Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 18, 2026Malaria JournalOpen Access

Deep learning convolutional neural networks for morphological identification of field-caught Anopheles stephensi in Ethiopia, using smartphone images of specimens

View Full Paper
Ask AI
Bookmark
Share

Authors

PMParthvi MehtaAKAjai KumarAZAtul Zacharias

Discussion

Loading...

Member takes

Overview

Validation study demonstrates high accuracy in identifying Anopheles stephensi via smartphone image classifiers, highlighting a scalable tool for decentralized malaria vector surveillance.

Key Points

  • To develop and evaluate deep learning convolutional neural network models capable of identifying the invasive malaria vector Anopheles stephensi from field-acquired smartphone images.
  • Fine-tuned an EfficientNet-B1 architecture using transfer learning on field-collected mosquito specimens from Ethiopia and Uganda, validated against PCR and morphological ground-truth identifications.
  • Trained and evaluated three models using 5-fold cross-validation and majority voting: a VectorCam multiclass classifier (7 species classes), a VectorCam binary classifier, and a handheld smartphone binary classifier using 10X and 15X clip-on macro lenses.
  • The VectorCam multiclass classifier achieved an An. stephensi sensitivity of 98.74% and an overall accuracy of 96.02%.
  • The VectorCam binary classifier achieved an An. stephensi sensitivity of 97.48% and specificity of 99.25%, whereas the handheld binary classifier achieved 91.09% sensitivity and 84.40% specificity.
  • Gradient-weighted Class Activation Mapping confirmed that model predictions were primarily driven by wing venation and leg morphology, matching expert entomological criteria.

Cite This Study

Mehta et al. (2026) studied this question.

synapsesocial.com/papers/6aad0bb6de0393d728b8a116https://doi.org/10.1186/s12936-026-06122-5
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Robust mosquito species identification from diverse body and wing images using deep learning2024 · 1 citations
  2. 2Robust mosquito species identification from diverse body and wing images using deep learning2024 · 7 citations
  3. 3Towards transforming malaria vector surveillance using VectorBrain: a novel convolutional neural network for mosquito species, sex, and abdomen status identifications2024
  4. 4AI-driven convolutional neural networks for accurate identification of yellow fever vectors2024 · 1 citations
  5. 5Potentials and limitations in the application of Convolutional Neural Networks for mosquito species identification using wing images2025