aDepartment of General Surgery, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
bCentre for Optical and Electromagnetic Research & International Research Center for Advanced Photonics, College of Optical Science and Engineering, Zhejiang University, Hangzhou 310058, China
cDepartment of Orthopedics, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
dInstitute for Lasers, Photonics, and Biophotonics, Department of Chemistry, State University of New York, Buffalo, NY 14260-3000, USA
eZhejiang Engineering Research Center of Cognitive Healthcare, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
fCollege of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou 310027, China
aDepartment of General Surgery, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
bCentre for Optical and Electromagnetic Research & International Research Center for Advanced Photonics, College of Optical Science and Engineering, Zhejiang University, Hangzhou 310058, China
cDepartment of Orthopedics, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
dInstitute for Lasers, Photonics, and Biophotonics, Department of Chemistry, State University of New York, Buffalo, NY 14260-3000, USA
eZhejiang Engineering Research Center of Cognitive Healthcare, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou 310016, China
fCollege of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou 310027, China
Optical imaging in the second near-infrared (NIR-II; 900-1880 nm) window is currently a popular research topic in the field of biomedical imaging. This study aimed to explore the application value of NIR-II fluorescence imaging in foot and ankle surgeries. A lab-established NIR-II fluorescence surgical navigation system was developed and used to navigate foot and ankle surgeries which enabled obtaining more high-spatial-frequency information and a higher signal-to-background ratio (SBR) in NIR-II fluorescence images compared to NIR-I fluorescence images; our result demonstrates that NIR-II imaging could provide higher-contrast and larger-depth images to surgeons. Three types of clinical application scenarios (diabetic foot, calcaneal fracture, and lower extremity trauma) were included in this study. Using the NIR-II fluorescence imaging technique, we observed the ischemic region in the diabetic foot before morphological alterations, accurately determined the boundary of the ischemic region in the surgical incision, and fully assessed the blood supply condition of the flap. NIR-II fluorescence imaging can help surgeons precisely judge surgical margins, detect ischemic lesions early, and dynamically trace the perfusion process. We believe that portable and reliable NIR-II fluorescence imaging equipment and additional functional fluorescent probes can play crucial roles in precision surgery.
Xiaoxiao Fan, Jie Yang, Huwei Ni, Qiming Xia, Xiaolong Liu, Tianxiang Wu, Lin Li, Paras N. Prasad, Chao Liu, Hui Lin, Jun Qian.
Initial Experience of NIR-II Fluorescence Imaging-Guided Surgery in Foot and Ankle Surgery.
Engineering, 2024, 40(9): 20-29 DOI:10.1016/j.eng.2024.04.011
With advancements in anatomy and surgical techniques, the field of surgery has transitioned from its reliance on empirical methods to a more precise approach [1], [2]. In 2015 Dong and Zhang [1] advocated the concept of precision surgery, emphasizing the importance of maximizing the resection of all diseased lesions and preserving normal tissues. Fluorescence imaging-guided surgery seamlessly aligns with the principles of precision surgery. Surgeons use visible or near-infrared (NIR) fluorescent probes to visualize diseased and normal tissues and facilitate resection. NIR fluorescence imaging-guided surgery has demonstrated remarkable achievements in the surgical field, facilitating the identification of sentinel lymph nodes, visualization of tumors, assessment of blood supply, and detection of nerves, among other applications [3], [4], [5], [6].
However, traditional NIR fluorescence imaging-guided surgery uses a bioimaging window ranging from 760 to 900 nm. The second near-infrared (NIR-II) window refers to the spectral region of 900-1880 nm, which has moderate light absorption, significantly reduced photon scattering, and minimal autofluorescence in biological tissues [7]. NIR-II fluorescence bioimaging provides greatly enhanced image contrast, particularly when deciphering structures from deep tissues in vivo. Since it was first proposed by Welsher et al. in 2009 [8], research on NIR-II bioimaging has drawn considerable attention [8], [9], [10]. Many types of fluorescent probes, such as single-walled carbon nanotubes, quantum dots, rare-earth-doped nanoparticles, and organic materials have been developed for this purpose [7], [11], [12]. Different types of NIR-II imaging systems, including NIR-II macroscopic imaging systems, NIR-II wide-field microscope, and NIR-II confocal microscope, have also been established [13]. With the development of fluorescent probes and imaging systems, NIR-II fluorescence imaging has significantly accelerated basic biological and medical research. It has been used to decipher cerebrovascular structure/function [14], [15] and the distribution of macrophages/T cells in lymph nodes [16], as well as for molecular imaging of tumors [17].
NIR-II fluorescence bioimaging presents a novel and promising opportunity for a diverse array of clinical applications, albeit at a nascent stage. Hu et al. [18] have made significant contributions to the clinical translation of NIR-II fluorescence imaging. In 2020, they introduced this novel technique for the surgical navigation of liver tumor. Compared to NIR-I fluorescence imaging-guided surgery, intraoperative NIR-II fluorescence imaging offers higher sensitivity for tumor detection and a higher tumor-to-normal tissue contrast ratio [18]. They further explored the application of NIR-II fluorescence imaging for the precise resection of brain gliomas. In their study, NIR-IIa (1300-1400 nm) and NIR-IIb (1500-1700 nm) bioimaging windows were first used clinically for brain blood vessel visualization during glioma resection [19]. The images acquired in the NIR-IIa and NIR-IIb windows performed better than those acquired in the traditional NIR-II and NIR-I windows. The limited number of studies conducted thus far has revealed significant potential for the application of NIR-II fluorescence imaging in clinical practice.
Few clinical studies have verified the applicability of NIR-II fluorescence imaging in foot and ankle surgery. In this study, we introduced the NIR-II imaging technique in orthopedic surgeries using our lab-established surgical navigation system. Better performance of NIR-II fluorescence imaging (beyond 1100 nm wavelength) was observed compared with NIR-I fluorescence imaging. Subsequently, we applied this innovative imaging technique to diabetic foot, calcaneal fracture, and pedicle flap surgeries. The utilization of NIR-II images, characterized by a higher signal-to-background ratio (SBR) and more imaging details, has facilitated the attainment of more accurate orthopedic surgeries. Our study demonstrates the great potential of NIR-II fluorescence imaging for foot and ankle surgeries, thus promoting its clinical translation.
2. Methods
2.1. Lab-established NIR-II fluorescence imager and commercial NIR-I fluorescence imager
A lab-established NIR-II fluorescence imager consisting of an NIR-II fluorescence imaging subsystem and laser excitation subsystem was developed. The NIR-II fluorescence imaging subsystem comprised of a cooled InGaAs camera (SW640; Xi'an Tianying Photoelectric Technology Co., Ltd., China) to record NIR-II images, an optical filter (FELH900 or FELH1100; Thorlabs, Inc., USA) to select the fluorescence wavelength, and an NIR lens (TKL35, F/1.6; Tianying Photoelectric Technology Co., Ltd.) to collect the images. As we used the emission tail of indocyanine green (ICG) to achieve NIR-II fluorescence imaging, the power density of the excitation laser could be very high if we selected a 1300 nm long-pass optical filter. Considering the balance between the imaging performance and power intensity of the excitation laser/imaging speed, the 1100 nm long-pass filter was selected for further clinical application. It efficiently acquires the fluorescence signal at the emission spectral tail of ICG sample and provides NIR-II fluorescence images with high contrast. The excitation subsystem was composed of an 808 nm laser (MW-GX-808; Changchun Laser Technology Co., Ltd., China) as the excitation source, a beam expander for adjusting the divergence, and an optical diffuser (EDP10-5; LBTEK, China) to ensure the uniformity of excitation. It can provide a suitable excitation power and excitation area (125 cm2) without excessive differences in the laser power between the center and edge. The PDE® near infrared fluorescence imager (Hamamatsu Photonics K.K., Japan) was a commercial NIR-I fluorescence imager and used as a contrast.
2.2. Measurement of absorption and emission spectra, animals, and NIR-I and NIR-II fluorescence imaging of blood vessels
The absorption spectra were collected using a UV-3600scanning spectrophotometer (Shimadzu Corporation, Japan). Emission spectra were obtained using an FLS980 photoluminescence (PL) spectrometer (Edinburgh Instruments, UK; signal detection: a liquid-nitrogen-cooled InGaAs diode detector). New Zealand rabbits (1.5 kg) were purchased from the Laboratory of the Animal Center of Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University. The rabbits were anesthetized by marginal ear vein injection of pentobarbital sodium salt solution (10 mg∙mL−1 in saline, 3 mL∙kg−1) to minimize any suffering. The abdominal hair of the rabbits was shaved and completely removed using a hair removal lotion. Then the rabbit was placed in the supine position and administered with 2 mL ICG (5 mg∙mL−1) via marginal ear vein injection. A 690 nm laser was used to provide uniform irradiation to the abdomen of the rabbit. A silicon-based charge coupled device (CCD) (GA1280; Tianying Photoelectric Technology Co., Ltd.) and an InGaAs camera were utilized to record the NIR-I fluorescence images through two 700 nm long-pass filters (FELH0700; Thorlabs, Inc.) and a 900 nm short-pass filter (FESH0900; Thorlabs, Inc.), and the NIR-II fluorescence images through a 900 nm long-pass filter or a 1100 nm long-pass filter (FELH900 or FELH1100) by an InGaAs camera. To minimize bias caused by ICG metabolism, the NIR-I and NIR-II systems were placed close to each other (Fig. S1 in Appendix A). Thus, we acquired NIR-I and NIR-II images almost simultaneously. The Experimental Animal Ethics Committee of Zhejiang University approved the study protocol (approval number: ZJU20210209).
2.3. Patient recruitment, clinical data collection, and imaging-guided surgery
Eight patients, including five with diabetic foot, two with calcaneal fractures, and one with lower-extremity trauma, were enrolled in this study. These patients were planned to receive a NIR imaging-guided surgery to assess the blood supply, using the PDE® near infrared fluorescence imager. After being fully informed about the study, the patients underwent NIR-II fluorescence imaging examination along with traditional ICG-assisted imaging-guided surgery. This study was approved and supervised by the Ethics Committee of Sir Run Run Shaw Hospital (approval number: 20220279). Basic data (age, weight, body mass index (BMI), sex, past medical history, allergic history, and smoking history) and operation-related data (ICG dosage and operative time) were collected.
During imaging-guided surgery, the imaging distance was approximately 50 cm from the target tissues to the NIR-II camera. We used the infrared component in the shadowless operating light to adjust the focal length to the appropriate position. The operating light was turned off and ICG was injected into the patient. We also prepared an operating light without an NIR component to replace the commonly used shadowless lamp for imaging-guided surgery. The exposure time was 10-25 ms to ensure fluency of the operative video. In general, after approximately 3 min, the fluorescence signal began to appear. The entire NIR-II fluorescence images were recorded.
2.4. Fluorescence images analysis and molecular docking
A two-dimensional (2D) Fourier transform was used to perform a spectral analysis of the fluorescence images and quantitatively compare the details in the images. Fast Fourier transform (FFT) was utilized on selected regions of all fluorescence images. The MATLAB function FFT2 (MATLAB version R2021b) was used to perform this transformation. The FFT was performed on the selected region. Statistically, the location of each pixel in the spectral image from the center and its intensity value were measured by plotting the curve with the frequency on the x-axis and intensity on the y-axis. The intensity integrals of each spectral image were normalized to the same value. The intensity profile of the image is plotted according to the lines in each figure. To calculate the SBR, the highest value was used as the signal, and the valley between the two signal peaks was used as the background value. The intensity ratio of the signal to the background was the SBR of the reference region. Quantitative analysis of the fluorescence intensity was performed using ImageJ software (version 1.6.0; National Institutes of Health, USA). Origin Pro software (version 9.0; OriginLab Corporation, USA) was used to generate graphs.
Coefficient of variation (CV) was used to assess the SBR or texture complexity of the images, and its calculation formula was CV = SD/mean × 100%. The mean represents the average value of the intensities across all pixels, whereas SD corresponds to the standard deviation. Squared mean differences of variance (SMD2) is an evaluation metric that focuses on the grayscale differences between adjacent pixels. The calculation formula was
where f(x, y) is the grayscale value of the corresponding pixel; x, y represents for the 2D coordinates of each pixel.
Paired two-tailed Student’s t-tests were used to assess differences between groups. Statistical significance was set at P < 0.05.
For molecular docking, the three-dimensional (3D) structure of albumin was downloaded from the Protein Data Bank. Molecular structures were constructed, and energy was minimized using Chem3D. Discovery Studio software (v19.1.0.18287) was used to simulate the molecule-to-protein binding site and stacking configuration.
3. Results
3.1. Human albumin significantly enhanced the fluorescence intensity of ICG in NIR-II imaging window
ICG is a clinically approved NIR fluorescent dye. With the development of NIR-II imaging techniques, researchers have gradually realized that the strong NIR-II emission tail of ICG provides a fluorescence image with a high SBR [20], [21]. Our group previously reported that ICG can achieve satisfactory results in both micro- and macro-NIR-II imaging [22], [23]. Even in the NIR-IIb imaging window, ICG exhibited excellent imaging performance [24]. ICG is a typical aggregation-caused quenching (ACQ) molecule with an emission peak located in the NIR-I imaging window. However, when ICG molecules are combined with proteins, such as albumin, the optical properties change significantly. Human serum albumin (HSA) plays an important role in human physiology. HSA is responsible for 60% of the plasma protein and nearly 80% of the osmotic pressure in the blood [25]. HSA often serves as carriers owing to the fact that a range of endogenous and exogenous ligands with association constants typically in the range of 104-106 mol∙L-1 bind reversibly to HSA [26]. Discovery Studio software was used to simulate the molecule-to-protein binding site and stacking configuration of ICG and HSA. Using score-ranking operations on the binding models, the model with the lowest absolute energy status (125.025) is presented (Fig. 1(a)). Receptor-ligand interactions were analyzed, and the outcome is presented as a 2D diagram (Fig. 1(b)). The ICG molecule interacts with Tyr452 via a conventional hydrogen bond and with Ala191 via a carbon-hydrogen bond. ICG interacts with Asp187, Lys436, Glu292, and Glu153 via π-cation or π-anion interactions. In addition, ICG interacted with Lys432, Phe157, His288, and Ala291 via π-alkyl interactions. By virtue of the interaction between ICG and HSA, π-π stacking of the aromatic structure of ICG was significantly reduced and intramolecular rotation was restricted. Previous reports have indicated that ICG has two emission peaks at low concentrations in water [27]. Similar results were observed in the present study (Fig. 1(c)). The fluorescence of ICG in a 5% HSA solution and ICG in human serum was significantly enhanced (Fig. 1(c)). It was also found that the fluorescence intensity of ICG in human serum was higher than that in a 5% HSA solution, which implied potential interactions of ICG with other proteins, such as high-density lipoproteins and low-density lipoproteins [28], in human serum. Notably, the photoluminescence spectra of ICG in 5% HSA solution and human serum were similar and exhibited an obvious red shift of approximately 40 nm compared with that of ICG in water.
To verify the application potential of ICG in NIR-II fluorescence imaging, images (in the 900 and 1100 nm long-pass spectral regions) of the abdominal wall blood vessels of a rabbit were obtained. ICG-based NIR-I fluorescence imaging was performed for comparison. All images were obtained using identical fields of view. In each image, the regions of interest were chosen to have identical dimensions and locations within a specified section (the region marked by the red box in Fig. 1(d)). A 2D FFT was used to analyze the frequency distribution of the images. After the FFT, a frequency shift was made to the Fourier-transformed image so that the low-frequency information was shifted to the center of the image. Using the center of the frequency spectral map as the origin, the average intensity of each image was obtained at different spatial frequencies after calculating the spectral intensity for different radii (Fig. 1(e)). Compared with the images acquired in the NIR-I window and 900 nm long-pass NIR-II window, the image taken in the 1100 nm long-pass NIR-II window possessed a larger area with a high spatial frequency (Fig. 1(e)), showing more imaging details and better resolution capability. To further compare the SBRs in the NIR-I and NIR-II bioimaging windows, the same blood vessels as those in the three fluorescence images were selected. The SBRs were 1.21, 1.62 and 3.02 in NIR-I, 900 nm long-pass NIR-II, and 1100 nm long-pass NIR-II windows, respectively. A higher SBR was detected in the NIR-II bioimaging window than that in the NIR-I bioimaging window, and the performance further improved beyond 1100 nm (Fig. S2 in Appendix A). Three rabbits were used to repeat this experiment, and the fluorescence images acquired beyond a wavelength of 1100 nm exhibited the highest average SBR (Fig. S3 in Appendix A).
3.2. Patient characteristics
Eight patients who had undergone three different surgeries were enrolled in this study. The three different types of surgeries included wound debridement for diabetic foot, open reduction and internal fixation (ORIF) for calcaneal fracture, and flap surgery. All these operations require an accurate judgement of the ischemia range. Detailed patient information is provided in Table S1 in Appendix A. The dosage of ICG used in our study was 7.5 mg each time because we found that it was sufficient to obtain satisfactory NIR-II fluorescence images for those surgeries, regardless of the patients’ body weight.
3.3. Lab-established NIR-II fluorescence imaging-guided surgery system
The PDE® near infrared fluorescence imager was used for traditionally (NIR-I) imaging-guided surgery (Fig. 2(a)). It is a mature commercial system widely used in clinical practice for liver cancer imaging, sentinel node tracing, and orthopedic surgery. Our lab-established NIR-II surgical navigation system mainly included an InGaAs camera, 808 nm laser, computer, and mobile cart (Figs. 2(b) and (c)). The maximum power density of the 808 nm laser was 30 mW∙cm-2.
Different types of foot and ankle surgeries were guided by a commercial system, which ensured that patients received standardized treatments. During surgery, NIR-II fluorescence images were collected to compare the imaging quality between the commercial NIR-I and our lab-established NIR-II imaging-guided surgery systems. Two representative regions, the tibialis anterior and foot dorsum regions, were selected. The contrast of NIR-II imaging was better than that of NIR-I imaging. For further quantitative analysis, representative parts (red dashed box) from the NIR-I and NIR-II images were cut and analyzed for their frequency spectral maps using FFT, as shown in Fig. 2(d). In the foot dorsum region, the proportion of high-frequency information in the NIR-II fluorescence image was 0.65, which was higher than that in the NIR-I fluorescence image (0.41). Similar results were observed in the tibialis anterior region. The proportion of high-frequency information in the NIR-II image (0.57) was higher than that in the NIR-I image (0.22). It is clear that NIR-II imaging provides more details than NIR-I imaging. To further compare image quality, the SBRs of the NIR-I and NIR-II images of this patient were also analyzed. In the foot dorsum region, the SBR of the NIR-II image was 2.67, which was higher than that of the NIR-I image (1.18). A higher SBR in the NIR-II images was also detected in the tibialis anterior region (NIR-I vs NIR-II: 1.31 vs 2.12, Fig. S4 in Appendix A). According to the aforementioned findings, the NIR-II imaging technique demonstrated a significantly better performance in surgical navigation than the traditional NIR-I technique.
3.4. NIR-II fluorescence imaging helps the judgement of necrotic areas in diabetic foot disease
Diabetic foot disease is one of the most troublesome complications of diabetes and the leading cause of diabetes disability burden [29]. An estimated 19%-34% of patients with diabetes develop foot ulcers during their lifetime [29]. Surgery is an indispensable component in the treatment of diabetic feet. Approximately 20% of patients with diabetic foot would require minor or major lower-extremity amputation [30]. Determining the extent of resection is crucial in diabetic foot surgery. A balance between maximizing necrotic tissue removal and maximizing the remaining healthy part of the lower limb is a major challenge for surgeons, and techniques have been developed to solve this problem. However, conventional methods including systolic pressure measurements, ankle-brachial indices, and transcutaneous oxygen pressures have low repeatability and accuracy [31]. Owing to neuropathy and microvascular dysfunction, the assessment of microcirculation is more difficult in diabetic patients than in non-diabetic patients [32]. To date, the range of debridement or amputation has been mainly determined by the surgeon’s experience. Considering the high SBR and spatial resolution of NIR-II fluorescence imaging, we used this technique for diabetic foot surgery.
Five patients with diabetic foot were included in this study, all of whom underwent minor amputation. During the operation, the patients were injected with ICG (7.5 mg) and underwent NIR-II fluorescence examination in addition to traditional NIR-I fluorescence imaging. We first gathered bright-field, NIR-I fluorescence, and NIR-II fluorescence images to compare the quality of the NIR-I and NIR-II fluorescence (Figs. 3(a)-(c)). Two evaluation parameters, CV and SMD2, were employed for quantitative analysis of the image quality. As depicted in Fig. 3(d), the CV values of the NIR-II images surpassed those of the NIR-I images in all three cases, indicating a better imaging performance of the NIR-II images. As shown in Fig. 3(e), the SMD2 values of the NIR-II images consistently exceeded those of the NIR-I images, further substantiating the enhanced imaging capability of the NIR-II imaging system. We also found that the background in the NIR-II images was clearly less than that in the NIR-I images, implying that NIR-II was less affected by environmental light.
Diabetic foot disease is caused by diabetes-related peripheral neuropathy, peripheral arterial disease, and infections [33]. The three main pathogenic factors of diabetic foot disease are shown in Fig. 4(a). Understanding the natural progression of diabetic feet is crucial for its prevention and management, which can be categorized into five distinct stages: superficial ulcers, deep ulcers, osteitis, and gangrene (partial or whole foot; Fig. 4(b)) [34]. Patient 3 had two types of diabetic foot ulcers. The lesion in the red circle is considered a superficial ulcer, whereas the lesion in the blue circle is considered a deep ulcer. In the NIR-II fluorescence images, it was observed that the superficial ulcer did not exhibit signal decrease (Fig. 4(c)). However, for the deep ulcer in patient 3, the NIR-II signals significantly decreased at the center of the ulcer and increased at the edge. This characteristic may help distinguish superficial ulcers from deep ulcers, thereby facilitating accurate treatment strategies for distinct ulcer types. Patient 4 presented an interesting case of diabetic foot disease. In this case, no obvious necrosis was detected in the center of the foot with the naked eye. However, in the NIR-II images, the fluorescence signal was minimal from the lateral margin to the central plantar region, suggesting the presence of necrotic tissue within the foot (Fig. 4(c)). After the skin was incised under NIR-II fluorescence imaging guidance, necrotic tissues were exposed (Fig. S5 in Appendix A). This case implies that NIR-II imaging could help identify necrotic areas before they can be observed by the naked eye. For the patient 5, obtained NIR-II images of the dorsal and plantar sides of the foot were obtained. The fluorescence signal on the plantar side of the great toe was negligible, suggesting gangrene (Fig. 4(c)). In addition, on the dorsal side of the foot, apart from the great toe, the fluorescence signal in the medial margin was lower than that in the rest of the foot (Video S1 in Appendix A). This case also demonstrated that changes in NIR-II fluorescence intensity were more obvious than morphological changes, which could help surgeons achieve more precise surgery. The wounds of all five patients recovered.
3.5. NIR-II fluorescence imaging facilitates the assessment for blood supply of surgical incision in calcaneal fracture surgeries
Closed calcaneal fractures account for 2% of all fractures and often require ORIF [35]. Wound infection and necrosis are two common short-term complications of ORIF that may lead to longer hospital stays, reoperation, and even disability [35]. The rate of post-operative wound necrosis in calcaneal ORIF (Fig. 5(a)) was higher than that in other ORIF surgeries. This is because of the unique anatomical characteristics of the local blood supply to the calcaneus and surrounding skin [36]. A comprehensive assessment of the wound blood supply is also important for calcaneal ORIF surgery. Previous studies used laser-assisted ICG angiography to assess wound blood supply [37]. The bioimaging window used in previous studies is located in the NIR-I spectral range. As previously stated, the value of NIR-II imaging in calcaneal ORIF is worth exploring.
Two patients with calcaneal fractures were enrolled. These two patients underwent ICG-assisted fluorescence imaging examination immediately and five days after the operation. Patient 1 was a 56-year-old male who underwent ORIF after a calcaneal fracture. The initial post-operative examination indicated satisfactory blood supply to the wound, with no non-fluorescent areas (Fig. 5(b)). In subsequent examinations, both the wound healing process and NIR-II fluorescence signal were good. For patient 2, a female, the NIR-II fluorescence signal surrounding the wound was negligible post-operation (Fig. 5(b), Video S2 in Appendix A). The wounds were closely monitored. Five days after surgery, necrosis was observed in a small area at the corner of the wound, and the NIR-II fluorescence signal in this region could not be observed. Wound healing is a dynamic process. Poor post-operative blood supply did not cause a frustrating outcome; however, it increased the risk of wound necrosis.
3.6. NIR-II fluorescence imaging assists the flap surgery
Owing to the thin soft tissues in the lower extremities, severe injury to the foot or ankle is always accompanied by bone and tendon exposure, which poses a great challenge to reconstruction surgery [38]. Flap reconstruction, including free and pedicle flaps, is a common treatment for large bone and tendon exposure [39]. Ensuring flap survival is a common concern of surgeons. Therefore, assessment of the local blood supply to the flap before and after flap transposition surgery is important, especially for pedicle flaps. The current methods include ultrasonography [40] and infrared thermography [41]. However, the value of NIR-II imaging in flap surgery remains unclear.
A patient with severe injury to the lower extremities underwent a flap delay procedure (Figs. 6(a) and (b)). This surgery included two steps: flap pre-treatment and flap transposition. The flap pre-treatment procedure (upper panel of Fig. 6(b)) blocked the blood supply surrounding the flap. It only reserves the blood supply from the flap pedicle, which could increase blood flow from the pedicle and improve the survival rate of the flap in the subsequent transposition step. The blood supply model gradually turned unidirectional after one week, and transposition surgery was performed. Before transposition, we observed the fluorescent signals of the flap using our laboratory-established NIR-II imaging system after ICG injection. As shown in Fig. 6(c) and Video S3 in Appendix A, the blood vessels originating from the pedicle were readily identifiable, and the distribution of the blood supply varied across different regions of the flap. The NIR-II signals in the area close to the pedicle appeared earlier than those in the distal area. Owing to the high resolution of NIR-II imaging, we accurately exhibited the entire process of flap imaging and subsequently proved that the blood supply was unidirectional, as expected. Approximately 2 min after the emergence of the signal, the entire flap exhibited a gradual manifestation of NIR-II fluorescence, indicating satisfactory blood supply to the flap (Fig. 6(d)). Transposition surgery was then performed (Fig. 6(b), Fig. S6(a) in Appendix A). The patient received a subsequent injection of ICG to assess the vascular perfusion of the flap following its transposition. The flap also presented good blood supply after 180° rotation (Fig. S6(b) in Appendix A). The fluorescence intensity of the flap exhibited a gradual increase, reaching its maximum within 90 s following the initiation of the signal alteration. This observation suggests that the alterations in the blood supply to the flap before and after transposition were negligible. The flap in the patient survived well after transposition surgery.
4. Discussion
With continuous advancements in fluorescence imaging technologies and their integration into the concept of precision medicine, personalized and intelligent orthopedic surgery has become a reality in clinical practice. In this study, we introduced NIR-II fluorescence imaging technology into foot and ankle surgeries by utilizing a lab-established surgical navigation system and achieved satisfactory navigation effects in diabetic foot, calcaneal fracture, and flap surgeries. The results demonstrate that NIR-II fluorescence imaging can produce high-quality images with superior SBR and better imaging details than the currently used NIR-I fluorescence imaging.
Owing to the moderate light absorption and low photon scattering of NIR-II fluorescence signals in biological tissues, NIR-II imaging technology has shown higher contrast than NIR-I when performing deep-tissue bioimaging; thus, it has gradually become a more promising technique in clinical practice [7]. With the assistance of ICG, a traditional US Food and Drug Administration-approved NIR fluorescent dye, NIR-I fluorescence imaging has been used extensively in the clinic for angiography, laparoscopic surgical guidance, malignant lymph node tracing, and tumor resection navigation [10]. Because the intense spectral tail of ICG extends into the NIR-II window, researchers have begun to explore the great potential of NIR-II imaging for surgical navigation.
This study expands the application of NIR-II imaging technology to orthopedic surgery and preliminarily verifies its application value. Assessment of blood supply during foot and ankle surgeries is important for intraoperative decision-making, and NIR-II imaging significantly facilitates this process. Changes in the blood supply usually occur prior to morphological alterations. Using NIR-II imaging during surgery could help surgeons accurately assess the blood supply and make proper intraoperative decisions. Doppler measurement is one of the most widely used methods for assessing blood supply [19], and its advantages include convenience and real-time examination. Compared with this traditional method, contactless, real-time, and high-resolution imaging with a larger field of view can be achieved using the NIR-II imaging technique. In addition, because our lab-established NIR-II fluorescence imaging equipment is portable, it has the potential for bedside examination or outpatient examination. In addition to the applications described in this study, NIR-II fluorescence imaging has potential applications in other orthopedic surgeries. For patients with chronic osteomyelitis, this technique can help identify necrotic bone tissue and ensure complete debridement. In patients with lower-extremity lymphedema, the lymphatic duct can be visualized by NIR-II fluorescence imaging after the subcutaneous injection of ICG. This technique can also help identify sentinel nodes in lower extremity malignancies. Although we emphasize the superior performance of NIR-II fluorescence imaging, this does not mean that the NIR-II imaging technique will replace the NIR-I imaging technique in the future. This can be an important supplement to the current surgical navigation systems. Combined with NIR-I imaging, multichannel imaging-guided surgery can provide novel navigation patterns.
There are three prospective avenues for future exploration in the field of NIR-II imaging-guided surgery. The first entails the creation of innovative NIR-II fluorescent probes characterized by extended peak emission wavelengths and increased quantum yields. Although our previous study indicated that another clinically approved dye, methylene blue, could also achieve NIR-II fluorescence imaging [42], the intensity was not high enough to fully meet clinical needs. Another study demonstrated that fluorescence imaging in the NIR-IIx sub-window (1400-1500 nm) could achieve the best performance in NIR-II imaging because the scattered NIR-IIx fluorescent photons could be most effectively attenuated by the moderate light absorption of biological tissues in this spectral region [7]. However, the brightness of the fluorescent probes remains the main limitation hindering the application of NIR-IIx imaging in clinical practice. Therefore, more fluorophores with better optical properties should be synthesized to fulfill the requirements of NIR-IIx imaging [43]. In addition, the clinical translation of NIR-II fluorescence probes with longer circulation time in the blood or tissue-specific targeting capability is expected [44], [45], [46]. Second, we believe that the development of a laparoscopic NIR-II fluorescence imaging system is another direction. Improving the pixels of the InGaAs camera and developing a laparoscopic lens for NIR-II imaging may be key points. Another potential direction is to introduce artificial intelligence (AI) to maximize the advantages of NIR-II imaging. Our team recently developed an AI-enhanced NIR-II fluorescence wide-field microscope with significantly improved optical throughput [47]. A similar idea can be utilized for clinically oriented NIR-II fluorescence imaging, and more precise surgeries are expected based on AI-enhanced higher-quality NIR-II fluorescence images.
Nevertheless, the main limitation of this study was the small number of patients, which made it difficult to draw conclusions through statistical analysis. To obtain more reliable results, future studies should aim to include a larger number of patients and set strict and reasonable control and experimental groups. In addition, this study only qualitatively assessed the blood supply to the lesion site and analyzed the absolute fluorescence intensity. In a cohort study, the relative fluorescence intensity, defined as the ratio of measured intensity to maximal intensity, is more suitable for statistical analysis because the imaging distance, exposure time, and power intensity of excitation vary in different patients. A more objective evaluation of the fluorescence intensity can be expected in future studies. Additional parameters must be included to draw more solid conclusions. Perfusion time, defined as the time from the onset of fluorescence to the peak value, is a valuable parameter. This parameter reflects blood microcirculation conditions. The perfusion time may increase when blood microcirculation is poor. In addition, the duration from maximum fluorescence to disappearance is worth recording, which can help assess venous function. The order of blood vessel appearance is another interesting parameter that can help distinguish arteries from veins. Finally, it is imperative to consider additional factors that could potentially impact fluorescence intensity, including scar tissue, crusts, edema, and other relevant variables. These factors have the potential to influence the accuracy of the clinical assessments performed during surgical procedures.
5. Conclusions
In this study, the application of the NIR-II fluorescence imaging technique in foot and ankle surgeries was demonstrated. Three different clinical application scenarios were included: diabetic foot, calcaneal fracture, and lower extremity trauma. Our lab-established NIR-II fluorescence imaging equipment enabled the acquisition of images of superior quality compared to the conventional NIR-I imager. NIR-II imaging has the potential to help surgeons judge surgical margins precisely, detect ischemic lesions early, and dynamically trace perfusion processes.
Acknowledgments
This work was supported by the Fundamental Research Fund for the Central Universities (K20220220), the National Key Research and Development Program of China (2018YFC1005003, 2018YFE0190200, and 2022YFB3206000), the National Natural Science Foundation of China (U23A20487, 82001874, 61975172, and 82102105), the Zhejiang Engineering Research Center of Cognitive Healthcare (2017E10011), the Natural Science Foundation of Zhejiang Province (LQ22H160017), the Zhejiang Province Science and Technology Plan Project (2022C03134), and the Science and Technology Innovation 2030 Plan Project (2022ZD0160703).
We sincerely thank Parikshit Asutosh Khadaroo from the Royal Melbourne Hospital to polish this manuscript.
Compliance with ethics guidelines
Xiaoxiao Fan, Jie Yang, Huwei Ni, Qiming Xia, Xiaolong Liu, Tianxiang Wu, Lin Li, Paras N. Prasad, Chao Liu, Hui Lin, and Jun Qian declare that they have no conflict of interest or financial conflicts to disclose.
DongJ, QiX. Liver imaging in precision medicine. EBioMedicine2018; 32:321-2.
[3]
CousinsA, ThompsonSK, WeddingAB, ThierryB. Clinical relevance of novel imaging technologies for sentinel lymph node identification and staging. Biotechnol Adv2014; 32(2):269-79.
[4]
MieogJSD, AchterbergFB, ZlitniA, HuttemanM, BurggraafJ, SwijnenburgRJ, et al. Fundamentals and developments in fluorescence-guided cancer surgery. Nat Rev Clin Oncol2022; 19(1):9-22.
[5]
ZhangZ, HeK, ChiC, HuZ, TianJ. Intraoperative fluorescence molecular imaging accelerates the coming of precision surgery in China. Eur J Nucl Med Mol Imaging2022; 49(8):2531-43.
[6]
LiuY, GuM, DingQ, ZhangZ, GongW, YuanY, et al. Highly twisted conformation thiopyrylium photosensitizers for in vivo near infrared-II imaging and rapid inactivation of coronavirus. Angew Chem Int Ed Engl2023; 62(13):e202214875.
[7]
FengZ, TangT, WuT, YuX, ZhangY, WangM, et al. Perfecting and extending the near-infrared imaging window. Light Sci Appl2021; 10(1):197.
[8]
WelsherK, LiuZ, SherlockSP, RobinsonJT, ChenZ, DaranciangD, et al. A route to brightly fluorescent carbon nanotubes for near-infrared imaging in mice. Nat Nanotechnol2009; 4(11):773-80.
[9]
WelsherK, SherlockSP, DaiH. Deep-tissue anatomical imaging of mice using carbon nanotube fluorophores in the second near-infrared window. Proc Natl Acad Sci USA2011; 108(22):8943-8.
[10]
Sevick-MuracaEM. Translation of near-infrared fluorescence imaging technologies: emerging clinical applications. Annu Rev Med2012; 63 (1):63217-31.
[11]
QianJ, FengZ, FanX, KuzminA, GomesASL, PrasadPN. High contrast 3D optical bioimaging using molecular and nanoprobes optically responsive to IR light. Physics Reports2022; 962:1-107.
[12]
LiY, ZhuH, WangXB, CuiY, GuLJ, HouXW, et al. Small-molecule fluorophores for near-infrared IIb imaging and image-guided therapy of vascular diseases. CCS Chemistry2022; 4(12):3735-50.
[13]
YuW, GuoB, ZhangH, ZhouJ, YuX, ZhuL, et al. NIR-II fluorescence in vivo confocal microscopy with aggregation-induced emission dots. Sci Bull2019; 64 (6):410-6.
[14]
ZhangM, YueJ, CuiR, MaZ, WanH, WangF, et al. Bright quantum dots emitting at approximately 1,600 nm in the NIR-IIb window for deep tissue fluorescence imaging. Proc Natl Acad Sci USA2018; 115 (26):6590-5.
[15]
LiY, FanX, LiY, LiuS, ChuahC, TangY, et al. Molecular crystal engineering of organic chromophores for NIR-II fluorescence quantification of cerebrovascular function. ACS Nano2022; 16(2):3323-31.
[16]
WangF, RenF, MaZ, QuL, GourguesR, XuC, et al. In vivo non-invasive confocal fluorescence imaging beyond 1,700 nm using superconducting nanowire single-photon detectors. Nat Nanotechnol2022; 17(6):653-60.
[17]
ZhongY, MaZ, WangF, WangX, YangY, LiuY, et al. In vivo molecular imaging for immunotherapy using ultra-bright near-infrared-IIb rare-earth nanoparticles. Nat Biotechnol2019; 37(11):1322-31.
[18]
HuZ, FangC, LiB, ZhangZ, CaoC, CaiM, et al. First-in-human liver-tumour surgery guided by multispectral fluorescence imaging in the visible and nearinfrared- I/II windows. Nat Biomed Eng2020; 4(3):259-71.
[19]
CaoC, JinZ, ShiX, ZhangZ, XiaoA, YangJ, et al. First clinical investigation of near-infrared window IIa/IIb fluorescence imaging for precise surgical resection of gliomas. IEEE Trans Biomed Eng2022; 69(8):2404-13.
StarosolskiZ, BhavaneR, GhaghadaKB, VasudevanSA, KaayA, AnnapragadaA. Indocyanine green fluorescence in second near-infrared (NIR-II) window. PLoS One2017; 12(11):e0187563.
[22]
CaiZ, ZhuL, WangM, RoeAW, XiW, QianJ. NIR-II fluorescence microscopic imaging of cortical vasculature in non-human primates. Theranostics2020; 10 (9):4265-76.
[23]
YuX, FengZ, CaiZ, JiangM, XueD, ZhuL, et al. Deciphering of cerebrovasculatures via ICG-assisted NIR-II fluorescence microscopy. J Mater Chem B Mater Biol Med2019; 7(42):6623-9.
[24]
HeM, WuD, ZhangY, FanX, ZhuangS, YangW, et al. Protein-enhanced NIR-IIb emission of indocyanine green for functional bioimaging. ACS Appl Bio Mater2020; 3(12):9126-34.
[25]
CarterDC, HoJX. Structure of serum albumin. Adv Protein Chem1994; 45:45153-203.
[26]
OlsonRE, ChristDD.In: Annual reports in medicinal chemistry. Elsevier; 1996. p. 327-36.
[27]
ChonB, GhannW, UddinJ, AnvariB, KundraV. Indocyanine green (ICG) fluorescence is dependent on monomer with planar and twisted structures and inhibited by h-aggregation. Int J Mol Sci2023; 24(17):13030.
[28]
YoneyaS, SaitoT, KomatsuY, KoyamaI, TakahashiK, Duvoll-YoungJ. Binding properties of indocyanine green in human blood. Invest Ophthalmol Vis Sci1998; 39(7):1286-90.
[29]
LazzariniPA, PacellaRE, ArmstrongDG, van Netten JJ. Diabetes-related lowerextremity complications are a leading cause of the global burden of disability. Diabet Med2018; 35(9):1297-9.
[30]
McDermottK, FangM, BoultonAJM, SelvinE, HicksCW. Etiology, epidemiology, and disparities in the burden of diabetic foot ulcers. Diabetes Care2023; 46(1):209-21.
[31]
NorgrenL, HiattWR, DormandyJA, NehlerMR, HarrisKA, FowkesFGR, et al. Inter-society consensus for the management of peripheral arterial disease (TASC II). Eur J Vasc Endovasc Surg 2007;33(Suppl 1):5-67.
[32]
VenermoM, SettembreN, AlbäckA, VikatmaaP, AhoPS, LepäntaloM, et al. Pilot assessment of the repeatability of indocyanine green fluorescence imaging and correlation with traditional foot perfusion assessments. Eur J Vasc Endovasc Surg 2016;52(4):527-33.
[33]
RiandiniT, PangD, TohMPHS, TanCS, LiuDYK, ChoongAMTL, et al. Diabetesrelated lower extremity complications in a multi-ethnic asian population: a 10 year observational study in Singapore. Diabetologia2021; 64(7):1538-49.
LinS, XieJ, YaoX, DaiZ, WuW. The use of cryotherapy for the prevention of wound complications in the treatment of calcaneal fractures. J Foot Ankle Surg2018; 57(3):436-9.
[37]
WylesCC, JacobsonSR, HoudekMT, LarsonDR, TauntonMJ, SimFH, et al. The Chitranjan Ranawat Award: running subcuticular closure enables the most robust perfusion after TKA: a randomized clinical trial. Clin Orthop Relat Res2016; 474(1):47-56.
[38]
EgelerSA, de JongT, LuijsterburgAJM, MureauMAM. Long-term patientreported outcomes following free flap lower extremity reconstruction for traumatic injuries. Plast Reconstr Surg2018; 141(3):773-83.
[39]
StephensCJ, SpectorJA, ButcherJT. Biofabrication of thick vascularized neopedicle flaps for reconstructive surgery. Transl Res2019; 211:211.
[40]
WuG, ZhangZ, ZhangF, ZhangY, WangQ, YuW. The free flap based on a single proximal perforator of the radial artery: ultrasonography study and clinical applications in reconstruction of soft tissue defects in finger. Eur J Med Res2022; 27(1):85.
[41]
ShokriT, LighthallJG. Perfusion dynamics in pedicled and free tissue reconstruction: infrared thermography and laser fluorescence video angiography. Am J Otolaryngol2021; 42(2):102751.
[42]
XueDW, WuD, LuZ, NeuhausJ, ZebibulaA, FengZ, et al. Structural and functional NIR-II fluorescence bioimaging in urinary system via clinically approved dye methylene blue. Engineering2023; 22:149-58.
[43]
ShengZ, LiY, HuD, MinT, GaoD, NiJS, et al. Centimeter-deep NIR-II fluorescence imaging with nontoxic AIE probes in nonhuman primates. Research, 2020;(1):652-65.
[44]
LiY, GaoJ, WangS, DuM, HouX, TianT, et al. Self-assembled NIR-II fluorophores with ultralong blood circulation for cancer imaging and imageguided surgery. J Med Chem2022; 65(3):2078-90.
[45]
ZengX, LiaoY, QiaoX, LiangK, LuoQ, DengM, et al. Novel NIR-II fluorescent probes for biliary atresia imaging. Acta Pharm Sin B2023; 13(11):4578-90.
[46]
LyuS, LuS, GuiC, GuoC, HanJ, XiaoY, et al. A NIR-II photoacoustic/NIR-IIa fluorescent probe for targeted imaging of glioma under NIR-II excitation. J Med Chem2024; 67(3):1861-71.
[47]
ChenR, PengS, ZhuL, MengJ, FanX, FengZ, et al. Enhancing total optical throughput of microscopy with deep learning for intravital observation. Small Methods2023; 7(9):e2300172.
PDF
(2789KB)
