Development of an App to Support Field Diagnostics in Veterinary Medicine

Name: Christiane V Loehr
Affiliation: OSU College of Veterinary Medicine
Phone: 541-737-9673
E-mail: christiane.loehr@oregonstate.edu
Website: http://vetmed.oregonstate.edu/departments/academic-faculty/faculty/l%C3%B6hr; http://mime.oregonstate.edu/people/sharp
Knowledge Required: The composite team should have knowledge and skills in smartphone app development including use of onboard camera, computer vision/image analysis, and the ability to retrieve and compare data from a database to that acquired with a smartphone.
Motivation: The primary motivation is to provide expertise in veterinary pathology (postmortem examination of animals and strategies to test for diseases) in underserved regions such as farms in rural areas (e.g., Eastern Oregon), wildernesses, and developing countries with limited resources but access to cell phone or satellite service.
Description: This project provides a unique opportunity to partner with veterinary medicine and mechanical/humanitarian engineering to develop a prototype of an android app that can be used to enable remote diagnostics for animal diseases. The app will serve as the bridge between the on-the-ground field personnel and veterinary pathologists located at laboratories. The app will also provide instantaneous feedback to the photographer by rejecting images with poor lighting, saturation, and contrast, and other problems such as excessive highlights. In addition to supporting disease diagnostics for the remote user, the tool will enable collection of important epidemiological data that can aid in the identification of disease trends and provide critical inputs when dealing with issues of reportable diseases or issues of public health concern.
Objectives: The goal of the project is to design and implement a prototype of an android app for image and data collection from necropsies performed at the Oregon Veterinary Diagnostic Laboratory in the College of Veterinary Medicine on the OSU campus. The database architecture will be designed and implemented for images taken of bodies, organs, and lesions, and for metadata including information about the patient, location, and history of the case. (Database scope will be necessarily limited in order to have appropriate scope for this initial prototype). The app will be required to interface with this database. The app will also provide a platform to serve up feedback from the pathologist to the person in the field to help with identification of lesions and to make recommendations about sampling and additional testing.
Deliverables: The app will provide instantaneous feedback to the photographer by rejecting images with poor lighting, saturation, and contrast, and other problems such as excessive highlights. The app will capture images and other points (mostly text entry and GPS coordinates). A database will be designed and built to enable linking between the app and the database. The database will be limited in scope for this prototypical implementation but it should be easily searchable and expandable. The app will provide a platform to present feedback from the pathologist to the person in the field. Ideally, the prototype will pass a ‘field test’ where images are captured in the OVDL laboratory and relayed to a pathologist in another campus location, who can both access the database and provide real-time feedback to the test field user.
Other comments: This is not a project for the squeamish as images will be somewhat graphic. This is an excellent fit for a CS student with skills in image analysis, app development, database architecture and implementation, and interests in biomedical applications and information and communications technologies (ICT) for rural development.
This project is a collaboration between College of Veterinary Medicine (Christiane Loehr) and Humanitarian Engineering Program/Mechanical Engineering (Kendra Sharp)

   D. Kevin McGrath
   Last modified: Fri Oct 20 09:31:13 2017