Biofilm Detection by a Fiber-Tip Ball Resonator Optical Fiber Sensor

Abstract

Bacterial biofilms are one of the most important challenges that modern medicine faces due to the difficulties of diagnosis, antibiotic resistance, and protective mechanisms against aggressive environments. For these reasons, methods that ensure the inexpensive and rapid or real-time detection of biofilm formation on medical devices are needed. This study examines the possibilities of using optical- and fiber-based biosensors to detect and analyze early bacterial biofilms. In this study, the biofilm-forming model organism Pseudomonas aeruginosa was inoculated on the surface of the optical sensor and allowed to attach for 2 h. The biosensors were made by a fiber-tip ball resonator, fabricated through a CO₂ laser splicer on a single-mode fiber, forming a weak reflective spectrum. An optical backscatter reflectometer was used to measure the refractive index detected by the sensors during different growth periods. The early biofilm concentration was determined by crystal violet (CV) binding assay; however, such a concentration was lower than the detection limit of this assay. This work presents a new approach of biofilm sensing in the early attachment stage with a low limit of detection up to 10^-4 RIU (refractive index units) or 35 ± 20 × 10³ CFU/mL (colony formed units).

Keywords: biofilm detection; biofilm formation; biomedical sensors; distributed sensors; fiber-tip ball resonator; optical fiber sensor.

Figure 1

Illustration depicting the biofilm formation and biofilm markers. Left: artwork describing the process of biofilm formation on the surface of an optical fiber spherical tip used as biosensor. Right: equivalent optical model for the early stage growth, showing a ball resonator surrounded by a medium with a varying refractive index.

Figure 2

Ball resonators fabrication: (a) BRs profilometry; (b) microscopic view of BR in the Fujikura’s interaction window with the operator; (c) BRs 3D modeling view; (d) BR view with 500 um diameter; (e) spectral change in BR during calibration with 200 μL steps of sucrose 40, while the inset shows the spectral feature detected by intensity measurement; (f) fitting of calibration R² = 0.9997, and sensitivity is −108.38 dB/RIU.

Figure 3

Set up of the experiment: (a) OBR Luna 4600 with a computer used for data acquisition; (b) incubator; (c) rack with probes consisting of microcentrifuge tubes (2 mL), parafilm wrapping, intravenous catheters (G18) containing ball resonator; (d) multiplexing set up consisting of splitter and optical fibers.

Figure 4

Microbiological methodology: Step 1—cultivate bacteria on Petri dish; step 2—inoculation of one pure colony in 3 tubes with NB; step 3—inoculation of 3 separate replicants for each experiment in NB; step 4—dilution to targeted OD for experiment.

Figure 5

Spectral response of ball resonators in different bacteria concentrations; the charts report the response of three different sensors with sensitivity values −127.52, −144.28 and −104.46 dB/RIU at different values of OD recorded over 60 min period. (a–c) Reflection spectra of each sensor for different measurement times for OD equal to 0.05 (a), 0.1 (b), and 0.5 (c). Insets of each figure show the spectral valley used for the intensity level tracking. (d) Normalized response for each sensor, reporting the amplitude change over time; the response of each OD is separated into 3 regions: 0–10 min, 10–30 min, 30–60 min.

Figure 6

Time response of ball resonators in different bacteria concentrations. (a) Normalized response during the 0–60 min time interval for each value of OD (0.05, 0.1, 0.5) and controls (DC, NB); charts report the average (solid line) and ±standard deviation (shadowed region) of 3 different sensors. (b) Bar chart reporting the temporal response in each condition, at each time stamp (bar = mean; error bar = ±standard deviation). (c) Integrated response, showing the integral of the average response at each time. (d) Response rate, estimating the slope of each normalized response over three time intervals (0–10, 10–30, 30–60 min); bar = mean; error bar = ±standard deviation.

Figure 7

CV binding assay for 1 and 2 h of bacteria incubation.