CombiANT Reader - Deep learning-based automatic image processing and measurement of distances to robustly quantify antibiotic interactions

Antibiotic resistance is a severe danger to human health, and combination therapy with several antibiotics has emerged as a viable treatment option for multi-resistant strains. CombiANT is a recently developed agar plate-based assay where three reservoirs on the bottom of the plate create a diffusion landscape of three antibiotics that allows testing of the efficiency of antibiotic combinations. This test, however, requires manually assigning nine reference points to each plate, which can be prone to errors, especially when plates need to be graded in large batches and by different users. In this study, an automated deep learning-based image processing method is presented that can accurately segment bacterial growth and measure more than 150 distances from key points on the CombiAnt assay at sub-millimeter precision. The software was tested on 100 plates using photos captured by three different users with their mobile phone cameras, comparing the automated analysis with the human scoring. The result indicates significant agreement between the users and the software. Moreover, the automated analysis remains consistent when applied to different photos of the same assay despite varying photo qualities and lighting conditions. The software can easily be integrated into a potential smartphone application. Integrating deep learning-based smartphone image analysis with simple agar-based tests like CombiANT could unlock powerful tools for combating antibiotic resistance. Author SummaryAntibiotic resistance is a significant problem worldwide with increasing prevalence of multi-resistant bacteria that may require the simultaneous administration of several different antibiotics. With the right antibiotics and concentration, such combination therapy may treat a strain that is otherwise resistant to each antibiotic individually. CombiANT is a novel test that can be used to identify suitable or inappropriate antibiotic combinations. However, it requires the human evaluator to grade each plate manually, which is time-consuming, and errors can easily be made, especially if the human evaluator needs to grade many plates in succession. In this study, an image processing pipeline is developed using a deep neural network to grade CombiANT test assays automatically.