Automating the process of antibiotic susceptibility testing

摘要

The proposed project presents a methodology to detect how susceptible or resistant certain bacteria are to an applied antibiotic. This detection is achieved by calculating the area of Zone of Inhibition (ZOI) regions present in the petri dish and comparing the results to the prescribed standards. The ZOI regions are empty areas formed around an antibiotic disc when placed on a petri dish containing a sample of the bacterial culture. Digital image processing techniques are employed to automate the process of ZOI detection. Experimental results show that the proposed project is successful in detecting ZOI regions of various shapes, such as perfectly circular, irregular, and overlapping. The experimental results also show that the accuracy of detection is typically over 95%, and it remains above 90%, even when the image is degraded by additive Gaussian noise.