Guide mapping for effective superficial photothermal coagulation of the esophagus using computer simulations with ex vivo sheep model validation study

Abstract

Objectives The transfer and widespread acceptance of laser‐induced thermal therapy into gastroenterology remain a topic of interest. However, a practical approach to the quantitative effect of photothermal injury in the esophagus needs further investigation. Here, we aim to perform computer simulations that simulate laser scanning and calculate the laser‐induced thermal damage area. The simulation engine offers the results in a guide map for laser coagulation with a well‐confined therapeutic area according to laser irradiance and surface scanning speed. The study also presents validation experiments that include histology analyses in an ex vivo sheep esophagus model. Methods The simulation engine was developed based on the Monte‐Carlo method and the Arrhenius damage integral. The computational model mimicked laser scanning by shifting the position of the calculated heat source in the grating system along the axis to be scanned. The performance of the simulations was tested in an ex vivo sheep esophagus model at a laser wavelength of 1505 nm. Histological analysis, hematoxylin–eosin staining, light microscope imaging, and block‐face scanning electron microscopy were used to assess thermal damage to the tissue model. Results The developed simulation engine estimated the photothermal coagulation area for a surface scanning speed range of 0.5–8 mm/second and laser power of up to 0.5 W at a 0.9‐nm laser diameter in a tissue model with a volume of 4 × 4 × 4 mm3. For example, the optimum laser irradiation for effective photothermal coagulation in the mucosa and superficial submucosa depths was estimated to be between 16.4 and 31.8 W/cm2, 23.2 and 38.1 W/cm2 at 0.5 and 1 mm/second, respectively. The computational results, summarized as a guide map, were directly compared with the results of ex vivo tissue experiments. In addition, it was pointed out that the comparative theoretical and experimental data overlap significantly in terms of energy density. Conclusions Our results suggest that the developed simulation approach could be a seed algorithm for further preclinical and clinical trials and a complementary tool to the laser‐induced photothermal coagulation technique for superficial treatments in the gastrointestinal tract. In future preclinical studies, it is thought that the simulation engine can be enriched by combining it with an in vivo model for different laser wavelengths.

An endoscopic approach providing near‐infrared laser‐induced coagulation with accurate depth limits

“…First, we successfully simulated the approach that traps the therapeutic effect in shallow tissue layers by creating a heat well. The laser‐induced thermal effect of the endoscopy apparatus was modeled using a simulation engine we previously developed [16] based on the Monte Carlo method [17]. The engine determined the distribution of photons in the tissue as a function of wavelength‐dependent parameters (i.e., refractive index, absorption coefficient, scattering coefficient, and anisotropy factor).…”

“…In our previous study [16], block face scanning electron microscopy examinations were utilized as the basis for determining the difference between coagulated and uncoagulated esophageal tissue in optical microscopy images of H&E-stained samples. Thus, thermal damage was characterized according to a hue and tone system in BF and DF microscope images.…”

Predicting dark-field images of H and E-stained esophageal specimens

“…Datasets were created from the microscopy images produced in another study 4 , in which a large matrix of parameters such as laser power, surface scanning speed, beam diameter, and irradiation duration was optimized. The dataset's images are 28 H&E-stained bright-field and dark-field microscopy images of ex vivo sheep esophageal tissue samples used in laser-induced thermal injury.…”

“…In particular, the entire work can be considered labor-intensive and time-consuming, given the number of evaluations required to optimize the large parameter matrix. Also, different imaging (e.g., dark-field microscopy) performed alongside bright-field microscopy has a multiplier effect on the workload, although it offers new opportunities to analyze more specific thermal damage 4 . Therefore, automated digital generation of microscopic images 5 could lead to exciting work in this research field.…”