Protozoan cysts (Cryptosporidium oocysts and Giardia cysts) cause serious human health risks not only in urbanized areas but also in the cold and remote regions. Since these protozoan cysts are hardly inactivated in conventional drinking water treatment, reliable and rapid detection of the pathogenic cysts is urgently demanded, especially for communities without advance disinfection facilities, such as ozonation.

This project proposes to develop a novel sensor system where water samples are examined under optical/fluorescent microscopes and the pathogenic cysts on the microscopic images are detected by artificial intelligence (AI). Detailed research objectives and tasks are: (1) to build a sufficient database of microscopic image for machine learning training of AI; (2) to develop a filtration/resuspension system that selectively collects Cryptosporidium oocysts and Giardia cysts from other particles in natural water (natural organic matter particles, non-pathogenic microorganisms); and (3) to apply fluorescent-labeled antibodies and genomic fragment amplification to enhance the sensor accuracy and reliability. The new sensor system will be transformative as it will help improve water safety and control waterborne human diseases.