<p>In this work, we developed an automated system for the detection and classification of soil-transmitted helminths (STH) and <i>Schistosoma (S.) mansoni</i> eggs in microscopic images of fecal smears. We assembled an STH and <i>S. mansoni</i> dataset comprising over 3,000 field-of-view (FOV) images containing parasite eggs, extracted from more than 300 fecal smear prepared using the Kato-Katz technique. These images were acquired using Schistoscope—a cost-effective automated digital microscope. After annotating the STH and <i>S. mansoni</i> eggs, we employed a transfer learning approach to train an EfficientDet deep learning model, using 70% of the dataset for training, 20% for validation, and 10% for testing. The developed model successfully identified STH and <i>S. mansoni</i> eggs in the FOV images, achieving weighted average scores of <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_2755_Article_IEq1.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="103" /> </InlineMediaObject> <EquationSource Format="TEX">\(95.9\%(\pm 1.1\%)\)</EquationSource> </InlineEquation> Precision, <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_2755_Article_IEq2.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="103" /> </InlineMediaObject> <EquationSource Format="TEX">\(92.1\%(\pm 3.5\%)\)</EquationSource> </InlineEquation> Sensitivity, <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_2755_Article_IEq3.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="111" /> </InlineMediaObject> <EquationSource Format="TEX">\(98.0\%(\pm 0.76\%)\)</EquationSource> </InlineEquation> Specificity, and <InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_2755_Article_IEq4.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="111" /> </InlineMediaObject> <EquationSource Format="TEX">\(94.0\%(\pm 1.98\%)\)</EquationSource> </InlineEquation> F-Score across four classes of helminths (<i>A. lumbricoides</i>, <i>T. trichiura</i>, hookworm, and <i>S. mansoni</i>). Our system highlights the potential of the Schistoscope, enhanced with artificial intelligence, for detecting STH and <i>S. mansoni</i> infections in remote, resource-limited settings and for supporting the monitoring and evaluation of neglected tropical disease (NTD) control programs.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Deep learning-based automated detection and multiclass classification of soil-transmitted helminths and Schistosoma mansoni eggs in fecal smear images

  • Prosper Oyibo,
  • Brice Meulah,
  • Tope Agbana,
  • Lisette van Lieshout,
  • Wellington Oyibo,
  • Gleb Vdovin,
  • Jan-Carel Diehl

摘要

In this work, we developed an automated system for the detection and classification of soil-transmitted helminths (STH) and Schistosoma (S.) mansoni eggs in microscopic images of fecal smears. We assembled an STH and S. mansoni dataset comprising over 3,000 field-of-view (FOV) images containing parasite eggs, extracted from more than 300 fecal smear prepared using the Kato-Katz technique. These images were acquired using Schistoscope—a cost-effective automated digital microscope. After annotating the STH and S. mansoni eggs, we employed a transfer learning approach to train an EfficientDet deep learning model, using 70% of the dataset for training, 20% for validation, and 10% for testing. The developed model successfully identified STH and S. mansoni eggs in the FOV images, achieving weighted average scores of \(95.9\%(\pm 1.1\%)\) Precision, \(92.1\%(\pm 3.5\%)\) Sensitivity, \(98.0\%(\pm 0.76\%)\) Specificity, and \(94.0\%(\pm 1.98\%)\) F-Score across four classes of helminths (A. lumbricoides, T. trichiura, hookworm, and S. mansoni). Our system highlights the potential of the Schistoscope, enhanced with artificial intelligence, for detecting STH and S. mansoni infections in remote, resource-limited settings and for supporting the monitoring and evaluation of neglected tropical disease (NTD) control programs.