【摘要】 目的研发自动疟原虫图像扫描及识别系统,用以代替传统人工镜检,提高疟疾检测工作效率和准确率。方法使用疟原虫自动扫描及识别系统对保存的疟疾病例血涂片进行扫描及识别,与人工镜检结果进行比较分析。结果自动疟原虫显微扫描识别系统可实现对薄血片的快速扫描、图像保存及阳性图像标注,对疟原虫阳性涂片的检出率为96.3%,与人工阅片检出结果一致,两者无统计学差异;对恶性疟原虫虫种鉴别的准确率最高可达到90.9%。结论疟原虫自动图像扫描及识别系统可对薄血片中的疟原虫进行准确快速的识别,可极大提高疟疾镜检工作效率和准确性。更多还原

【Abstract】 Objective To develop an automatic malaria image scanning and recognition system to replace the traditional manual microscopy and improve the efficiency and accuracy of malaria detection. Methods The blood smears of patients with malaria were scanned and identified by the automatic scanning and identification system of plasmodium,and the results were compared with those of manual mi croscopy. Results The system could quickly scan, save and label the positive image of thin blood smear. The detection rate of positive smear was 96.3%,which was consistent with that of manual reading;The highest accuracy of plasmodium falciparum species identification could reach 90.9%. Conclusion The automatic image scanning and recognition system of plasmodium can identify the plasmodium in thin blood film accurately and quickly,which can greatly improve the efficiency and accuracy of malaria microscopic examination.更多还原