Introducing BAIT (Biofilm Architecture Inference Tool): a software program to evaluate the architecture of oral...

Biofilm model systems are used to study biofilm growth and predict the effects of anti-biofilm interventions within the human oral cavity. Many in vitro biofilm model systems use a confocal laser scanning microscope (CLSM) in conjunction with image analysis tools to study biofilms. The aim of this study was to evaluate an in-house developed image analysis software program that we call BAIT (Biofilm Architecture Inference Tool) to quantify the architecture of oral multi-species biofilms following anti-biofilm interventions using a microfluidic biofilm system. Differences in architecture were compared between untreated biofilms and those treated with water (negative control), sodium gluconate ('placebo') or stannous fluoride (SnF2). The microfluidic system was inoculated with pooled human saliva and biofilms were developed over 22 h in filter-sterilized 25 % pooled human saliva. During this period, biofilms were treated with water, sodium gluconate, or SnF2 (1000, 3439 or 10 000 p.p.m. Sn²⁺) 8 and 18 h post-inoculation. After 22 h of growth, biofilms were stained with LIVE/DEAD stain, and imaged by CLSM. BAIT was used to calculate biofilm biovolume, total number of objects, surface area, fluffiness, connectivity, convex hull porosity and viability. Image analysis showed oral biofilm architecture was significantly altered by 3439 and 10 000 p.p.m. Sn²⁺ treatment regimens, resulting in decreased biovolume, surface area, number of objects and connectivity, while fluffiness increased (P<0.01). In conclusion, BAIT was shown to be able to measure the changes in biofilm architecture and detects possible antimicrobial and anti-biofilm effects of candidate agents.