Table 13. Comparison of application, sample type, and accuracy/limit of detection (LOD) for various optical modalities.

3. Future Prospective and Current Clinical Limitations

Advancements in the design of new optical systems have demonstrated the immense potential for various optical spectroscopy, imaging, and sensing tools to assess pathogens within biological samples. These techniques offer rapid, and in some cases, culture-free approaches to detect and characterize bacterial species with simplistic measurement protocols.

When aiming to study and characterize bacteria biofilms in their native environments, as in the case of OM, investigators can seek to use portable OCT systems to examine biofilm structures and, where applicable, types of bacterial effusions. Advances in its in vivo applications now involve using machine learning to automate OM classification based on the images acquired to better aid medical personnel ease in interpreting the OCT images ^[98][178]. Beyond OM utility, researchers are taking advantage of OCT imaging capabilities for other in vivo biofilm assessment relating to rhinitis, dental plaques, and antibiotic therapy monitoring^[104] [204].

If the goal is to detect and identify bacteria species, strain, or study antibiotic response, Raman spectroscopy and Fourier-transform infrared (FTIR) spectroscopy are the best tools to use—especially if the species is unknown. However, these modalities are best used when coupled with known databases and machine learning algorithms to improve specificity. Moving towards culture-free detection and in cases where culture-negative bacteria are being investigated, SERS-based techniques can prove beneficial and may be best suited for this application. The significant amplification of the intrinsic Raman signal that the nanoparticles provide is advantageous for culture-free methods. In recent years, nanoparticles have been coupled at the tip of fiber optical probes to create a more reproducible means of acquiring the SERS signal^[105] [205]. Therefore, with improved sensitivity over conventional Raman optical probes, this technique can provide researchers further flexibility in its use, depending on the ex vivo and in vivo environment. One drawback that researchers need to be aware of is that the SERS-based optical probes and label-free assays need to be in direct contact with the sample. Further research needs to be conducted to assess the sensitivity and specificity of the optical probes in identifying a particular species without labels. If the bacterial species is known, fluorescence, SERS, and interference-based platforms can be used for direct targeting using antigen-specific binding. Of the three, fluorescence spectroscopy is the simplest and most inexpensive tool to use. It is also the most versatile because it can be used for both imaging and quantitative analysis.

To assess bacterial growth kinetics, laser speckle contrast imaging and OCT are the most useful tools. However, with simpler optics, laser speckle is more advantageous and provides faster information, especially in measuring antimicrobial susceptibility. Compared to the gold standard and commercial antimicrobial susceptibility test (AST), which can take anywhere between 8 and 20 h for standard methods, laser speckle contrast can accomplish this within two-to-five hours^[106] [202]. If interested in colony growth patterns, polarization, and laser-scattering offer users the ability to capture growth patterns over time and utilize those patterns to understand how the cells divide and self-organize into colonies.

Towards point-of-care diagnostics is the miniaturization of fluorescence and SERS-based systems into a mobile or handheld platform for field applications. The simplicity in the optics required for fluorescent-based spectroscopy and imaging allows it to be fashioned into compact handheld devices that can be coupled with microfluidic platforms to be used at the point-of-care^{[29][27][107]} [164,166,206]. Similarly, with recent advancements in the design of SERS-based probes and compact spectrophotometers, these systems can couple with both paper- and microfluidic-based tools for point-of-care analysis. However, SERS-based systems are more expensive and have a higher degree of complexity than fluorescent ones, but they do provide the possibility for multiplexed detection.

Bacteria identification and the visualization of the distribution within a localized space can be achieved using various multimodal optical systems. These systems involve the merger of two or more optical modalities to benefit from the individual systems and provide users with more comprehensive information for bacteria detection and characterization. For example, coupling the biomolecular sensitivity of Raman with OCT, which provides depth-resolved imaging of a bacterial environment, can allow users to interrogate bacteria within a localized space^[101] [181]. Another multimodal system involves the coupling of autofluorescence with OCT for dental biofilm assessment. The fluorescence imaging provides information about the maturity of pathogenic plaque and OCT about the total plaque buildup^[108] [207].

However, while these optical modalities have shown tremendous promise in detecting bacteria, there are still several challenges associated with these techniques that must be addressed. The broader challenge centers around developing analytical software for quick and reliable data processing and real-time interpretation. For some modalities, such as FTIR, this is already being addressed with the creation of large databases containing information on various biological samples that are open access and can be shared and used as standards by which unknown data can be interpreted. Such databases are in their early stages, and unanimous consensus among scholars in each field is needed to ensure the validity of the information.

Another major challenge is the cost and expertise required to construct the various technologies. For fluorescence devices, this is less of an issue, as discussed. The recent acceptance and use of Raman spectroscopy in forensic science and pharmacology have allowed for the commercialization of various RS tools from benchtop to optical probes and handheld devices. This technique still relies on the use of sensitive detection hardware and narrowband laser sources, which are relatively expensive components. Still, the reduction in prices of these tools is inevitable as the market for these kinds of systems matures. In addition, as more clinicians become aware of the recent FDA-cleared OCT handheld device for OM detection, this technology also has the potential to become more widely available for the common use of biofilm characterization. Research studies with other technologies such as laser speckle contrast, inference, and polarization are still in nascent stages. Therefore, these systems are not readily available, as most are designed and built in-house.

The optical detection of pathogenic bacteria is currently an area of tremendous ongoing research because different modalities have the potential to provide information about the biomolecular make-up of a species, growth pattern recognition, single-cell versus biofilm characterization, cell motility and viability, cell mutation, and antibiotic resistance status. However, further research must be pursued to enable the eventual use of these modalities by clinicians.