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

This article investigates an endoscopic approach that utilizes negative pressure to achieve laser‐induced thermal coagulation limited to the esophageal wall's mucosal and superficial submucosal layers. The study was built upon a series of studies combining numerical simulation based on the Monte‐Carlo technique and ex vivo porcine tissue experiments, including apparatus design and histology analysis. An endoscopy apparatus was developed using 3D printing to validate the tissue stretching‐based approach. A fiber‐pigtailed diode was used as the near‐infrared source, emitting 208.8 W/cm2 laser irradiance at 1.5 μm. Simulation results suggested that the approach successfully created a local heat well to prevent residual thermal effects (>65°C) from penetrating the deeper submucosal layer. Histology analysis of ex vivo tissues showed that at a fluence of 5.22 kJ/cm2, the depth of thermal coagulation was reduced by half compared to the control. With further preclinical studies, including endoscopy apparatus design, the approach can be applied to the larger esophageal surface.


| INTRODUCTION
Therapeutic endoscopy is a versatile intervention used in the outpatient and inpatient treatment of the upper gastrointestinal tract.The procedure allows for various treatments, including hemostasis of ulcer bleeding, esophageal dilatation, removal of gastric polyps, and stenting [1].It also stands out as a platform that enables the development of new approaches in the treatment of precancerous tissues [2].For example, the endoscopic thermal therapy (ETT) method is widely used to diminish or eliminate Barrett's esophagus (BE) [3][4][5], a precancerous condition of the esophagus and associated with the risk of developing esophageal adenocarcinoma [6].This method is based on heating the target mucosal layer with absorbed energy.Direct contact transmission of high-frequency alternating (<500 kHz) current, the non-contact transmission of highfrequency energy through the flow of argon gas, and transmission of near-infrared laser radiation in the mm-scale distance between the probe tip and the tissue surface are prominent techniques for exposing the tissue to heat energy.Although ETT provides improvements that contribute to BE reversal, the main challenge with current energy delivery strategies is that the treatment depth extends more profoundly than the submucosal layer (>1 mm), causing an uncontrollable thermal effect in the non-target tissue.In addition, more than one treatment session may cause adverse conditions such as organ narrowing, bleeding, or perforation with this treatment failure [7,8].On the other hand, ETT combined with practical approaches that accurately apply the thermal effect confined to the mucosa (<0.8 mm thickness [9,10]) can address the challenges in the treatment of abnormal tissue at skin depth.Furthermore, the availability of endoscopic procedures with such approaches is of substantial interest for studies of therapeutic endoscopy technology [11].
We present an endoscopic approach that limits the depth of thermal damage by exploiting negative pressure stretching of the mucosal and superficial submucosal layers.Conceptual validation of the proposed approach in an ex vivo porcine esophagus and stomach models was performed using the laser-induced thermal coagulation method.In addition to providing well-known mathematical models of laser-tissue interaction in theoretical analysis, the laser method using fiber-coupled diode technologies offered high controls in laser energy delivery.In this context, an endoscopy apparatus developed in line with ex vivo experiments delivered laser radiation at a wavelength of 1.5-μm while vacuum force lifted the mucosal lining a finite distance from deeper layers (e.g., muscularis externa).Numerical analysis based on Monte Carlo simulations revealed the potential of the proposed approach to limit the thermal effect (i.e., 65 C and above) within the recessed area (slot) of the apparatus.In addition to photothermal coagulation up to 750 μm deep, the porcine esophagus model results suggested that the apparatus slot defines the area of thermal injury, thereby providing photothermal damage that is relatively independent of the laser energy density (i.e., fluence) applied in the proposed approach.

| Design of the endoscopic apparatus
We designed an endoscopy apparatus to validate the proposed endoscopic approach in the porcine esophagus.
Figure 1A shows schematic illustrations of the overall outline of the endoscopy apparatus (SOLIDWORKS, education version).The apparatus is 3 cm long and 1 cm wide, fitting the pig model [12,13], and consists of two parts: a cover and a body.A large hole in the body (diameter: 6.5 mm) is the air duct that transmits the negative pressure created by the vacuum pump.The heat well slot is 4.8 mm long and 1.5 mm wide.The width is chosen according to a value (approximately twice the pig mucosa layer thickness) that will not cause mechanical damage to the stretched mucosa and superficial submucosal layers [14].A GRIN lens is coupled with the diode laser and transmits a collimated laser beam at a wavelength of 1.5 μm.For soft tissues containing >70% water, water is considered the primary chromophore in the near-IR.The wavelength of 1.5 μm provides an optical penetration depth of >0.57mm in water [15], matching the width of the slot design.

| Working principle of the approach
The proposed endoscopic approach creates a heat well that traps the temperature increase in the recessed area despite heat conduction.Additionally, this approach, taking advantage of the elasticity of soft tissues, uses F I G U R E 1 (A) Schematic illustration of the overall outline of the endoscopy apparatus.(B) A simulation result of transient temperature to ensure that the approach results in the confinement of the therapeutic effect to the recessed area.Beam diameter = 0.5 mm, grid size = 20 μm, laser power = 400 mW for 1 second laser duration (corresponding to 203.7 J/cm 2 ).(C) A crosssectional image of a hematoxylin-and-eosin-stained porcine esophagus tissue sample to confirm the delimitation of mucosal and superficial submucosal layers from deeper layers via the negative pressure approach-the images show the merging of seven images acquired at 10Â magnification (0.25 NA).An enlarged view highlights the area of interest.
negative pressure to stretch the mucosal and superficial submucosal layers into a recessed area where a heat well is created.Within this context, two studies were planned to exemplify the working principle of the approach.
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 wavelengthdependent parameters (i.e., refractive index, absorption coefficient, scattering coefficient, and anisotropy factor).The calculated photon distribution profile was used as a heat source for the time-dependent calculation of the thermal response based on a partial differential equation governed by the thermal properties of the tissue model and tissue size (stretched and surrounding tissue).The optical and thermal properties of the tissue model used in the simulation were compiled from the literature [18][19][20].A transient temperature result of the simulation for 203.7 J/cm 2 (Ø0.5 mm, 400 mW for 1 s laser duration) is shown in Figure 1B.As in the design of the apparatus, the stretched tissue size was set to 1.5 mm.The result confirmed that the tissue in the recessed area reached a temperature sufficient for thermal damage, defined as photothermal coagulation at 65 C and above, even with 1 s of irradiation.Additionally, the simulation suggested that the approach traps heat rather than heat flux to surrounding and deep tissues.It has been shown that tissue temperature does not reach >40 C at a depth of >1 mm despite thermal conductivity.
The second study aimed to histologically prove the concept of confining esophageal mucosal and superficial submucosal layers to a specific volume by vacuum aspiration.Laser-induced thermal damage (coagulation) was not performed at this stage.In this context, the apparatus inserted into the porcine esophagus, ex vivo, was immersed in 10% neutral formalin during vacuum aspiration (i.e., applying negative pressure in the recessed area of the apparatus).The tissue was thus fixed in its current state (i.e., the state in which it was stretched into the recessed area) and then embedded in paraffin to prepare for sectioning.The pressure was measured as À295 mmHg (corresponding to À39.3 kPa) within the safe range against mechanical damage to soft tissue [21,22].Figure 1C shows a cross-section histological image of a tissue section (5 μm) stained with hematoxylin-and-eosin (H&E-stained) acquired at 10Â magnification by brightfield (BF) microscopy.It was observed that tissue protrusion occurred based on tissue stretching due to vacuum aspiration.As can be seen from the image, the physical form of the protrusion on the mucosal surface was defined by the slot of the apparatus.The histological study confirmed the delimitation of mucosal and superficial submucosal layers of the ex vivo porcine esophagus.The results also showed that the vacuum force could stretch target tissue structures to a limited distance from deeper layers, such as the muscularis externa (muscle), and trap it in the recessed area of the apparatus.This straightforward approach was an essential capability of the apparatus to create a wellconfined depth of therapy during the therapeutic process.

| Experimental setup
Figure 2A shows a schematic illustration of the experimental setup combining all its components for wellconfined laser-induced thermal damage studies in the porcine tissue model, ex vivo.Components included an electric vacuum pump with a vacuum gauge connected in series, an endoscopy apparatus, and a diode laser for thermal coagulation.Figure 2B depicts a photograph of the apparatus that uses the negative pressure approach with a slot to stretch the mucosal and superficial submucosal tissue layers into the recessed area.The parts were printed with a 3D printer from biocompatible polylactic acid material.The visible (635 nm) aiming beam indicated that the invisible laser beam was aligned towards the recessed therapy area for laser-induced thermal injury.
A fiber pigtailed (SMF-28e+) diode laser (BrightLock, QPC Lasers) emitting up to 500 mW at 1505 nm was used as a near-infrared source.The laser provided an optical penetration depth of >0.57mm in water within the therapy depth defined by the housing of the apparatus (recessed therapy area), as shown in Figure 3A, and was operated in CW mode to achieve the fastest deposition of the total energy density required.A fiber-coupled (SMF-28e+) GRIN lens was fixed in place to collimate a laser beam.The optical fiber passed through vacuumsealed tubes, and airtightness was completed using sealing tape to generate a sufficient vacuum in the recessed therapy area.
Figure 3B shows the measured laser irradiance generated as a function of the electric current applied to the diode laser chip (100-mA steps up to an electric current of 4 A).The green area shows the irradiance range (140-210 W/cm 2 ) at which the endoscopy apparatus provides a temperature rise of over 65 C. The asterisk marks 208.8 W/cm 2 used in all ex vivo experiments.Figure 3C shows the beam radius calculated as approximately 0.5 mm using the knife-edge measurement technique at different knife-edge positions.
A digital camera (AM4815ZTL, Dino-lite) and a thermal camera (0.93 emissivity and 80 mK thermal sensitivity, Xi400, Optris) were also used in the experimental setup.While the digital camera facilitated the visual monitoring of the tissue, the thermal camera measured the temperature change of the tissue from the muscle surface (i.e., the longitudinal muscle layer of the muscularis propria).Thus, the thermal effect in the deep tissues was visualized from the outer surface.

| Ex vivo tissue studies
In the early validation phase of the approach, we planned ex vivo tissue studies as a guide, albeit limited, and thus also considered the replacement principle of the 3Rs (i.e., replace, reduce, and refine).Fresh tissue was moistened with saline solution during the experiments to prevent desiccation.Therefore, we assumed that the percentage of water did not change significantly, and that the absorption coefficient of the tissue model was as constant as possible.We conducted all studies in a fume hood (Köttermann), which provides a biologically safe environment and is approved by the Izmir Biomedicine and Genome Center Biosafety Board.
We first examined the tissue stretch-based approach in the porcine esophagus, which resembles the human esophagus in terms of morphology and layer thickness [12].The following steps were followed in this part of the study: (1) The apparatus was inserted in a portion of the fresh pig esophagus obtained from the slaughterhouse.Before the procedure, the esophagus was cleaned with a saline solution.Figure 4A shows an example of the apparatus (developed for this purpose) inserted into the esophagus during the experiment.(2) The position of the tissue was adjusted so that the slot of the inserted apparatus was perpendicular to the field of view of the thermal camera.(3) The vacuum pump was started for suction aspiration (vacuum).The negative pressure was set to À295 mmHg, as determined in the previous section.(4) The 1.5 μm diode laser was operated in CW mode to induce thermal damage to the tissue stretched into the recessed area.(5) The laser and vacuum pump were turned off, respectively, and the apparatus was shifted to another spot along the esophageal surface for statistical analysis, and Steps 2-4 were repeated.The laser irradiance for all trials was 208.8 W/cm 2 .The laser duration ranged from 5 to 25 s.The corresponding laser fluence was between 1044 and 5220 J/cm 2 .
The stretching of the tissue is based on using the unique biomechanical property of the esophageal tissue layers, namely their elasticity [23].An essential advantage of this approach is that it works for all soft tissue types.However, it is conceivable that tissues such as columnar epithelium could potentially differ from stratified squamous epithelium in terms of elasticity.Consequently, such a difference in biomechanical properties may raise a question regarding the approach's potential, such as the inability of the apparatus to stretch the tissue into the recessed area/slot.Finally, the tissue stretch-based approach was applied to partially address this issue in the ex vivo porcine stomach model because the mucosa of the fundus region of the stomach consists of cardiac glands and is covered with simple columnar epithelium [24,25].Figure 4C shows an example of laser coagulation while operating the vacuum pump after the apparatus contacts the stomach surface during the experiment.

| Histology and microscopy
Immediately after laser coagulation studies, tissue samples ($5 Â 5 mm 2 square pieces) were collected from the mucosal whitening part seen in gross anatomy to measure the depth of thermal damage.The collected samples were fixed in 10% neutral formalin to preserve their current state and embedded in paraffin to prepare for the sectioning process.Next, the relevant region was sectioned into longitudinal sections of 5 μm thickness from paraffin blocks to be stained with hematoxylin-eosin (H&E) for further microscopy examination.Brightfield (BF) and darkfield (DF) microscope images were obtained from the stained tissue samples.
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.The published results revealed that, in addition to the information obtained from BF and DF microscope images, mathematical color subtraction of DF microscopy images from BF images (e.g., magenta-red = blue) provides a high-contrast image with morphological details.The color subtraction approach identified uncoagulated tissue fragments as pink in the epithelium, yellow-orange in the lamina propria and submucosa, and red in the muscularis mucosa, and coagulated tissue fragments as pale pink and yellow in the epithelium, pink in the lamina propria, and magenta-pink in the muscularis mucosa.Microscopy images produced to evaluate the response of esophageal tissue to laser-induced coagulation with the apparatus and to laser-induced coagulation without the apparatus were examined using the same image analysis.This image processing also contributed to the differentiation of individual tissue layers and structural details to analyze thermal tissue damage and estimate the depth, width, and area of injury by pixel calculation.

| Statistical analysis
A total of eight pig esophagus and three pig stomach trials were performed in this study.Three esophagi were used as controls in the no-apparatus condition.Iterations of laser coagulation ranging from a minimum of three to a maximum of five were performed for each laser power set.To assess statistical significance, we used ordinary one-way analysis of variance (ANOVA) for multiple groups of laser energy density (laser fluence) and unpaired t-tests for datasets with and without apparatus at each laser energy density.A p-value less than 0.05 was typically statistically significant in the analysis.

| Comparison results in an esophageal model
In this part of the study, the performance of an apparatus capable of implementing the proposed endoscopic approach was evaluated in an ex vivo porcine esophagus model.The evaluation included comparing the results of the esophagus directly exposed to the laser beam without the apparatus, as in conventional laser therapy.Laser output power and beam diameter were kept constant in both cases, so laser fluence was the only variable dependent on irradiation time.thermally induced visual changes during laser-induced thermal injury.
Figure 6A,D shows an example of these optically detectable changes (i.e., mucosal whitening) for 1.044, 1.67, 2.088, 2.715, 3.132, 3.759, 4.176, and 5.22 kJ/cm 2 , respectively.Note that the whitening size increases as a function of laser fluence.In addition to observing whitening, samples were collected at the end of each trial for light microscopy examination to assess the response of biological tissues to thermal damage.Figure 6B,E shows representative cross-sectional histological images of H&E-stained tissue acquired at 4Â magnification in BF microscopy.The examination identified morphological changes in photothermal coagulation assessment based on hue.For example, the epithelium became lighter, while the connective tissue darkened if affected by photothermal coagulation.Disruption of intercellular junction integrity, formation of intracellular spaces, and loss of cell contours gave the epithelium a transparent blurry appearance, resulting in a light pink manifestation.The darker color of connective tissue was probably due to the high affinity of the free amino groups of the degraded proteins to eosin.On the other hand, explosive tissue disruption resulted from the relatively short working distance between the tip of the GRIN lens and the tissue and the rupture of water vapor vacuoles in heated tissues [26].Figure 6C,F shows samples and subtraction results of BF and DF microscopy images separately for cases with (dashed green lined box) and without (dashed blue lined box) endoscopy apparatus.Condensed or lost cell nuclei and intracellular voids were observed in the epithelial layer of the tissue.In addition, subepithelial clefting in the basal lamina and homogenization of collagen due to protein denaturation in the lamina propria (LP) were observed as thermal damage indicators.
Figure 7 summarizes all data estimated from the subtraction results of BF and DF microscopy images.As expected, the change in depth and width of laser-induced thermal damage was not statistically significant regardless of laser energy, as the slot of the apparatus defined the volume of the stretched tissue.The negative pressure also increased the tissue density by collecting the surrounding mucosa in the slot, thus ensuring the heat was trapped in the mucosal layer despite the heat conduction.Conversely, the depth and width of thermal damage increased as a function of laser energy at which the laser beam was collimated on the esophageal tissue without using the apparatus.Figure 7 in panel C comparing esophageal injury at the layer level shows that direct laser irradiation caused thermal damage to the submucosal layer by penetrating LP and muscularis mucosa (MM) layers at all laser energy values.In the control (no-apparatus case), the increased laser energy expanded the submucosa's damaged area and the esophagus's total damaged area.Conversely, the apparatus made the thermal effect independent of laser energy, thereby limiting thermal damage to the mucosa (LP and MM) and superficial submucosa.

| Ex vivo pig stomach studies
The tissue stretch-based approach allowed elastic soft tissues, such as esophageal tissue layers, to be pulled using vacuum force and held in the recessed area of the apparatus.In this section, we applied the apparatus to the ex vivo fundus region of the pig stomach model, which can closely mimic the abnormal, albeit limited, esophageal tissue type rather than the healthy esophageal layers.Similar laser energy density characteristics used in the comparison study were also reproduced with the exact irradiation durations for this study.Figure 8 shows histology analysis results obtained while operating the vacuum pump after the apparatus contacts the stomach surface during the experiment.The results matched the total damaged area in the healthy esophageal sample.For example, the apparatus produced a thermally damaged area of 0.430 mm 2 at 6.32 J (3.219 kJ/cm 2 ) laser energy, while the damaged area was measured as 1.054 mm 2 in the control (no-apparatus).At another laser energy level (7.54 J, corresponding to 3.84 kJ/cm 2 ), the damaged area was 0.432 mm 2 compared to 1.150 mm 2 without using the apparatus.This way, whether the columnar epithelial layer on its surface and the apparatus were working as planned was verified.

| DISCUSSION
Malignancies in the gastrointestinal tract may be associated with precancerous mucosal lesions [27].Adequate eradication of these precancerous lesions is critical in preventing gastric cancer progression, such as esophageal adenocarcinoma [28].For this reason, approaches that apply endoscopic thermal therapy in the early stages of abnormal tissues with minimal damage to healthy deep tissues continue to attract attention.Besides, the photothermal coagulation technique, whose potential has been regained with the developments in optical imaging, fiber optic components, and diode laser technology in the telecommunications wavelength range, can contribute to therapeutic technologies.
Our Monte-Carlo-based simulations predict that creating a heat well limits laser-induced thermal damage to the superficial tissue layer.In this context, we hypothesize an endoscopic approach based on stretching the mucosa towards a recessed area (i.e., heat well) by exploiting the flexibility of soft tissues.We present an apparatus design that supports this hypothesis.The apparatus created negative pressure (i.e., vacuum aspiration) for tissue stretching and delivered a 1.5 μm laser beam for thermal injury.The results demonstrated that the approach could confine laser-induced thermal damage to the mucosal and superficial submucosal layers in ex vivo models, including the pig stomach fundus region.Our findings also indicated that the extent of thermal damage became independent of laser energy as the tissue was trapped in a recessed area.Therefore, the temperature rises as a function of laser energy only began to determine the degree of thermal damage relative to the conservation of energy.
This study and its exciting findings reveal several limitations beyond its scope and should be discussed in future studies.First, the endoscopic apparatus developed to conceptualize the proposed approach produced local thermal damage incompatible with a typical clinical treatment surface size.Despite its restrictions, the apparatus yields adequate results for ex vivo studies, but it needs to be redesigned to be compatible with conventional endoscopes to support the clinical translation steps of the approach.It is also conceivable that the new design would have a scanner to deliver the laser energy to the sizeable esophageal surface.
Second, it has been assumed that the fundus region of the pig stomach, consisting of cardiac glands and is lined with simple columnar epithelium, mimics the abnormal esophageal tissue type, albeit limited, and the results obtained in the studies conducted on this model have F I G U R E 8 An example of histology results obtained at 6.32 J (corresponding to 3.219 kJ/cm 2 ) laser energy of the endoscopy apparatus placed in the fundus region of the pig stomach.Color subtraction results of brightfield (BF) and darkfield (DF) microscopy images are also applied.4Â magnification (0.13 NA).Scale bar = 0.5 mm.strengthened the applicability of the approach.However, the preclinical performance evaluation of the approach should be carefully evaluated in an animal model with an abnormal esophagus convenient for endoscopic intervention.In addition, a long-term study should be conducted to determine the extent of progressive chronic tissue damage.In fact, the apparatus design can be improved for RF ablation.
Third, the laser wavelength was 1.5 μm, corresponding to an optical penetration depth of >0.57mm in water [15], and the laser was operated in CW mode.On the other hand, various laser wavelengths such as 514, 532, 810, and 1064 nm were used in Barrett's esophagus treatment studies [29][30][31], providing a relatively deep optical penetration depth in water.Thus, the study can be extended by considering different laser wavelengths and pulsed laser modes.In addition, another study may be planned that involves modifying the apparatus design for RF ablation and comparing it to conventional RF ablation.
Finally, we prioritized the laser-induced thermal damage analysis of the study and did not focus on the adjective of mechanical damage detection other than the physical damage caused by photothermal ablation.However, a more rigorous histological analysis of thermal damage to the esophagus is required along with the in vivo animal model.Nitroblue tetrazolium chloride staining can contrast healthy and laser-induced thermally damaged tissue while excluding mechanical damage.

| CONCLUSION
This article investigates and conceptually validates a novel approach that exploits negative pressure stretching of the mucosal and superficial submucosal layers for well-confined laser coagulation.The tissue stretch-based approach provides a sufficient depth of thermal damage to the shallow layers of the target esophageal wall.The practical relevance of the approach was developed from an apparatus design compatible with an ex vivo model.Evolving from the results, it was concluded that this approach creates a heat well that traps the temperature rise in the recessed area despite thermal conduction.Moreover, it has been shown that the area of thermal damage to the tissue becomes independent of the applied laser energy.In other words, unintended tissue damage in terms of thermal spread is reduced with the extra benefit of less control of laser energy.The pig stomach model experimental results confirmed whether the apparatus worked as planned in the columnar epithelial layer on its surface.This demonstrates the relative potential of the approach's feasibility on abnormal tissue surfaces.With all merits mentioned above, the proposed approach can be enriched for coagulating a large esophageal mucosa with a novel endoscopy apparatus design in an in vivo model for further studies.
AUTHOR CONTRIBUTIONS Merve Turker-Burhan was involved in building the system, planning, and executing all experiments, and preparing the figures.Ender Berat Ellidokuz was involved in in designing and the planning of the ex vivo tissue model.Husnu Alper Bagriyanik was involved in the histology analysis.Serhat Tozburun was involved in obtaining support, conceptualization, investigation, writing-original draft, project management.
Schematic drawing of the experimental setup.(B) A photograph of the apparatus printed with a 3D printer from biocompatible polylactic acid.The slot is 4.8 mm long and 1.5 mm wide.The 635 nm aiming beam aligns the near-infrared laser toward the recessed area for thermal coagulation.

F
I G U R E 3 (A) A graph of the optical penetration depth in water over the optical wavelength.(B) Estimated laser irradiance as a function of electric current applied to the diode.The green area indicates sufficient range for coagulation.The asterisk marks 208.8 W/cm 2 used in the study.(C) Normalized laser powers along knife edge position and Gaussian-fitted first derivatives.The estimated beam diameter produced by the GRIN lens was $0.5 mm (1/e 2 ).

F
I G U R E 4 (A) Example showing the endoscopy apparatus in the ex vivo esophagus.(B) A photograph showing the control study for laser-induced coagulation without using the endoscopy apparatus.(C) A photograph showing the apparatus placed over the fundus region of the pig stomach that is covered with simple columnar epithelium.

Figure 5
presents representative thermal images of the esophagus acquired from the muscle surface to measure the thermal effect in the deep tissue layers in experiments with (a) and without (b) endoscopy apparatus.As can be seen from the thermal images, direct irradiation caused a temperature change of >7 C in the muscle layer.The apparatus, which exposed the target tissue to the laser beam in the recessed region, produced less than half of this temperature change in the deep tissue at the same laser fluence.The fluence was 1.67 kJ/cm 2 for 8 s of irradiation.Tissue thickness was measured as $3.4 mm, including all esophageal layers.Hypothetically due to coagulation necrosis, the mucosal whitening pattern in gross anatomy corresponds to F I G U R E 5 Representative thermal images of the esophagus acquired before and during laser irradiation from the muscular surface of the esophagus (outermost surface of the esophagus).(A) Endoscopic apparatus.(B) Direct-to-tissue collimation of laser irradiation as a control group.F I G U R E 6 Exemplary results in an ex vivo esophageal model for comparison between the endoscopic apparatus and direct-to-tissue collimation of laser irradiation as a control group.(A,D) Mucosal whitening is seen in gross anatomy.Dashed lines indicate where tissue samples were sliced.(B,E) Cross-sectional images of H&E-stained tissue sections.4Â magnification (0.13 NA).Scale bar = 0.5 mm.(C,F) Subtraction results of brightfield (BF) and darkfield (DF) microscopy images separately for cases with (dashed green lined box) and without (dashed blue lined box) endoscopy apparatus.E, epithelium; LP, lamina propria; M, muscle; MM, muscularis mucosa; SM, submucosa.2Â magnification (0.06 NA).Scale bar = 1 mm.In panels C and F, the region of interest is marked in bright green color for with apparatus and cyan color for control.

F I G U R E 7
Summary of all data predicted from the comparison study's subtraction results of brightfield and darkfield microscopy images.The pixel-based calculation determined the thermal injury's depth (A), width (B), and area (C) separately.*p ≤ 0.05, ***p ≤ 0.001, ****p ≤ 0.0001.