Mobile Bot Application for Identification of Trypanosoma evansi Infection through Thin-Blood Film Examination Based on Deep Learning Approach

Simple item page
dc.contributor.authorRangsan Jomtarak
dc.contributor.authorVeerayuth Kittichai
dc.contributor.authorMorakot Kaewthamasorn
dc.contributor.authorSuchansa Thanee
dc.contributor.authorApinya Arnuphapprasert
dc.contributor.authorKaung Myat Naing
dc.contributor.authorTeerawat Tongloy
dc.contributor.authorSiridech Boonsang
dc.contributor.authorSanthad Chuwongin
dc.contributor.correspondenceS. Chuwongin; College of Advanced Manufacturing Innovation, King Mongkut's Institute of Technology Ladkrabang, Bangkok, Thailand; email: santhad.ch@kmitl.ac.th
dc.date.accessioned2025-03-10T07:34:44Z
dc.date.available2025-03-10T07:34:44Z
dc.date.issued2023
dc.description.abstractTrypanosomiasis caused Trypanosoma evansi is current public health concern especially, in south Asia and Southeast Asia. Although polymerase chain reaction is currently used as a standard method, the techniques required skilled personnel, were performed in multiple steps, and required expensive instruments. Fundamental microscopic approach also has limitation in use by facing both inter- and intra-variability of interpretation by examiners. New automatic tool with the microscopic examination is needed. The study aimed to develop the mobile application-based YOLO neural network algorithms to predict T. evansi blood stages from thin-blood film examination. YOLO v4 tiny model is outperformed to localize and classify unseen images with the best performance at 95% of sensitivity, specificity, precision, accuracy and F1 score, respectively, with less misclassification rate than 5%. Simulation implementation platform, calling CiRA bot, give the empirical result and reliably comparable to that from the computational experiment studied with the area under ROC and precision-recall curves as 0.964 and 0.962, respectively. The result obtained from the CIRA bot platform is good enough for further distribution in field site. In the future, the study could contribute human and animal public health staff to simply identify the unicellular parasitic flagellate infection and also benefit them for designing the strategy in prevention and treatment of the disease. © 2023 IEEE.
dc.identifier.citationInternational Conference on Cybernetics and Innovations, ICCI 2023
dc.identifier.doi10.1109/ICCI57424.2023.10112327
dc.identifier.isbn979-835033657-3
dc.identifier.scopus2-s2.0-85159860271
dc.identifier.urihttps://repository.dusit.ac.th//handle/123456789/4546
dc.languageEnglish
dc.publisherInstitute of Electrical and Electronics Engineers Inc.
dc.rights.holderScopus
dc.subjecta microscopic examination
dc.subjectMobile application
dc.subjectmodel performance
dc.subjectTrypanosoma evansi
dc.subjectYOLO algorithms
dc.titleMobile Bot Application for Identification of Trypanosoma evansi Infection through Thin-Blood Film Examination Based on Deep Learning Approach
dc.typeConference paper
mods.location.urlhttps://www.scopus.com/inward/record.uri?eid=2-s2.0-85159860271&doi=10.1109%2fICCI57424.2023.10112327&partnerID=40&md5=c05d53f1787eaaca52c264620e74438d
Files
Collections
SCOPUS 2023