 Depolarization
completely out of phase with the polarization of the beam. Analysis. Clin Lab Haematol 25.2: 77--86.
2. de Langen A, van Dillen J, de Witte P, Muchetto S, Nagelkerke
Figure 2: Initial Device Prototype N, Kager P, 2003. Detection of Malaria Pigment: Feasibility for
• Image analysis was performed using software based on Malaria Diagnosing in an Area with Seasonal Malaria in
ImageJ (NIH, Bethesda). Noise threshold levels were Figure 2 shows the initial device prototype. The Northern Namibia. Trop Med Int Health 11.6: 809--816.
calibrated using the images from the imaged P. falciparum detection of these crystals was easily automated using 3. Noland G, Briones N, Sullivan D Jr, 2003. The Shape and Size
negative samples. The area used for particle counting was few components: of Hemozoin Crystals Distinguishes Diverse Plasmodium
based on the size of hemozoin5. • Stepper motors and flexible drive shafts to provide Species. Mol Biochem Parasitol 130.2: 91--99.
sample translation 4. Lawrence C, 1999. Remembered: Malaria Haemozoin in
•An inexpensive microscope for imaging Leucocytes. Lancet 353.9167: 1852.
• Standard polarizing film. 5. Ross N, Pritchard C, Rubin D, Dusé A, 2006. Automated
Image Processing Method for the Diagnosis and Classification
of Malaria on Thin Blood Smears. Med Biol Eng Compu 44: