Intraoperative Speckle Variance Optical Coherence Tomography for Tissue Temperature Monitoring During Cutaneous Laser Therapy

Shoujing Guo, Shuwen Wei, Soohyun Lee, Mary Sheu, Se‐Won Kang, Jin U. KangIEEE Journal of Translational Engineering in Health and Medicine 2019 · DOI

TL;DR

This paper introduces a novel, non-invasive method using speckle variance optical coherence tomography (svOCT) to monitor tissue temperature in real-time during cutaneous laser therapy. By establishing a linear relationship between speckle variance and temperature before tissue coagulation, svOCT can accurately track rapid temperature changes, leading to safer and more effective laser treatments.

What Problem Does This Paper Address?

Cutaneous laser therapy is a widely used treatment for various skin conditions. However, ensuring treatment efficacy while avoiding collateral damage requires precise control over tissue temperature. Current methods for monitoring temperature during these procedures are often invasive (like thermocouples) or lack the speed and accuracy needed for real-time feedback, especially with pulsed lasers. This leaves treatment parameter selection largely reliant on operator experience, which can vary significantly. The paper aims to bridge this gap by developing a non-invasive, high-speed, and accurate method for intraoperative tissue temperature monitoring.

Methods

The core of this research is the application of speckle variance optical coherence tomography (svOCT) for temperature monitoring. svOCT leverages the fact that as tissue heats up, molecular motion increases, altering the speckle patterns captured by OCT. By analyzing the variance of these speckle patterns across multiple OCT frames, researchers can infer temperature changes.

The authors developed a system combining a swept-source OCT with a pulsed laser. To calibrate their svOCT system, they used excised human skin samples heated incrementally in a water bath. They then performed linear regression to establish a relationship between normalized speckle variance and temperature, specifically identifying a threshold of 52°C, beyond which tissue coagulation occurs and the linear relationship breaks down.

For validation, they conducted experiments where the skin samples were irradiated with a pulsed laser while svOCT monitored temperature. Additionally, they employed COMSOL Multiphysics software to create a numerical thermal model of the laser-tissue interaction, allowing for comparison between simulated and experimental temperature profiles.

Datasets

The study utilized excised human skin samples from three individuals, collected from the arm, face, and back. These samples were used for both the temperature calibration phase and the laser therapy experiments. The "dog-ear" samples, typically redundant skin from excisional surgery, were chosen for their composition of normal, full-thickness skin. No new benchmark datasets were introduced; rather, the focus was on characterizing the behavior of biological tissue under controlled laser irradiation.

Key Insights and Ideas

  • Speckle Variance as a Thermometer: The most significant insight is the demonstration that speckle variance in OCT images directly correlates with tissue temperature, acting as a sensitive, non-invasive thermometer.
  • Pre-Coagulation Linearity: A crucial finding is the identification of a clear, linear relationship between normalized speckle variance and tissue temperature before the onset of tissue coagulation (around 52°C). This linearity is key for accurate temperature quantification.
  • Coagulation as a Distinct Event: The study clearly shows that tissue coagulation at temperatures above 52°C leads to a significant and rapid decrease in normalized speckle variance, marking a distinct thermal event that breaks the linear trend.
  • Cross-Validation with Thermal Modeling: The agreement between experimental svOCT measurements and numerical thermal simulations provides strong validation for both the svOCT method and the thermal model.
  • Potential for Real-time Feedback: The high temporal resolution (10 ms) and depth resolution (4.5 µm) of the svOCT system suggest its suitability for real-time feedback during laser procedures.

Key Results

  • A linear relationship was established between normalized speckle variance and tissue temperature for excised human skin samples in the range of 25°C to 52°C, with high correlation coefficients (e.g., r = 0.9007 for the back sample).
  • The svOCT system successfully monitored rapid temperature increases during pulsed laser irradiation, reaching up to approximately 51.7°C in the irradiated area.
  • Tissue coagulation was observed to occur around 52°C, leading to a significant decrease in normalized speckle variance.
  • The experimental results showed good agreement with numerical simulations, with a cross-correlation coefficient of r = 0.88 for temperature trends.
  • The system demonstrated repeatability across different skin sample sites (arm, face, back) and within different regions of interest on the same sample, suggesting a universal calibration might be possible for similar skin types.

Limitations and Future Work

  • Coagulation Limit: The primary limitation acknowledged is that the linear relationship breaks down after tissue coagulation, making it difficult to monitor temperature accurately once this event occurs. This also means calibration cannot be performed on the same spot after coagulation.
  • Ex-Vivo vs. In-Vivo: The study was conducted on ex-vivo skin samples. While efforts were made to estimate water content and thermal properties, in-vivo conditions (e.g., blood perfusion, dynamic tissue response) might differ.
  • Cooling Rate Discrepancy: The experimental cooling rate was observed to be faster than predicted by the simulation, potentially due to factors like non-uniform tissue surfaces and water evaporation.
  • Future Work: The authors propose expanding test datasets with more diverse human skin samples, performing calibration and monitoring during clinical laser therapy, and investigating methods to improve the sensitivity and resolution of the svOCT system. They also aim to explore how svOCT can be used as a research tool to further understand laser-tissue interactions.

Summary updated

Abstract

<i>Background:</i> Tissue temperature monitoring during cutaneous laser therapy can lead to safer and more effective treatments. In this study, we investigate the use of speckle variance optical coherence tomography (svOCT) to monitor real-time temperature changes in the excised human skin tissue sample during laser irradiation. <i>Methods:</i> To accomplish this, we combined the pulse laser system with a reference-based svOCT system. To calibrate the svOCT, the ex-vivo human skin samples from three individuals with tissues collected from the arm, face, and back were heated with 1-degree increments. Additionally, linear regression was used to extract and evaluate the linear relationship between the temperature and normalized speckle variance value. Experiments were conducted on excised human skin sample to monitor the temperature change during laser therapy with a svOCT system. Thermal modeling of ex-vivo human skin was used to numerically simulate the laser-tissue interaction and estimate the thermal diffusion and peak temperature of the tissue during the laser treatment. <i>Results and Conclusion:</i> These results showed that normalized speckle variance had a linear relationship with the tissue temperature before the onset of tissue coagulation (52°) and we were able to measure the rapid increase of the tissue temperature during laser therapy. The result of the experiment is also in good agreement with the numerical simulation result that estimated the laser-induced peak temperature and thermal relaxation time.
BibTeX
@article{guo2019intraoperative,
  author = {Shoujing Guo and Shuwen Wei and Soohyun Lee and Mary Sheu and Se‐Won Kang and Jin U. Kang},
  title = {{Intraoperative Speckle Variance Optical Coherence Tomography for Tissue Temperature Monitoring During Cutaneous Laser Therapy}},
  journal = {IEEE Journal of Translational Engineering in Health and Medicine},
  year = {2019},
  doi = {10.1109/jtehm.2019.2943317},
}
Figure