Fast and cost-effective machine learning based system for microorganism growth analysis

Project title:

“Fast and cost-effective machine learning based system for microorganism growth analysis”

Project contract number:

1.1.1.1/19A/147

Project partners:

SIA "Laboratorija AUCTORITAS (Edgars Baranovičs)

Riga Technical University (Dmitrijs Bļizņuks)

Project implementation deadline:

36 months (01.07.2020. - 30.06.2023.)

Total project funding:

647 881,72 EUR

Research manager:

Aleksejs Ļihačovs (Biophotonics laboratory)

Project administrator:

Liene Ozoliņa

About the project:

The aim of the project is to develop and validate a new, cost-effective system for the early evaluation of the activity of microorganisms in solid media. The microorganism activity assessment system will use non-contact laser spectral contrast imaging technology in combination with deep machine learning based on artificial neural network (MNT) image processing. The system will significantly reduce the time required to obtain microbial test results and automate bacterial counting and classification. During the project, the proposed prototype of the device and the algorithm based on the MNT algorithm will be tested in real laboratory conditions and validated in a microbiological test laboratory, where microbiological tests of various food and environmental samples are performed. The project team will consist of experienced microbiology, biophotonics, electronics and information technology experts, young scientists and students from the University of Latvia, Riga Technical University and private food hygiene laboratory SIA “AUCTORITAS”.

Expected main results include:

1. Portable system for laser spot generation and image acquisition (TRL4);
2. Improved MNT-based algorithms for laser specimen image analysis and microbial activity assessment;
3. Stand-alone, compact laser specimen display device for counting and classifying microorganisms (TRL7);
4. 9 original scientific articles (including in three journals with a high impact factor), intellectual property (know-how) of the developed product, as well as a license agreement (to be concluded after the project implementation);
5. A new method for the rapid enumeration of colony - forming units of micro - organisms, used to quantify the total number of colony - forming units and to identify specific groups of bacteria.