Answers are grounded in this paper and cite their sources.
Airborne rubber powdery mildew conidia are difficult to collect and distinguish from morphologically similar particles in microscopic samples. This study investigated a microfluidic chip combined with SGS-YOLO-assisted image analysis for the localization and identification of Oidium heveae conidia under controlled laboratory conditions. The microfluidic chip was designed to regulate airflow distribution and promote the localization of airborne spore-type particles in the collection region. COMSOL simulations characterized the velocity and pressure distributions within the chip. Aerosol sampling experiments were conducted at five inlet flow rates ranging from 35 to 55 mL/min, with a sampling duration of 30 min and three independent replicates per condition. At 45 mL/min, the O. heveae distribution fraction reached 88.5 ± 3.7%, while the target-particle proportion in the collection region reached 95.8 ± 1.5%. To further distinguish the target conidia from interfering particles, a lightweight SGS-YOLO model was developed using a separately acquired six-class microscopic image dataset. The model contained 0.793 million parameters and achieved an AP50 of 0.968 ± 0.003 for O. heveae conidia. These findings support the feasibility of combining microfluidic particle localization with image-based identification under controlled laboratory conditions.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: