Cholecystectomy is a commonly used clinical procedure for the treatment of many biliary tract diseases, such as acute cholecystitis, gallstones, biliary obstruction, and biliary colic. However, this procedure is associated with a complication rate of 0.45%.1 Once complications occur after cholecystectomy, the patient suffers from continuously repeated treatments, with potentially seriously life-threatening results.2 Previously, Asbun et al.3 found that the main cause for cholecystectomy-associated complications was the misidentification of the biliary system due to inflammation, adhesion, fat, and the patients’ other clinical conditions. Clinically, intraoperative cholangiography is applied to prevent iatrogenic biliary injuries.4 But the operation procedure is complicated and involves a certain risk of wound infection.5 The intracorporeal ultrasound and magnetic resonance cholangio-pancreatography represent that other modalities are commonly used for diagnosis and preoperative detection of biliary injuries.6 These methods are resource, labor, and time intensive, require a high level of technical skills for clinicians, and are not suitable for real-time cholangiography.
In recent years, there has been an increasing interest in using near-infrared fluorescence imaging technology for surgical navigation.7 It has been used for sentinel lymph node mapping to treat breast cancer, liver cancer resection, and biliary system imaging.8,9 Indocyanine green (ICG) is a safe and effective near-infrared fluorescent dye already approved by the FDA for use in humans. It can effectively penetrate skin, gallbladder, bile duct, and other tissues to a depth of 5 to 8 mm.10
In this letter, we report a low-cost and portable fluorescence microscopic system (PFMS) that uses near-infrared fluorescence imaging to guide cholecystectomy. The PFMS device comprises a light source module, a detection module, and a Raspberry Pi computer, as shown in Figs. 1(a) and 1(b). The light source module consists of eight 690-nm LEDs for excitation and one 850-nm LED for background illumination, as shown in Fig. 1(c). The power distribution of the excitation light is simulated by the TracePro software package. The averaged intensity of the excitation light source is , and the maximal intensity is . The surgical field is , and the working distance is 50 mm. With numerical simulation, we are able to optimize the arrangement of the LED locations for uniform light distribution without irreversible photobleaching [Fig. 1(d)]. Based on Fig. 1(e), the light distribution measured by experiment coincides with that of simulation.
The detection module consists of a 5-megapixel OV5647 CMOS camera (OmniVision, Santa Clara, California) coupled with a ML15 lens system (Meimei Metering Electricity Technology Co., Shanghai, China) and a BLP01-785R optical filter (Semrock Inc., New York). The filter is placed between the lens system and the camera, with its angle of incidence and cutoff wavelength carefully designed to ensure appropriate fluorescence detection in the desired field of view. The detection module has an adjustable magnification from to . A Raspberry Pi computer is used for the primary imaging acquisition and analysis tasks, including synchronization of fluorescence excitation and background illumination, acquisition of fluorescence and background image, and image fusion at a frame rate of two frames per second.
Fluorescence imaging performance of our PFMS system is characterized at different ICG concentrations and compared with that of an in vivo Imaging System (IVIS) Lumina III small animal imaging system (PerkinElmer, Waltham, Massachusetts). According to Fig. 2(a), our PFMS system has a sensitivity similar to that of the IVIS system in an ICG concentration range from 0 to , with the peak fluorescence emission achieved at , which is fully consistent with the former research by Yuan et al.11 We have also evaluated the imaging depth of the system by embedding an agar phantom of ICG at in chicken breast tissue of different thicknesses. According to Figs. 2(b) and 2(c), while the turbid nature of the tissue precludes imaging at depths much , the signal still remains marginally above background out to 5 to 7 mm, suggesting that fluorescence signals from buried objects may be sensed (albeit with low SNR) at this depth.
Furthermore, in vivo validation of the PFMS system is carried out following the animal protocol approved by the Animal Care and Use Committee at the School of Life Sciences of University of Science and Technology of China (Protocol No. USTCACUC1501015). A total of nine adult female rats (Beijing Vital River Laboratory Animal Technology Co. Ltd., Beijing, China) with a weight of 25 to 35 g are equally divided for the following procedures: (1) no ligation (NL), (2) cystic duct ligation (CL, simulating correct cholecystectomy), and (3) common bile duct ligation (CBL, simulating incorrect cholecystectomy). Each animal is anaesthetized and fixed by medical tape, with its hepatobiliary system exposed by laparotomy. The PFMS device is used to guide the NL, CL, and CBL procedures, respectively, followed by tail vein injection of 0.2 mL ICG solution at a concentration of for ﬂuorescence imaging at an interval of 5 min for 2.5 h. Figure 3 shows the representative background (left column) and fluorescence (right column) images acquired by the PFMS device after each procedure. According to Figs. 3(a) and 3(b), strong fluorescence emission is observed in the biliary structure and the gall bladder, indicating NL. According to Figs. 3(c) and 3(d), the gall bladder shows markedly reduced fluorescence emission, indicating appropriate ligation of cystic duct. According to Figs. 3(e) to 3(g), fluorescence emission is observed in the gall bladder but absent from the duodenum, indicating iatrogenic ligation of common bile duct.
In summary, we propose a PFMS device for simultaneous background and fluorescence imaging of biliary structure during cholecystectomy. In comparison with other cholangiographic techniques, the proposed device has advantages of simple operation, portability, and real-time imaging at low cost without radiation hazard. Our benchtop study demonstrates that the PFMS device has imaging sensitivity and dynamic range similar to that of a commercial Xenogen IVIS system. Our in vivo study demonstrates the technical feasibility for precise localization of the biliary structure and identification of potential iatrogenic injury. Considering that human biliary size and anatomy are different from that of a rat, this letter only demonstrates the technical feasibility of biliary imaging without further assessment of the clinical feasibility of reduced postoperative complications. Further validation tests in large animals are necessary before clinical deployment of such a device. In addition to cholecystectomy, this device may find use in many other surgical procedures where fluorescence imaging is needed at both microscopic and macroscopic levels.
The authors have no relevant financial interests in this letter and no potential conflicts of interest relevant to disclose.
The project was partially supported by the National Natural Science Foundation of China (Nos. 81271527 and 81327803) and the Fundamental Research Funds for the Central Universities (No. WK2090090013). The authors are grateful for the help from Dr. Ting Yue, University of Science and Technology of China (USTC), School of Life Sciences and Zelin Yu (USTC, Department of Precision Machinery) in performing the ICG concentration experiment, Meng Li and Shulin Zhang (USTC, Department of Precision Machinery) for assistance with programming, Rong Ma (Chongqing Medical University) for assistance with animal experiments, and Chaoyu Yang (USTC, Department of Precision Machinery) for assistance with 3-D printing.