Photoacoustic imaging of posterior periodontal pocket using a commercial hockey-stick transducer

Lei Fu, Ling Chen, Zhicheng Jin, Jessica Luo, Jorge Palma‐Chavez, Zhuohong Wu, Jingcheng Zhou, Jiajing Zhou, Brian Donovan, Baiyan Qi, Aditya Mishra, Tengyu He, Jesse V. JokerstJournal of Biomedical Optics 2022 · DOI

TL;DR

This paper introduces a novel "hockey-stick" shaped photoacoustic transducer for imaging periodontal pockets, overcoming the limitations of larger conventional transducers that struggle with posterior teeth. The developed system demonstrates accurate and reproducible measurements of pocket depth, showing promise for improved diagnosis and monitoring of periodontitis.

What Problem Does This Paper Address?

Periodontitis, a common inflammatory disease affecting the tissues supporting teeth, can lead to bone loss and tooth loss if untreated. Accurately measuring the depth of periodontal pockets is crucial for diagnosis, treatment monitoring, and assessing disease progression. Current methods, like manual probing with a Williams probe, are subjective, prone to errors due to probing force and angulation, and can be painful for patients. While photoacoustic imaging (PAI) offers a more objective and potentially more accurate alternative for measuring pocket depth, conventional PAI transducers are too large to effectively image the posterior teeth (pre-molars and molars), which are often affected by periodontitis. This limits the clinical applicability of PAI for comprehensive periodontal assessment.

Methods

The core innovation of this work is the adaptation of a commercially available "hockey-stick" shaped ultrasound transducer for photoacoustic imaging. This transducer, with a central frequency of 9 MHz and a bandwidth of 7-15 MHz, was modified by integrating a 14-fiber bundle to deliver laser light for photoacoustic excitation. The unique angled design of the hockey-stick transducer is key, allowing it to access the more posterior regions of the mouth that are difficult for standard linear transducers.

The system was characterized using a pencil lead array to evaluate resolution and imaging range. For ex vivo validation, swine jaws were used, with artificially created deep pockets to mimic human conditions. Cuttlefish ink was employed as a photoacoustic contrast agent to highlight the periodontal pocket. Pocket depth measurements were performed using both the developed PAI system and the traditional Williams probe, with blinded comparisons to assess accuracy and correlation. Inter-rater reliability was also evaluated to assess the reproducibility of the measurements. Finally, the system was tested in vivo on a healthy human subject.

Datasets

The study utilized:

  • Swine jaws: Used for ex vivo validation of the photoacoustic imaging system's ability to measure periodontal pocket depths. Artificial pockets were created to simulate human pocket depths.
  • Human subject: A single healthy 29-year-old adult was imaged in vivo to demonstrate the system's applicability in a clinical setting.

No new benchmark datasets were introduced; rather, the study focused on validating the performance of the developed PAI system against a gold standard (Williams probe) and assessing its reproducibility.

Key Insights and Ideas

  • The "Hockey-Stick" Advantage: The angled, narrow profile of the hockey-stick transducer is a game-changer for accessing posterior teeth, a significant limitation of previous PAI systems in dentistry.
  • PAI vs. Probing: Photoacoustic imaging, when coupled with a contrast agent, provides a more objective and potentially more precise measurement of periodontal pocket depth compared to manual probing.
  • Reproducibility is Key: The study demonstrates good inter-rater reliability, suggesting that the PAI method is teachable and can be reliably used by different operators.
  • Bridging the Gap: The system successfully bridges the gap between the need for PAI in posterior regions and the limitations of existing transducer technology.

Key Results

  • Resolution: The axial and lateral resolutions for the photoacoustic imaging were measured at 230 µm and 258 µm, respectively.
  • Accuracy in Swine Model: The photoacoustic measurements showed a bias of approximately -0.3 mm compared to the Williams probing method in swine teeth, with 95% of replicates falling within 3.2 mm of the difference.
  • Inter-Rater Reliability: Inter-class correlation (ICC) values for agreement between raters ranged from 0.61 (good) to 0.90 (excellent), indicating good to excellent reproducibility.
  • Human Subject Imaging: The system successfully imaged incisor #8 and pre-molars #5, #12, and #29 in a healthy human subject, demonstrating its in vivo capability. Measured pocket depths in the human subject were small (0.3-0.9 mm), consistent with good oral hygiene.

Limitations and Future Work

  • Coverage Limitations: The current hockey-stick transducer, even with its improved design, cannot image the very last molars due to its size and the required illumination angle for light coupling.
  • Transducer Size: The bulk added by the fiber module is a limitation.
  • Future Directions: The authors envision developing even smaller, high-frequency transducers specifically designed for dental applications to cover all teeth. They also suggest exploring alternative contrast agents and further optimizing the system for clinical use.

Summary updated

Abstract

The unique angle shape of the hockey-stick transducer allows it to image more posterior teeth than regular linear transducers. This study demonstrated the ability of a hockey-stick transducer to measure the periodontal pocket via photoacoustic imaging.
BibTeX
@article{fu2022photoacoustic,
  author = {Lei Fu and Ling Chen and Zhicheng Jin and Jessica Luo and Jorge Palma‐Chavez and Zhuohong Wu and Jingcheng Zhou and Jiajing Zhou and Brian Donovan and Baiyan Qi and Aditya Mishra and Tengyu He and Jesse V. Jokerst},
  title = {{Photoacoustic imaging of posterior periodontal pocket using a commercial hockey-stick transducer}},
  journal = {Journal of Biomedical Optics},
  year = {2022},
  doi = {10.1117/1.jbo.27.5.056005},
}
Figure