[cosc-grad-students-list] RESEARCH PROPOSAL SEMINAR NOTICE - Madhukar Kuchavaram: GEOAI-ENABLED FRAMEWORK FOR MOSQUITO SURVEILLANCE AND VECTOR-CONTROL DECISION SUPPORT

[Eulenfeld, Menda]

We cordially invite you to attend Madhukar Kuchavaram's Research Proposal. Your support would be great!

RESEARCH PROPOSAL SEMINAR NOTICE
GEOSPATIAL COMPUTER SCIENCE PROGRAM
COLLEGE OF ENGINEERING AND COMPUTER SCIENCE
TEXAS A&M UNIVERSITY-CORPUS CHRISTI

SUBJECT: GEOAI-ENABLED FRAMEWORK FOR MOSQUITO SURVEILLANCE AND VECTOR-CONTROL DECISION SUPPORT
SPEAKER: Madhukar Kuchavaram
ADVISOR: Dr. Yuxia Huang
COMMITTEE: Dr. Wenlu Wang, Dr. Rajesh Melaram, Dr. Zheng Wei, and Dr. Lei Jin (Graduate Faculty Representative)
DATE: October 16, 2026 (Friday)
TIME: 1:00 PM Central Time (US and Canada)
PLACE: NRC 1232 Conference Room at CBI
Microsoft Teams Link: https://teams.microsoft.com/meet/239076420565400?p=9La9C9GkMNkLcDmYjU<https://nam12.safelinks.protection.outlook.com/ap/t-59584e83/?url=https%3A%2F%2Fteams.microsoft.com%2Fmeet%2F239076420565400%3Fp%3D9La9C9GkMNkLcDmYjU&data=05%7C02%7Ccosc-grad-students-list%40listserv.tamucc.edu%7C4cb96937ab73419b053c08df2943ddea%7C34cbfaf167a64781a9ca514eb2550b66%7C0%7C0%7C639275043626970755%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=%2B8lpTpm86M1bK89CqIc3t8Lt1puxR0OI08Fpy3IcKow%3D&reserved=0>
Meeting ID: 239 076 420 565 400
Passcode: 48Fp6rp7

Abstract
Mosquito-borne diseases such as West Nile virus and dengue remain a persistent public health burden. In the United States, municipal and county vector-control programs conduct mosquito surveillance and control. These programs depend on labor-intensive manual counting and on scarce taxonomic expertise for species identification, while trap counts, laboratory results, and field operations are often held in separate records. Although automated counting and image-based identification have advanced, few tools have been evaluated within an operating program, and the spatial interpretation of sparse trap networks remains limited. This dissertation develops an integrated GeoAI-enabled framework for mosquito surveillance and vector-control decision support, built in collaboration with the Vector Control program of the Corpus Christi-Nueces County Public Health District, Texas. The research begins with the design and deployment of MOSAIC, a web-based geospatial platform that links each trap collection to its location, photograph, count, laboratory results, and expert determinations. The platform provides computer-vision-assisted counting, while an associated offline evaluation examines multimodal large language model (LLM) species identification using a local species list, taxonomic knowledge, and labeled exemplars against expert-reviewed reference labels. Building on these records, the study interprets trap evidence spatially, using mixed-effects models and rank-concordance statistics to quantify differences between trap locations and their persistence, and a capture-probability model adapted from MGSurvE to identify areas weakly represented by the trap network. Trap evidence and spatial context are then integrated in a two-stage multi-criteria framework, with expert-elicited weights and Stochastic Multicriteria Acceptability Analysis, to prioritize weekly inspection. Finally, the proposed work extends the framework to network design, formulating trap placement as a constrained optimization problem to improve modeled coverage in areas weakly represented by the existing network. Together, these components integrate AI-assisted specimen processing, spatial interpretation, and network design within a common surveillance framework, offering an interpretable, evidence-based approach that vector-control programs can adapt to their own data and operational needs.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.tamucc.edu/pipermail/cosc-grad-students-list/attachments/20261013/24f613d6/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/calendar
Size: 9743 bytes
Desc: not available
URL: <http://listserv.tamucc.edu/pipermail/cosc-grad-students-list/attachments/20261013/24f613d6/attachment-0001.ics>