An automatic vision-based malaria diagnosis system.

Malaria is a worldwide health problem with 225 million infections each year. A fast and easy-to-use method, with high performance is required to differentiate malaria from non-malarial fevers. Manual examination of blood smears is currently the gold standard, but it is time-consuming, labour-intensive, requires skilled microscopists and the sensitivity of the method depends heavily on the skills of the microscopist. We propose an easy-to-use, quantitative cartridge-scanner system for vision-based malaria diagnosis, focusing on low malaria parasite densities. We have used special finger-prick cartridges filled with acridine orange to obtain a thin blood film and a dedicated scanner to image the cartridge. Using supervised learning, we have built a Plasmodium falciparum detector. A two-step approach was used to first segment potentially interesting areas, which are then analysed in more detail. The performance of the detector was validated using 5,420 manually annotated parasite images from malaria parasite culture in medium, as well as using 40 cartridges of 11,780 images containing healthy blood. From finger prick to result, the prototype cartridge-scanner system gave a quantitative diagnosis in 16 min, of which only 1 min required manual interaction of basic operations. It does not require a wet lab or a skilled operator and provides parasite images for manual review and quality control. In healthy samples, the image analysis part of the system achieved an overall specificity of 99.999978% at the level of (infected) red blood cells, resulting in at most seven false positives per microlitre. Furthermore, the system showed a sensitivity of 75% at the cell level, enabling the detection of low parasite densities in a fast and easy-to-use manner. A field trial in Chittagong (Bangladesh) indicated that future work should primarily focus on improving the filling process of the cartridge and the focus control part of the scanner.

A preliminary report on culture of Leishmania donovani in Mymensingh Medical College and evaluation of new immuno-chromatography test (ICT).

A total of 51 inpatients having prolonged low grade irregular fever with anaemia, hepatosplenomegaly, emaciation and other allied features were evaluated by findings of haematologic, seroimmunologic, microscopic examination for LD body and culture for L. donovani. The study was done during the period from September' 1999 to January' 2000. Bone marrow or splenic aspirates were examined for LD bodies and those samples were cultured in modified NNN media following a standard method. Out of 51 samples, 36 (70.5%) were positive for LD bodies in stained smears and 38 (74.5%) were positive in culture. The mean time of culture positivity was 5 +/- 2 days. Specific antileishmanial antibody were detected by Immunochromatography Test (ICT) in all 38 confirmed cases, whereas Aldehyde test (AT) were negative in 3 such cases. ICT was positive in 1 and AT in 7 cases where no parasite could be detected. So, sensitivity of ICT was 100% with a specificity of 92.3% and sensitivity of AT was 92.1% with a specificity of 46.1%.

Treatment of pemphigus: a local experience.

A clinical study was conducted to evaluate the efficacy and safety of three currently used treatment modalities for pemphigus. Seventy-two patients of pemphigus (pemphigus vulgaris and severe cases of pemphigus vegetans, pemphigus foliaceous and pemphigus erythematosus) were divided into three groups. Forty patients were treated with prednisolone, 15 with a combination of prednisolone plus azathioprine and 17 patients with betamethasone-cyclophosphamide pulse (BCP) therapy. All patients were followed from 9 to 21 months (mean 16 months). There was no statistical difference between steroid and azathioprine-corticosteroid therapy groups in terms of time taken to achieve initial control of the disease but the frequency of relapses and the incidence of complications were higher in patients treated with corticosteroids alone (p < 0.05). A marginally increased susceptibility to infections was seen in patients treated with BCP therapy as compared with azathioprine-corticosteroid group (p = 0.07). Seventy percent patients treated with BCP therapy required additional steroids in between the monthly pulses, indicating failure of BCP as sole therapy. It was concluded that azathioprine-corticosteroid treatment of pemphigus was more effective and comparatively safer than steroid alone or BCP therapy.