1. Introduction

Computer-assisted surgery is very valuable for surgeons in modern clinical surgery; however, it is not capable of differentiating a tumor from the healthy tissue locally. Although in conventional histology the neoplastic part can be distinguished from the removed tissue, the results depend on the perception of the examining histologist, as well as the location of the removed tissue [1]. White-light endoscopy is also a traditional technique for tumor detection; however, as it has been proved [2], it is very difficult for this technique to distinguish between the normal and the tumor tissue, sometimes leading to miss some flat lesions. Undetected tumor lesions may increase the possibility of tumor recurrence. Photodynamic diagnosis (PDD), in which the tumor and the normal tissue are indicated with different fluorescent colors, is an effective solution to the above problem and is widely employed in clinical applications [1]; the photosensitizer and the corresponding wavelength of laser affects the fluorescent colors.

Once the tumor tissue is found, the following step is to implement therapy in terms of oncology. Present minimally invasive surgery removes the tumor tissue, which may lead to loss of functionality after the operation [3]. Moreover, there are certain tumors that are very difficult to be removed, such as those infiltrating the brain parenchyma, and peripheral lung cancer. Photodynamic therapy (PDT), started approximately in 1960 by Lipson [4], who adopted a photosensitizer and the relevant red-colored laser to destroy tumor cells in terms of single-oxygen. PDT realizes the tumor therapy without tissue removal, which is beneficial for maintaining functionality after the operation.

The 5-Aminolevulinic acid (ALA) is able to be applied in diagnosis along with a laser of 375-400 nm laser, as well as in therapy along with 635 nm laser, respectively [5]. Therefore, it is available to use ALA as the photosensitizer for both diagnosis and therapy.

Several endoscope technologies for PDD or PDT are applied in clinical applications or research. Olympus produces a 4-mm out diameter endoscope for the PDD of bladder cancer, which is thin enough for most of human organ; however, it is too short for the diagnosis of tumor in the stomach or in other organs. In addition, the surgeon needs to change the endoscope after the diagnosis during operation, which is not convenient; and most importantly, using the therapy endoscope it is impossible or difficult to find the same endoscope area that was marked on the diagnosis endoscope image. Therefore, during operation, the surgeon may miss the tumor target when using different therapy and diagnosis endoscopes.

For PDT, two main laser endoscope technologies exist: one is the laser firing fiber separating from the endoscope [6]; another is the composite-type optical fiberscope with one working channel for inserting necessary surgery applications which are usually applied in clinics at present. Both technologies cannot control the laser irradiation field, which may hurt the surrounding normal tissue or blood vessel, thus causing serious complications or vessel necrosis. At present the surgeon covers the surrounding healthy tissues or blood vessel with silver paper, which increases the burden of surgeon, as well as elongates the operation time. Then because the channels for the laser fiber and the endoscope are different, they do not provide the same laser irradiation field as the view field; this may result in not-treatment of some tumor tissue, thus leading to tumor recurrence. Moreover, during laser irradiation, the target is invisible to the present endoscope because of the laser reflection, which has an effect on the surgeon observing the therapy process.

Yamanaka [7] proposed a rigid endoscope that could capture image and simultaneously transmit laser to target blood vessel for fetus surgery. However, it is rigid, which is not suitable for organs, such as the stomach, or for the target, which is not appropriate in front of the endoscope. In addition, it is designed to transmit 1064 nm laser. It is not suitable for PDT laser, the wavelength of which is within the range of visible light. In 2016, we published a flexible laser endoscope for PDT [8], which captures images and simultaneously transmits laser to selected targets; however, it does not support for photodynamic diagnosis (PDD) prior to the laser therapy. Changing endoscopes during operation creates a high risk of missing tumor tissue.

Therefore, a new endoscope is required, which would be able to address all or parts of the above-mentioned problems. That is to say, the new endoscope should satisfy the following conditions: it should be flexible to be inserted into the body; coaxial, with the same view field and laser irradiation field; it should have a controllable laser irradiation field, without hurting the surrounding tissue or blood vessle; the target should be visible, even during laser irradiation, to enable the surgeon to observe; it should integrate the diagnosis with the therapy function, meaning that only one endoscope will capture different images during the diagnosis and the therapy.

In this paper, we propose such a new flexible coaxial laser-steering endoscope, which captures images under the diagnosis and therapy modes. In order to realize these objectives, first, a flexible long imaging fiber bundle is applied to transmit the laser from the emitter to the selected target and image from outside to the camera. Then, a precision stage automatically moves the laser header and the laser focusing lens system to the target selected by the surgeon. The wavelengths of the PDD fluorescent light and the PDT laser are almost the same, but it is necessary to capture the diagnosis fluorescent light but shield the reflection laser during therapy. Therefore a flip mount with a notch filter and a color compensation filter is adopted in this system to adjust the filter according to different purposes. This system is believed to be an advanced device combining pharmaceuticals and automatic equipment, which completes the tumor diagnosis and therapy with minimal invasion.

The paper is structured as the following five parts. The flexible laser steering endoscope will be introduced in the second section, which follows the introduction. Next, the experimental results are shown in the third section. Finally the discussion and conclusion will be presented in the fourth and fifth section, respectively.

2. Flexible laser steering endoscope for photodynamic diagnosis and therapy (PDD/T)

The main purpose of this flexible endoscope is to capture images under photodynamic diagnosis and therapy modes, and to transmit laser to the selected target under therapy mode. To achieve this, the system comprises two sub-optical systems, the observation sub-optical system and the laser-induced sub-optical system. The former captures images under different modes, by combining the objective lens system, fiber optics, and the eyepiece lens system. The latter positions the laser to the selected target, and includes the objective lens system, fiber optics and the laser focusing lens system. The objective lens system and fiber optics are shared by these two sub-optical systems.

Figure 1(A) shows a schematic illustration and the experimental setup of the flexible laser steering endoscope. The outside visible light (or fluorescent light) is focused via the objective lens system into the imaging fiber bundle; it is then reflected by the beam splitter; and focused by the eyepiece lens, through the notch filter (or color compensation filter), finally onto the camera sensor. Once the target is selected, the XY stage moves the laser fiber header and laser- focusing lens to the corresponding position. Then the focused laser is fired, through the beam splitter into the imaging fiber bundle, and is dispersed by the objective lens onto the selected target. Most of the construction parts were printed with 3D printer, and the materials were biocompatible; therefore the experimental setup in Fig. 1(B) was built to be suitable for the operation room. The detail specifications of the experimental setup is shown in Table 2.

 

Fig. 1 Illustration of flexible laser steering endoscope. (A) Schematic illustration of the laser endoscope (not drawn to scale). (B) The experimental setup.

Download Full Size | PDF

In the following sections, we will explain each part in detail. First we will present the principle of the shared objective lens system and fiber bundle. Then, the constructions of objective lens system, laser focusing lens system and eyepiece lens system are separately introduced.

2.1 The principle of the shared objective lens system and fiber bundle

It is known that the objective lens system, along with the imaging fiber bundle and the eyepiece, is a conventional observation optical system, which enables users to capture images. The reason for the sharing objective lens system and imaging fiber bundle, which can be used to position laser to selected target, is illustrated in detail. In the paraxial optical system, the laser ray transmission can be expressed via the translation and refraction matrix [9]. The laser is transmitted from the fiber bundle to the objective lens system; then, the angle and the height of laser ray are expressed as $u$ and $h$ respectively. When the laser ray passes through the optical lens system with a focal distance of $f_0$, we assume that the laser ray angle is$u^{\prime }=\tan \theta$, and height is $h^{\prime }$. Then we can express their relationship via Eq. (1).

Next, we obtain the equation $\tan \theta =u^{\prime }=-\frac{h}{f_0}$ which reveals that the angle of laser output ray does not depend on the input ray angle; it is only related to the ratio between the ray height and focal distance. Because the imaging fiber bundle is coaxial, each fiber in the fiber bundle is arranged cohesively, so that the same position is maintained at the input side and the output side. Therefore, in order to change the laser steering angle from the objective lens system, we would only need to change the laser input height to the imaging fiber bundle.

The laser power density is important during PDT; therefore, so it is necessary to control the laser transmission power and spot size. If a ray enters into the optic fibers at an angle $\theta$, the emergent ray will be distributed in a manner that they would fulfill a circle of a cone twice of angle $\theta$ [10]. We performed an experiment on our fiber bundle to understand its characteristics, through laser beam diagnostics (Spiricon, SP503U), which revealed that the output laser disperses with the increase in the incidence angle (the results are presented in Fig. 11). When selecting the laser scanning system, parallel laser ejection should be an important condition for our system. The galvanometric or polygon scanning system with a focusing lens causes severe distortion during laser scanning [11,12], thus rendering it difficult to ensure that the laser enters in a parallel focus into the fiber bundle. Thus, in our system, the XY hollow stage is considered as a precision instrument for the movement of the laser lens components.

Moreover, because the therapy laser wavelength lies within the range of visible light, the polarized beam splitter is a suitable choice for the laser transmission laser and the reflection of color images. Color images contain more information than gray images, and are important for the determination of the surgeon.

2.2 The objective lens system

To ensure that the outside light retains a parallel focus into the imaging fiber bundle, in the proposed system, the tele-centric lens system is designed in this system as shown in Fig. 2(A). In addition, because the laser from the fiber bundle would disperse quickly, the distance between the objective lens system and the fiber bundle is designed to be as short as possible to reduce the laser loss caused by the laser dispersion. The outer diameters of the lenses are 3 mm, which are smaller compared to those in our previous work [8]. Furthermore, the field angle is larger than 35° for the imaging fiber radius of 0.7 mm, which is significantly larger than that in the previous system [8]. Therefore, the objective lens system is improved in the present system, with a smaller outer diameter and a larger field of view. The detailed parameters are listed in Fig. 2(B).

 

Fig. 2 The objective lens system construction. (A) The construction. (B) The lens system parameters.

Download Full Size | PDF

2.3 The laser-focusing lens system

The lens system, as shown in Fig. 3(A), focuses the fiber laser to an as-small-as-possible are in order for it to be concentrated into one or several fibers in the imaging fiber bundle. The diameter of the lenses should be larger than the laser ray diameter to prevent diffraction losses. The load capacity is limited to 50g, whereas the weight of the designed lens system and that of the laser fiber header is 35g. The parameters are shown in Fig. 3(B), where the focused laser radius is smaller than that in our previous system [8]. The smaller the focused laser diameter, the smaller the number of fibers transmitted via the laser; therefore, the applied laser-focusing lens system in the proposed system produces a more improved laser-focusing spot.

 

Fig. 3 The laser-focusing lens system configuration. (A) The configuration. (B) The lens system parameters.

Download Full Size | PDF

2.4 The eyepiece lens system

The eyepiece lens system, as shown in Fig. 4(A), magnifies the image up to nearly two times from the parameters shown in Fig. 4(B). This Plössl group eyepiece is employed in this system because it can be suitably corrected for aberration.

 

Fig. 4 The eyepiece lens system construction. (A) The construction. (B) The lens system parameters.

Download Full Size | PDF

3. Experiments

In this section, we present the experimental results based on which we evaluate this flexible laser steering endoscope with regard to the following parts: the observation optical system, the laser-induced optical system, and the positioning accuracy based on the entire process of endoscopy. Finally, the in-vitro experimental results will be presented.

3.1 Experiments on the observation optical system

In this section, the experimental results on the observation optical system are presented, as shown in Fig. 5. In the Fig. 5(A), the view field lengths are 30, 44, 57, and 72 mm at distances of 20, 30, 40, and 50 mm respectively. Based on the data presented in Fig. 5(A), the view field length at a distance of 10 mm is inferred approximately 16 mm. Typically, the PDT is suitable for a tumor diameter of less than 1 cm; thus the view field of the proposed endoscope is sufficient for the therapy of common tumors at various distances. The following three images were recorded at a distance of 35 mm between the endoscope tip and the object test paper.

 

Fig. 5 The experimental results regarding the observation optical system. (A)The view field length of the endoscope. The horizontal axis (Distance), represents the distance away from the endoscope tip, 20-50 mm. The vertical axis (View filed), represents the diameter of the image at different distances. (B) The resolution of the endoscope. (C) The modulation transfer function (MTF) of the endoscope. (D) The chromatic aberration (CA) of the endoscope.

Download Full Size | PDF

In the Fig. 5(B), the 1951 USAF resolution test chart (Edmunds Optics Inc.) was applied to evaluate the image resolution of the system. The resolution based on the group number and element number is estimated using the following equation

The line pairs in the element 3 of group 3, marked by the red circle in the figure, could be distinct, as according to the equation, the resolution of the system is 10.08 lp/mm, which is approximately five times higher than the standard laparoscope resolution that is 2.0 lp/mm [13]. Thus, the endoscope system is believed to provide good-resolution images to the surgeons.

In Fig. 5(C), the modulation transfer function (MTF), which is defined as the proportionate contrast at a granted spatial frequency (between the output and input contrast), is illustrated. The test pattern paper was created using Imatest Test Charts and printed by a copy machine with the maximum resolution. The first-rate signs of the image sharpness are the spatial frequencies in which the MTF is half of the low frequency value (MTF 50) [14]. MTF 20 is closest to the analog-television lines, which is a specification of the horizontal resolution power of the analog camera. Therefore, we applied MTF 50 and MTF 20 to evaluate the sharpness of our system image. As shown in the figure, the results of several positions include MTF 50 (up) and MTF 20 (down) which are marked in the white block. The MTF 50 and 20 at the center are higher compared to those at the rim, thus implying that the sharpness of the image at the center of the image is better than that at the rim.

In Fig. 5(D), the chromatic aberration (CA), which exists because the glass refraction index is different with respect to the light wavelength, is measured to evaluate the distortion of this endoscope image caused by the lens system. Most of the CA values are less than 1, which correspond to a minor aberration according to Imatest. The values from 1 to 1.5 and those above 1.5 around the rim correspond to moderate and serious aberration respectively. This divergence in aberration values is caused by the characteristics of the lens; we cannot remove all of the aberration, therefore, in the proposed system, we ensure that the image aberration at the center of the image not affect the surgeon observing.

3.2 Experiments on the laser-induced optical system

The laser is transmitted from the laser equipment to the object by the laser-induced optical system; thus, the related evaluation experiment undergoes here.

The predicted laser power transmission efficiency is expected to be 19.39%. Figure 6 shows the laser power and the corresponding transmission efficiency at different position, therefore, the transmission efficiency of the system is approximately 2.44%, which is acceptable for the system. First, the laser-focusing lens system lost about 7.39% more than predicted. Then, the beam splitter is supposed to cost 50% of the laser, but it only transmits 45.53%. Finally, the fiber bundle and objective lens system lead to an abundance loss, for which there are two reasons. The first is that the transmission efficiency for the laser, the diameter of which is larger than that of one fiber, in the fiber bundle is significantly lower than the predicted. The second is that the pinhole in the objective lens system cuts more laser than it was supposed to. Detailed information is presented in Fig. 12.

 

Fig. 6 The laser power transmission efficiency (N = 5). (A) The laser power is measured at different positions with different laser diode currents. (B) The laser power transmission efficiency through the laser focusing lens system, beam splitter, and finally, exiting from the objective lens system.

Download Full Size | PDF

The laser power density is an important factor in the tumor therapy results. At distances of 20-50 mm, the laser spot diameters were measured, ranging from 1.663 to 1.723 mm horizontal, 2.277 and 2.584 mm vertical, we can calculate the laser power density (PD) by dividing total power by the size of the beam. Then, the laser power density of the proposed system can be derived, and it is listed in Table 1, with respect to the laser current and distance from the endoscope tip. The current of laser device ranges within 200-670 mA, the distance from the endoscope tip ranges within 20-50 mm. We used the laser-beam diagnostics (SP503U, Spiricon) to get the size of the beam. The size of the Gaussian beam (13.5% of the peak beam width) became smaller from 40 to 50 mm, so the power density increased at 50 mm compared to 40 mm.

Table 1. Laser power density (mW/$c{m}^2$) of the system.

View Table | View all tables in this article

According to Mahmoud [15], during operation, the power density should be maintained below 200mW/cm², because a high power density causes unnecessary thermal outcomes. Thus, at different distances, the system satisfies the requirement merely through the adjustment of the laser device current.

3.3 Experiments on the laser positioning accuracy

Simulated to the camera calibration, we performed stage calibration to the laser endoscope system in order to achieve automatic positioning of the laser to the selected target. The stage calibration is based on the following equation, the inner and extrinsic parameters, with the coordinate of target in the image. In order to obtain these parameters, we performed the following procedure: First, the extrinsic parameters [16] are obtained via the camera calibration, which is based on the chess pattern, as shown in Fig. 7(A). Then without moving the target, we fired the laser on to the target paper, thus recording different laser spot coordinates in the image and the distances of stage moving from the origin, as shown in Fig. 7(B). Finally, these parameters are applied to get the stage calibration parameters as described in Eq. (3).

 

Fig. 7 The stage calibration. (A) Obtaining the extrinsic parameters (rotation and translation matrix) based on camera calibration. (B) The recording of different laser spot coordinates in image and the distances of stage moving from the origin used for stage calibration.

Download Full Size | PDF

Typically, it is thought that more data will result in more precise parameters. Therefore, 10 groups of different chess pattern poses and distances are obtained to count out the stage calibration parameters (50 laser spot coordinates for each group). Based on these data, we conducted experiments to prove the accuracy of the laser endoscope system.

We adopted the laser spot on the camera image for stage calibration, as well as the accuracy. The accuracy of the laser positioning is evaluated in the air at a distance of 20, 30, 40, and 50 mm from the endoscope tip, respectively. 21 laser positions distributed over the image are selected for the experiment as shown in Fig. 8(A). After positioning laser to the selected target, the images are recorded, as shown in inserts Figs. 8(B-a) and Figs. 8(B-c), because the center of the laser spot is considered to be the selected target. We extracted the laser spot from the captured images by threshold, then count out the center of the laser spot in the image coordinate as the positioned target.

 

Fig. 8 The results of laser positioning accuracy (N = 5). (A) The selected 21 positions all spreading the image. (B) The examples of extracting laser spot in order to get the center of the laser spot. (a) and (c) are laser spots without outside illumination, (b) and (d) are the corresponding extracted spots. (C) The positioning accuracies at distances from 20 to 50 mm, the colorful points represent the error distance at different positions in the image, the red crosses are the average error distances at the corresponding distance from the endoscope tip.

Download Full Size | PDF

Next, the error distance, or positioning accuracy, is the Euclidean distance between the selected target and the positioned target, as shown in Fig. 8(C). As it may be observed, the error distances at the dim section of the image are mostly larger compared to those at the center of image, which is considered to be caused by the distortion of the lens. The maximum error in our system is less than 0.7 mm, which is less than a quarter of the laser spot size. Therefore the requirements is fully satisfied (the maximum error being half of the laser spot size).

Because the tumor size is always larger than the laser spot size, laser scanning is adopted in the endoscope system for photodynamic therapy. In order to save time during operation, the laser only scans the selected target tumor area, so the light would be delivered over multiple vector scanning sweeps. It is known that the tumor is always irregular, thus we assume that the arbitrarily selected tumor area is closed to obtain the area [17]. We arbitrarily select 20 area and record the area, the number of scanning points and the elapsed time for one scan; we then obtain the following fit equation:

Then, the iteration counts of the scanning is considered to be related to the required power energy, the area of the selected field, and the power energy of one laser spot during one scan. By adjusting the related parameters, the system would complete the therapy less than 15 minutes, which is acceptable for photochemical reactions. Therefore, after the user selects the target field and required power energy, the endoscope system automatically stops the laser irradiation after achieving the necessary value.

3.4 In-vitro experiments

In this section, we will describe the preparation of this experiment, and then the results will be presented.

Preparation of in-vitro experiment

Protoporphyrin IX (Pp IX, by Sigma-Aldrich, Stockholm, Sweden) is dissolved in dimethylformamide (DMF, by Sigma-Aldrich, Stockholm, Sweden) of spectroscopic purity (99.8%). The Pp IX concentration is 0.05 mmol/l in the experiments. The formula weight of Pp IX is 562.66 g/mol, the Pp IX concentration is 28.133 mg/l. The concentration of DMF in water is less than 5% to make sure the solution PH be less than 5%.

The gelatin from porcine skin (Sigma-Aldrich, Stockholm, Sweden) is applied to check the effectivity of the laser endoscope. The gelatin is solved in purified water (KENEI Pharmaceutical), during experiment at a temperature of about 55-60 °C. The light wavelength for photodynamic diagnosis is 400-410 nm, provided by flashlight, whose power is less than 2 W. The laser wavelength for photodynamic therapy is 635 nm. The developed laser endoscope is used as the equipment for the diagnosis, the therapy and for observing the target. The cuvette containing the Pp IX solutions and protein materials is irradiated. First, under the diagnosis mode using blue light for illumination, the fluorescent image is recorded with a camera. Then we arbitrarily selected part of the area as therapy target. Finally, the laser scanning in the laser endoscope system undergoes, and the laser power density is adjusted to approximately 45 mW/cm².

The laser plus photosensitizer could be applied into the photodynamic diagnosis and therapy, which is approved by researchers and surgeon [18,19]. Thus, we apply in-vitro experiment here to prove the system effectiveness in the present study. As illustrated aforementioned, the applied photosensitizer is 5-ALA, which is a natural biochemical precursor for heme synthesis in living mammalian cells [20–22]. After the photosensitizer precursor is injected, the biological metabolizing pathway from ALA to protoporphyrin IX (Pp IX) is absolutely necessary for the photodynamic diagnosis and therapy, because the irradiation laser reacts with the protoporphyrin IX (Pp IX), not with the ALA, to emit the fluorescent light for the diagnosis, or the singlet oxygen to destruct the tumor cell for the therapy. Therefore, in our experiment, the cell is not planned to be applied, the Pp IX is directly applied instead of ALA for which cell or tissue is necessary. Moreover, the main principle of the photodynamic therapy is that the singlet oxygen destructs the tumor protein [23]. According to some research works [24], the gelatin film could disappear after photo activation. Instead of a live animal, the gelatin, the main component of which is protein, is applied in this experiment.

Results of in-vitro experiments

In this experiment, the images captured under white light and blue light are compared, as shown in Fig. 9. The image in Fig. 9(A) is the phantom applied in the experiment, the gelatin containing the photosensitizer (PS) is placed at the center of the phantom (diameter: 15 mm), which is considered to be the simulated tumor tissue, whereas the gelatin without PS is placed at the outer ring of the phantom (diameter: 25 mm), which is assumed to be the simulated healthy tissue. The endoscope images are captured at a distance of 30 mm between the endoscope tip and the phantom. The results in Fig. 9(B) are under natural outside light, which is considered as the control experiment.

 

Fig. 9 The photodynamic diagnosis images during in-vitro experiment. (A) Phantom for experiment, (B) Image under white light, there is no difference between the two parts, (C) Image under blue light, it is dark red in the center of the phantom, because it is not obvious, we increase the contrast of the image as (D), in which the center part is obvious shown as red.

Download Full Size | PDF

As shown in Fig. 9, it is obvious that the captured images during the PDD mode are useful for diagnosing tumor/cancer part (red part in the images). In order to prove the difference, we also derived the histograms of the simulated tumor tissue and the simulated healthy tissue for one of the experimental results. The histograms of the red channel of the two image are shown in Fig. 10(C) and Fig. 10(D), and the number of pixel in Fig. 10(D) is more than that in Fig. 10(C). Therefore, the flexible endoscope has the ability to produce images for distinguishing the tumor from the healthy tissue in order to assist the surgeon in diagnosing the tumor part.

 

Fig. 10 The histograms of a PDD image. (A) The simulated healthy tissue and its corresponding histogram in (C), (B) The simulated tumor tissue and its corresponding histogram in (D).

Download Full Size | PDF

In the PDD mode, the surgeon could select the tumor tissue area easily via the red parts in the image. After confirming the area, the surgeon can change the mode to PDT, and determine the laser firing parameters. Then the tumor therapy can begin.

In the in-vitro experiment, we conduct the laser scanning experiment after selecting the simulated tumor area via different laser power, different distance, different laser scanning time, and different waiting time after laser scanning, however we did not find that the gelatin decreased as shown in the reference paper. This is attributed to the complicated response between the gelatin (protein) and the singlet oxygen. However, because the laser with sufficient power together with the photosensitizer are able to produce singlet oxygen in order to kill the tumor cell, which is a fact that has been proven by numerous surgeons and researchers [1–4]. In this work, the system provides sufficient laser power during the therapy, therefore it is expected to kill the tumor cell with the photosensitizer during PDT after determining the tumor area in the PDD mode, without changing endoscope to decrease the risk of omission during therapy, and ensure high-precision positioning to avoid hurting neighbor healthy tissue.

4. Discussion

In this paper, a flexible laser endoscope was developed, which diagnosed the tumor tissue in the PDD mode, and performed laser scanning with high precision during therapy. During the experiment, we also found certain disadvantages in this system.

First is the design of objective lens system. In the experiment regarding the laser power density, the main reason for the lower laser transmission efficiency than predicted is the loss caused by the objective lens system and fiber bundle. To ensure the outside-light parallel focus into the fiber bundle, based on the existing product with such a small diameter, a pinhole is a satisfactory choice in the design. However, this results in laser loss from the fiber bundle to the selected target, which is more than predicted. If there are suitable lenses available for the tele-centric lens system, we would not have used such pinhole in the system to provide parallel focus. Considering to avoid hurt on neighboring blood vessels, not the laser output, small laser spot is preferred, and this lens system would render the necessary spot size.

Second, the highest laser power in our system would hurt the fiber bundle. The maximum temperature of the fiber bundle transformation is 300~400 °C. According to the of Yamanaka [25], when the lens without coating is irradiated by a laser of 50 W for more than 30 s, the temperature increases by 5 °C; thus, in our system the highest laser power is less than 400 mW, thus rendering it impossible to produce such a high temperature for such a system. However, actually after measuring the maximum laser power density in the laser endoscope system, we observed certain black points appearing on the surface of the fiber bundle. Thus the power density is assumed as a possible reason. When the laser diode current reaches its maximum value, 670 mA, and the laser power into the fiber bundle incident surface is 375 mW.

Then the laser power density is expected to be higher than 103 kW/cm² (the radius of the laser spot is 34 microns). 1030 nm laser with a power density of 80 kW/cm² could injure the fiber surface [26]. Using the equation $PD(y)=\raisebox{1ex}{$PD(x){\lambda }_y$}\!\left/ \!\raisebox{-1ex}{${\lambda }_x$}\right.$ [27], the CW laser of 635 nm, the power density of which is higher than 49.3 kW/cm² would cause fiber injury. Thus the highest laser power density damages the imaging fiber bundle surface, hence yielding the black points. Therefore, although the imaging fiber bundle is suitable for our system to transmit laser during photodynamic therapy, such a fiber bundle is not currently available for high power density application.

Third, because the laser spot center is considered as the selected target, part of the laser is outside the selected tumor area during laser scanning. As presented in the above experimental result, the laser positioning errors are highest when the distance between the endoscope tip and the target is 50 mm, and the maximum positioning distance is less than 0.7 mm. The radii of laser spot at this distance are 1.242 mm and 0.866 mm, therefore, the out-field laser irradiation length should be less than 1.566 mm at the minor axis and 1.942 mm at the major axis, which is entirely acceptable for photodynamic tumor therapy. Moreover this length decreases as the distance between the endoscope tip and the target decreases.

Fourth, the illumination light and the blue laser have not been integrated into the laser endoscope at present, hence during the experiment outside light is applied. Our main purpose is to develop a flexible laser endoscope that would conduct diagnosis and therapy via the same endoscope device; it would perform laser scanning for therapy without hurting the neighboring healthy tissue, and would produce images under both two modes for the surgeons to observe. All these are achieved in the previously described experiments, and could be applied to perform a certain operation, such as brain tumor therapy. Integrating the illumination optical path into this system just improves the system to become suitable for all other inside-organ therapy, which is our next step.

Fifth, the device adopted vector scanning to complete laser therapy. Usually, the photochemical action time can be 10-15 minutes. By applying high-speed moving device in this system, it is completely able to achieve the same reaction as that by typical light delivery. About the waiting for long periods of time for scanning, surgeons need to spend large time and energy to cover neighboring blood vessel or healthy tissue before laser irradiation during operation, however, this device would save such time and energy. Therefore, this system would not extend the operation time.

In addition, the system is able to be applied combined with other medical equipment. In the oncology surgery, the tumor size for photodynamic therapy should be less than 10 or 15 mm, hence for a larger tumor, a tissue-removing device is necessary. Our system holds the diagnosis function, which renders it suitable to be applied in clinical applications combined with other medical equipment, such as tissue removing devices. Under the diagnosis mode, the caught image shows different-color tissue to the surgeon, which helps him/her to differentiate the tumor tissue to remove the tumor more accurately than under white-light illumination. Then the surgeon could observe the front view by the caught image while inserting the endoscope into the body. However, it is difficult to track the position and shape of the endoscope only by the visual field of the endoscope. The flexible endoscope could change its shape in 3D space, and the shape of the endoscope would be complex in certain cases. A better sense of the position and shape of the endoscope is more beneficial for the surgeon. Therefore, this laser endoscope is able to be combined with some tracking sensor. In addition, it is known that the magnetic resonance imaging (MRI) images are usually employed as data that navigate the surgeon toward the tumor target. After tracking the endoscope, this system is able to be combined with MRI images to guide the surgeon to insert the endoscope to the neighboring area of the tumor; and then under the PDD mode, the surgeon can find the accurate tumor tissue. Therefore, the laser endoscope is also available to be combined with certain navigation systems.

5. Conclusion

In the present work, a flexible laser-steering endoscope was developed, which could capture images under photodynamic diagnosis and therapy modes, while simultaneously transmitting laser to the selected target under the therapy mode. The experimental results entirely proved the effectiveness of this system. The image resolution, laser power density, and positioning accuracy sufficiently satisfied the requirements.

After complete the most important step for the device, which is to realize high-accuracy laser positioning and fluorescence image capturing. In the future, for the system in order to become more suitable for in-vivo and final clinic operation in an interior cavity, we plan to improve the system: first, making it waterproof; second, the flexible imaging fiber would thus controllable when inserted into the body inside an organ; Third, the illumination would be integrated to the system in order to render it fit to function in an organ.

Appendix

We provide some related experimental results. The results in Fig. 11 provide the evidence that we supply the XY stage to move the laser fiber header.

 

Fig. 11 Output laser distribution responds to different incident angles in the fiber bundle. With the incident angle larger, the output laser disperses, which leads to the laser power density decreasing. The laser power density is a very important factor for the photodynamic therapy, and it is not a good idea to cause power density loss by such way, therefore the XY stage is applied in the system to move the fiber header.

Download Full Size | PDF

In the Fig. 12, we compare the laser power of this laser endoscope (named PDD/PDT laser system) with that of our previous endoscope (named PDT laser system) to get acknowledge of the laser power effect factors.

 

Fig. 12 Laser power effect factors. (A) The laser spot of PDT laser system on image. (B) The laser spot of PDD/PDT laser system on image. The laser spot size in (A) is about 19.5 times of single fiber in fiber bundle, and that in (B) about 8.5 times. (C) The laser power from the laser focusing lens system by two systems, and apparently the laser power by PDD/PDT laser system is much higher than that by PDT laser system, which means that the smaller the laser focusing spot, the higher the output laser power. (The lenses in the two laser system are the same serial with almost the same transmission efficiency.) (D) The laser power transmits through objective lens systems by different laser systems. The pinhole size in the system is presumed to be without effecting the laser power transmission efficiency, but the less-size pinhole in PDD/PDT laser system leads to less laser power from the figure. (E) The laser power transmits from fiber bundle header to endoscope tip in PDT laser system. (F) The laser power transmits from fiber bundle header to endoscope tip in PDD/PDT laser system. The laser power difference values in the two figures prove that the smaller the laser focusing spot size, the higher the laser transmission efficiency of laser in the fiber bundle.

Download Full Size | PDF

Table 2. Specification the system construction parts

View Table | View all tables in this article

Acknowledgments

Author contributions: Under the direction of Ken Masamune, Yan Hu built the system and wrote the paper.

References and links

1. P. Crow, N. Stone, C. A. Kendall, R. A. Persad, and M. P. J. Wright, “Optical diagnostics in urology: current applications and future prospects,” BJU Int. 92(4), 400–407 (2003). [CrossRef]   [PubMed]  

2. H. G. Sim, W. K. O. Lau, M. Olivo, P. H. Tan, and C. W. S. Cheng, “Is photodynamic diagnosis using hypericin better than white-light cystoscopy for detecting superficial bladder carcinoma?” BJU Int. 95(9), 1215–1218 (2005). [CrossRef]   [PubMed]  

3. S. Jheon, T. Kim, and J. K. Kim, “Photodynamic therapy as an adjunct to surgery or other treatments for squamous cell lung cancers,” Laser Ther. 20(2), 107–116 (2011). [CrossRef]   [PubMed]  

4. R. L. Lipson, E. J. Baldes, and A. M. Olsen, “Hematoporphyrin derivative: a new aid for endoscopic detection of malignant disease,” J. Thorac. Cardiovasc. Surg. 42, 623–629 (1961). [PubMed]  

5. M. H. Abdel-Kader, Photodynamic Therapy: From Theory to Application (Springer-Verlag, 2014).

6. B. Pettiford and R. J. Landreneau, “Endobronchial stents and bronchial sparing surgery in the management of lung cancer,” Rev. Inst. Nal. Enf. Resp. Mex. 20(1), 33–41 (2007).

7. N. Yamanaka, H. Yamashita, K. Masamune, T. Chiba, and T. Dohi, “An endoscope with 2 DOFs steering of coaxial Nd:YAG laser beam for fetal surgery,” IEEE/ASME Trans. Mechatron. 15(6), 898–905 (2010).

8. Y. Hu and K. Masamune, “Flexible coaxial laser endoscope with arbitrarily selected spots in endoscopic view for photodynamic tumor therapy,” Appl. Opt. 55(30), 8433–8440 (2016). [CrossRef]   [PubMed]  

9. R. Shannon, Applied Optics and Optical Engineering (Elsevier, 2012).

10. H. F. Wolf, Handbook of Fiber Optics: Theory and Applications (Garland, 1979).

11. Y. K. Zhang, Optical Coherence Tomography Guided Laser-cochleostomy (KIT Sci. Pub., 2015).

12. L. Beiser, “Fundamental architecture of optical scanning systems,” Appl. Opt. 34(31), 7307–7317 (1995). [CrossRef]   [PubMed]  

13. S. A. Pierre, M. N. Ferrandino, W. N. Simmons, C. Fernandez, P. Zhong, D. M. Albala, and G. M. Preminger, “High definition laparoscopy: objective assessment of performance characteristics and comparison with standard laparoscopy,” J. Endourol. 23(3), 523–528 (2009). [CrossRef]   [PubMed]  

14. G. D. Boreman, “Modulation Transfer Function in Optical and Electro-Optical Systems,” Russ. Chem. Rev. 71(2), 159–179 (2001).

15. R. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision (Cambridge University, 2003).

16. H. Kostron, “Photodynamic Diagnosis and Therapy and the Brain,” in Photodynamic Therapy (Human, 261–280, 2010).

17. P. Bourke, Calculating the area and centroid of a polygon, Swinburne Univ. Technol., July 1988, Retrieved 6 Feb 2013.

18. C. Fritsch, K. Lang, W. Neuse, T. Ruzicka, and P. Lehmann, “Photodynamic diagnosis and therapy in dermatology,” Skin Pharmacol. Appl. Skin Physiol. 11(6), 358–373 (1998). [CrossRef]   [PubMed]  

19. D. E. Dolmans, D. Fukumura, and R. K. Jain, “Photodynamic therapy for cancer,” Nat. Rev. Cancer 3(5), 380–387 (2003). [CrossRef]   [PubMed]  

20. A. Sugitachi, K. Otsuka, T. Kimura, M. Hakozaki, M. Yaegashi, M. Kamishima, K. Kume, Y. Ohmori, S. Nishizuka, and A. Aasaki, “Color-imaging histodiagnostics approach for cancer,” Intergr. Molec. Med. 2(4), 231–233 (2015).

21. M. Wachowska, A. Muchowicz, M. Firczuk, M. Gabrysiak, M. Winiarska, M. Wańczyk, K. Bojarczuk, and J. Golab, “Aminolevulinic acid (ALA) as a prodrug in photodynamic therapy of cancer,” Molecules 16(12), 4140–4164 (2011). [CrossRef]  

22. L. Teng, J. I. Hamada, M. Nakada, S. G. Zhao, and Y. Hayashi, Current Applications of 5-ALA in Glioma Diagnostics and Therapy (INTECH Open Access Publisher, 2013).

23. P. Mroz, A. Yaroslavsky, G. B. Kharkwal, and M. R. Hamblin, “Cell death pathways in photodynamic therapy of cancer,” Cancers (Basel) 3(2), 2516–2539 (2011). [CrossRef]   [PubMed]  

24. G. López-Carballo, P. Hernández-Muñoz, R. Gavara, and M. J. Ocio, “Photoactivated chlorophyllin-based gelatin films and coatings to prevent microbial contamination of food products,” Int. J. Food Microbiol. 126(1-2), 65–70 (2008). [CrossRef]   [PubMed]  

25. N. Yamanak, “Coaxial laser steering endoscope system for fetus surgery,” Ph.D Dissertation, Department of IST, the University of Tokyo, Tokyo, Japan, 2012.

26. T. Ishida, T. Baba, H. Kuzushita, K. Hatooka, and K. Yoshida, “Development of highly damage resistant anti-reflection coating,” Mitsubishi Cable Indust. Rev. 100, 89–93, (2003).

27. Edmund optics Japan, “Laser Damage Threshold Testing,” Edmund Optics, [Online]. Available: http://www.edmundoptics.jp/technical-resources-center/lasers/laser-damage-threshold-testing/. [Accessed 2015].