METHODS: We conducted a two year study in a high human density dengue-endemic urban area in Selangor, where Gravid Ovipositing Sticky (GOS) traps were set up to capture adult Aedes spp. mosquitoes. All Aedes mosquitoes were tested using the NS1 dengue antigen test kit. All dengue cases from the study site notified to the State Health Department were recorded. Weekly microclimatic temperature, relative humidity (RH) and rainfall were monitored.
RESULTS: Aedes aegypti was the predominant mosquito (95.6%) caught in GOS traps and 23% (43/187 pools of 5 mosquitoes each) were found to be positive for dengue using the NS1 antigen kit. Confirmed cases of dengue were observed with a lag of one week after positive Ae. aegypti were detected. Aedes aegypti density as analysed by distributed lag non-linear models, will increase lag of 2-3 weeks for temperature increase from 28 to 30 °C; and lag of three weeks for increased rainfall.
CONCLUSION: Proactive strategy is needed for dengue vector surveillance programme. One method would be to use the GOS trap which is simple to setup, cost effective (below USD 1 per trap) and environmental friendly (i.e. use recyclable plastic materials) to capture Ae. aegypti followed by a rapid method of detecting of dengue virus using the NS1 dengue antigen kit. Control measures should be initiated when positive mosquitoes are detected.
METHODS: The YOLOv4 model is modified using direct layer pruning and backbone replacement. The primary objective of layer pruning is the removal and individual analysis of residual blocks within the C3, C4 and C5 (C3-C5) Res-block bodies of the backbone architecture's C3-C5 Res-block bodies. The CSP-DarkNet53 backbone is simultaneously replaced for enhanced feature extraction with a shallower ResNet50 network. The performance metrics of the models are compared and analysed.
RESULTS: The modified models outperform the original YOLOv4 model. The YOLOv4-RC3_4 model with residual blocks pruned from the C3 and C4 Res-block body achieves the highest mean accuracy precision (mAP) of 90.70%. This mAP is > 9% higher than that of the original model, saving approximately 22% of the billion floating point operations (B-FLOPS) and 23 MB in size. The findings indicate that the YOLOv4-RC3_4 model also performs better, with an increase of 9.27% in detecting the infected cells upon pruning the redundant layers from the C3 Res-block bodies of the CSP-DarkeNet53 backbone.
CONCLUSIONS: The results of this study highlight the use of the YOLOv4 model for detecting infected red blood cells. Pruning the residual blocks from the Res-block bodies helps to determine which Res-block bodies contribute the most and least, respectively, to the model's performance. Our method has the potential to revolutionise malaria diagnosis and pave the way for novel deep learning-based bioinformatics solutions. Developing an effective and automated process for diagnosing malaria will considerably contribute to global efforts to combat this debilitating disease. We have shown that removing undesirable residual blocks can reduce the size of the model and its computational complexity without compromising its precision.