Intraoperative Doppler Ultrasonography with Superb Microvascular Imaging for Intestinal Perfusion Assessment in Colorectal Cancer Surgery: A Prospective Pilot Study Authors: Atsushi Hirata, Gaku Ohira, Satoko Arai, Tetsuro Maruyama, Toru Tochigi, Koichiro Okada, Koichi Hayano, Masaya Uesato, Shunsuke Kainuma, Michihiro Maruyama Published: 2026 ABSTRACT AbstractObjectives: Intraoperative indocyanine green fluorescence imaging (ICG-FI) reduces anastomotic leakage, but has limitations related to contrast agent use, repeatability, and timing-dependent interpretation. We conducted an exploratory trial of a Doppler ultrasonography method using a high-frequency probe with superb microvascular imaging (SMI) to assess intestinal perfusion during colorectal cancer surgery. Methods: In this single-center prospective pilot study, we enrolled patients undergoing colorectal cancer resection with anastomosis. Intraoperatively, SMI assessed intramural microvascular flow to determine the perfusion boundary, followed by ICG-FI. The SMI boundary was defined as the line where vertical vessels penetrating the muscularis propria were no longer visible. The primary endpoint was the difference in blood flow boundaries between ICG-FI and SMI. The secondary endpoints were inter-observer reproducibility of SMI boundary identification and safety within 30 days. Results: Seven patients were analyzed for efficacy (one excluded because of invalid SMI data) and eight for safety. The median difference in blood flow boundaries between ICG-FI and SMI was −4 mm (range, −60 to 0 mm), with SMI identifying a more proximal boundary. Inter-observer variability for SMI boundary detection showed a median absolute difference of 2 mm (range, 0–4 mm). No procedure-related adverse events occurred. Conclusions: Our findings revealed differences in perfusion boundary location between SMI and ICG-FI, and intraoperative SMI acquisition and boundary identification were technically feasible. Analyzable data were obtained in seven of eight patients, with high inter-observer reproducibility and no procedure-related adverse events. Further prospective validation with standardized SMI acquisition and interpretation is required before clinical implementation. Keywords: colorectal cancer, intestinal perfusion, indocyanine green (ICG), Doppler ultrasonography, superb microvascular imaging (SMI), high-frequency ultrasound Introduction Anastomotic leakage is a feared complication in colorectal cancer surgery, occurring in 3.2–14.5% of cases[1], with the incidence rising to 10–18% in rectal cancer procedures[2-4]. Anastomotic leakage is associated with increased local recurrence rates and reduced long-term survival, making its prevention critical for improving patient outcomes[5]. Ensuring adequate blood supply to the anastomotic site is crucial for preventing this complication. Intraoperative bowel perfusion assessment using indocyanine green fluorescence imaging (ICG-FI) has demonstrated efficacy in reducing anastomotic leakage rates in clinical trials[6-8]. However, ICG fluorescence angiography has several limitations. First, it is contraindicated in patients with iodine hypersensitivity, and the adverse reaction rate is 3.13%[9]. Second, rapid and repeated assessments are limited by the presence of residual contrast agent and recirculation. Third, objective and reproducible evaluations are hindered by technical and environmental factors, including camera distance, ambient light, and ICG dosage[10]. Fourth, there is no consensus on the optimal timing of perfusion assessment after ICG administration[10]. Over time, the contrast agent can diffuse into ischemic zones, leading to an underestimation of poorly perfused areas[11]. A fluorescence time (FT) exceeding 60 s is reportedly associated with a higher risk of anastomotic leakage[12]. Moreover, relying on a simple cutoff time provides an inaccurate perfusion assessment as it disregards critical real-time hemodynamic factors such as the heart rate and blood pressure that directly influence blood flow[13]. This timing issue was highlighted as a potential limitation in the EssentiAL trial[7]. These challenges highlight the need for a novel intraoperative perfusion assessment method that overcomes the limitations of ICG. High-frequency ultrasound provides high spatial resolution for superficial structures at the expense of the penetration depth[14]. Superb microvascular imaging (SMI; Canon Medical Systems, Otawara, Japan) is an innovative Doppler technique that separates low-velocity flow signals from motion artifacts, allowing the assessment of microvascular flow by visualizing these low-velocity signals that were previously undetectable with conventional Doppler methods[15]. Combined use of a high-frequency ultrasound probe and SMI enables precise visualization of the superficial anatomy and microvascular flow, thereby allowing clear imaging of the superficial microcirculation in the fingertip and skin tissues[16,17]. In our previous retrospective study, Doppler ultrasonography using a high-frequency probe with SMI enabled clear visualization of intramural microvascular flow signals in the intestinal wall[18]; that observational study was approved by the Ethics Committee of Chiba University Hospital (Approval number: HK202311-11). Based on these findings, in this prospective exploratory pilot study, we aimed to quantify differences in the perfusion boundary location as identified by Doppler ultrasonography using a high-frequency probe with SMI compared with ICG-FI. We also assessed the technical feasibility of intraoperative SMI acquisition and boundary identification, including measurement reproducibility and perioperative safety. Methods Study Design This single-center, open-label, single-arm, prospective study was conducted at Chiba University Hospital, Japan. The trial was registered with the Japan Registry of Clinical Trials (jRCTs032240428). Ethical Considerations The study protocol was approved by the Certified Review Board of Chiba University Hospital (Approval number: CRB0095-24), and the study was performed in accordance with the ethical principles of the Declaration of Helsinki. All enrolled patients provided written informed consent after being informed of the objective, potential risks, and benefits of the study. Additionally, written informed consent for the publication of anonymized data and images was obtained from all participants. Our study guaranteed the protection of privacy and confidentiality of the participants by ensuring that the data were anonymized. Participants were not provided with any compensation for participating in the study. Study Participants Patients scheduled for elective surgery with anastomosis for colorectal cancer were enrolled in this study (regardless of the reconstruction method or surgical approach). The inclusion criteria were as follows: (1) histologically proven colorectal cancer; (2) aged ≥20 years at the time of consent; (3) Eastern Cooperative Oncology Group performance status of 0–2; and (4) adequate major organ function, as judged by the investigator to be safe for surgery. The exclusion criteria were as follows: (1) known history of allergy or hypersensitivity to ICG or iodine, (2) pregnancy, possible pregnancy, within 28 days postpartum, or breastfeeding, and (3) any other condition judged by the investigator to be inappropriate for study participation. The clinical stage of each patient was determined according to the Japanese Classification of Colorectal, Appendiceal, and Anal Carcinoma[19]. Surgical Procedure and Examination Protocol Standard colorectal surgical procedures, including total mesorectal excision or complete mesocolic excision with central vascular ligation and lymph node dissection, were performed according to the Japanese Classification of Colorectal, Appendiceal, and Anal Carcinoma[19]. After mesenteric division and before bowel transection, the intraoperative perfusion assessment protocol began with Doppler ultrasonography using SMI, followed by ICG-FI. Figure 1 shows the intraoperative examination setting and perfusion assessment method using SMI. A sterile graduated ruler was placed on the bowel and held in position to prevent movement, allowing accurate length measurement. The mesenteric transection line was designated as the baseline for all subsequent measurements. Distances from the baseline were recorded at 2-mm intervals, with positive values indicating the distal side and negative values the proximal side. The entire intraoperative assessment was video recorded for documentation and subsequent analysis. However, the actual point of bowel transection was determined exclusively based on the surgeon's interpretation of the ICG-FI findings, consistent with the standard clinical practice. The video recordings were used postoperatively to assess the inter-observer variability of SMI measurements and to quantitatively evaluate ICG-FI. Intraoperative setting and protocol for perfusion assessment using superb microvascular imaging (SMI). (a) Operating-room setup and intraoperative SMI examination. The examiner applies the ultrasound probe to the colon exteriorized through a small incision to observe the blood flow, and the assistant operates the ultrasound system. The real-time SMI video feed is fused with the surgical field view from the scope and recorded. (b) Close-up view of the examiner’s hand during SMI examination. The line drawn on the colon indicates the baseline corresponding to the mesenteric transection line. (c) The baseline was defined as the mesenteric transection line. A sterile graduated ruler was placed on the bowel, and distances were recorded every 2 mm. Positive values indicate the distal side and negative values the proximal side. (d) Monochrome superb microvascular imaging (mSMI) mode. Images were acquired using a standardized preset, and acquisition settings remained unchanged throughout boundary identification. Blood flow signals from the muscularis propria and submucosal layer are visualized. (i) Vertical vessels penetrating the muscularis propria. (ii) Horizontal vascular arcade in the submucosal layer. Blood Flow Assessment Using SMI Intraoperative SMI assessment was performed using an Aplio i700 and 33 MHz linear probe (Canon Medical Systems Corporation, Otawara, Japan). The system was set to the monochrome Superb Microvascular Imaging (mSMI) mode, and a standardized preset for mSMI was used throughout the study (Figure 1d). The mechanical index, B-mode gain, dynamic range, and frame rate (frames per second) were 0.5, 80, 70, and 55, respectively. The focal zone was placed on the bowel wall (focus depth, 3 mm); the imaging depth was set to 1 cm. Under the mSMI mode, the color gain and velocity scale were set to 42 and 0.5 cm/s, respectively. These settings were kept constant during boundary identification in each case. In all cases, intraoperative SMI acquisition and boundary identification were performed by a single trained examiner. The examiner completed training during our previous retrospective observational study under the supervision of an experienced sonographer (20 years of experience in ultrasonography). This training followed the same probe-handling protocol as that used for the present study. The probe, covered with a sterile probe cover, was gently applied parallel to the bowel's long axis. To avoid compression, a 2–3 mm layer of acoustic gel was maintained between the probe and bowel surface while scanning from the proximal to the distal side at a slow speed of approximately 1 cm/s. The SMI blood flow boundary was defined as “the demarcation line where the vertical vessels penetrating the muscularis propria were no longer visible” (Figure 1d). After identifying this boundary, we recorded “the demarcation line where submucosal flow signals were absent” (Figure 1d). The attenuation width in SMI was defined as “the distance between these two boundaries.” The real-time SMI video feed was combined with the surgical field view from the scope and recorded. To assess inter-observer variability, video recordings of the SMI scans were reviewed by two independent colorectal surgeons, each with 19 years of experience (observers A and B). The observers were blinded to all intraoperative findings and clinical information. As part of the standardized training tutorial, the observers jointly reviewed multiple intraoperative SMI cine-loop recordings with the examiner, archived from our previous study, and were instructed on the definition of the SMI blood flow boundary. After completing this training, each observer independently identified the SMI blood flow boundary without discussion with the other observer; thereafter, their measurements were compared with those of the intraoperative examiner. Blood Flow Assessment Using ICG Fluorescence Imaging Immediately following the SMI assessment, ICG (Diagnogreen; Daiichi-Sankyo Pharma, Tokyo, Japan) was prepared by dissolving 25 mg in 10 mL of normal saline (2.5 mg/mL). A 5-mL bolus (12.5 mg) was administered intravenously, then the infusion route was flushed with saline. The time point of the saline flush was defined as 0 s. The bowel was then observed for 90 s from a fixed distance of 20 cm using a near-infrared camera system (VISERA ELITE II, Olympus, Tokyo, Japan; or da Vinci Surgical System Xi with Firefly™ imaging, Intuitive Surgical, Sunnyvale, CA, USA). As no objective method for quantifying the fluorescence intensity was available intraoperatively, the ICG-FI blood flow boundary was defined as “the clear demarcation between the brightly fluorescent and non-fluorescent areas, as visually determined by the surgical team (operating surgeon and assistants).” The delayed fluorescence boundary was defined as “the most distal demarcation line of faint fluorescence observed within 90 s of ICG injection.” The attenuation width in ICG-FI was defined as “the distance between these two boundaries.” Quantitative evaluation of ICG fluorescence was performed postoperatively using the recorded video files and ImageJ software (v1.54, National Institutes of Health, USA). A 1 × 1 cm region of interest (ROI) was placed on the well-perfused bowel 1 cm proximal to the ICG-FI boundary (Figure 2). A time-intensity curve was generated by plotting the average green channel fluorescence intensity within the ROI over the 90-s observation period. The following quantitative parameters were derived from the fluorescence intensity-time curves (Figure 2). The maximum fluorescence intensity (Fmax) was defined as the peak value of fluorescence intensity. The time to the initial fluorescence increase (T0) was defined as the interval between the intravenous injection of ICG and the first detectable rise in the fluorescence intensity. Time to maximum fluorescence (Tmax) was defined as the interval from the onset of the fluorescence rise to the point of Fmax. FT was defined as the interval from the intravenous injection of ICG to the time of Fmax. The slope (Slope) was calculated by dividing Fmax by Tmax (Fmax/Tmax), providing a simplified indicator of the steepness of the fluorescence increase. Quantitative assessment of blood flow with indocyanine green fluorescence imaging (ICG-FI) and time–fluorescence intensity curve with quantitative parameters. (a) A region of interest (ROI, 1 × 1 cm) was placed on the well-perfused bowel, located 1 cm proximal to the perfusion boundary identified by ICG-FI. The fluorescence signal was continuously recorded for 90 s. Postoperatively, the average green channel fluorescence intensity within the ROI was analyzed using ImageJ software. (b) The parameters of T0, Tmax, FT, Fmax, and Slope were calculated from the time–fluorescence intensity curve. Outcomes The primary outcome was the difference in the blood flow boundaries between SMI and ICG-FI, recorded in 2-mm increments. The secondary outcomes were defined to assess the efficacy and safety of the SMI method. For efficacy, we evaluated inter-observer variability of the boundary assessment. For safety, we assessed the incidence of intraoperative and 30-day postoperative adverse events, graded according to the Common Terminology Criteria for Adverse Events version 5.0, including anastomotic leakage and length of postoperative hospital stay. The exploratory outcomes included the time required to determine the boundary with each method, the attenuation width of the blood flow boundary for SMI and ICG-FI, and quantitative parameters derived from the ICG-FI analysis (T0, Tmax, FT, Fmax, and Slope). In this study, technical feasibility was described based on successful intraoperative acquisition of analyzable SMI boundary data together with inter-observer reproducibility and perioperative safety. The acquisition of analyzable SMI data was not a separate prespecified endpoint. Statistical Analysis All statistical analyses were descriptive and were performed using SAS Software 9.4 (SAS Institute Inc., Cary, NC, USA). Continuous variables are presented as medians and ranges, whereas categorical variables are reported as counts. The analyzed population sets were defined as follows: the Full Analysis Set (FAS), the primary population for efficacy analysis, included all enrolled patients for whom the efficacy data were available. The Per Protocol Set (PPS) was a subset of the FAS that excluded patients with major protocol violations. The Safety Analysis Set (SAS) comprised all enrolled patients who underwent at least one study-related procedure and was used for all safety assessments. Sample size As this was an exploratory pilot study, formal sample size calculation was not performed. The target sample size was pragmatically set at eight cases based on the expected number of eligible patients who could be enrolled during the limited loan period of the ultrasound system and our institutional case volume, as prespecified in the statistical analysis plan. Results Study Participants The flow of participants in this study is summarized in Figure 3. Among the eight patients initially enrolled in this study, one patient was excluded from the FAS because valid SMI data could not be obtained. Thus, analyzable SMI-based boundary data were available for seven of the eight patients (87.5%). Accordingly, seven patients were included in the FAS, which was identical to the PPS. As the FAS and PPS were the same, this group was defined as the Efficacy Set. The SAS comprised all eight enrolled patients. Efficacy outcomes were analyzed in the Efficacy Set, whereas safety outcomes were analyzed in the SAS. All patients in the SAS (8/8) completed the scheduled 30-day postoperative follow-up. Participant flow diagram. FAS, Full Analysis Set; PPS, Per Protocol Set; SAS, Safety Analysis Set Patients were prospectively enrolled between December 2024 and February 2025. Baseline demographic and clinical characteristics of the Efficacy Set (n = 7) are summarized in Table 1. The median age of the seven patients was 72 years (41, 85), and five were men. None of the patients underwent preoperative chemotherapy or chemoradiotherapy. All patients had left-sided colon or rectal cancer. The tumors were located in the sigmoid colon (n = 4), upper rectum (n = 2), and lower rectum (n = 1). The surgical procedures included sigmoid resection (n = 4) and low anterior resection (n = 3); a diverting stoma was created in one case. Anastomosis was performed using the ECHELON CIRCULAR™ Powered Stapler (Ethicon, Somerville, NJ, USA) in all cases, employing the double stapling technique in five patients and the single stapling technique in two. Table 1.Baseline Demographic and Clinical Characteristics of the Efficacy Set.Demographicsn=7Age (years), median (range)72 (41, 85)Body mass index (kg/m2), median (range)24.6 (15.2, 27.5)Sexmale:female5:2ECOG-PS0:1:22:2:3Comorbidityhypertension1chronic lung disease2chronic kidney disease1Tumor locationsigmoid colon:upper rectum:lower rectum4:2:1clinical TNM stageI:II:III3:2:2Operative approachrobot:laparoscopy6:1Operative proceduresigmoid resection:low anterior resection4:3AnastomosisDST:SST5:2Diverting stoma1Operative time (min), median (range)242 (159, 362)Estimated blood loss (ml), median (range)10 (1, 30)ECOG-PS, Eastern Cooperative Oncology Group performance status; DST, double-stapling technique; SST, single-stapling technique Baseline Demographic and Clinical Characteristics of the Efficacy Set. ECOG-PS, Eastern Cooperative Oncology Group performance status; DST, double-stapling technique; SST, single-stapling technique Primary Outcome The primary outcome was the difference in the blood flow boundary between SMI and ICG-FI. As shown in Figure 4, the blood flow boundary in ICG-FI lay close to the baseline in all seven cases, with a median of 0 mm (−12, 10). In SMI, the boundary had a median of −4 mm (−70, 0); it was near the baseline in Cases 1–5, with values between −4 mm and 0 mm, whereas Case 6 was at −32 mm and Case 7 at −70 mm, indicating that the SMI boundary was located markedly more proximally relative to the baseline in these two cases. Consequently, the primary outcome, defined as the difference in the blood flow boundary between SMI and ICG-FI, had a median value of −4 mm (−60, 0). The individual positions are displayed in Figure 4b and 4c. Comparison of blood flow boundaries assessed intraoperatively by superb microvascular imaging (SMI) and indocyanine green fluorescence imaging (ICG-FI) in individual patients. (a) Schematic of the colon showing the mesenteric transection baseline (dotted line). Positive values indicate the distal side and negative values the proximal side. (b) Individual positions of blood-flow boundaries relative to the baseline (2-mm increments), and (c) the corresponding table summarizing the numerical values for each modality. Secondary Outcomes To assess the reproducibility of SMI boundary identification, inter-observer variability was evaluated (Figure 5). The median difference between the measurements of the intraoperative examiner and those of the two postoperative observers was 2 mm (0, 4), suggesting high reproducibility. In the safety assessment, no intraoperative adverse events or device malfunctions occurred in any of the eight patients. Postoperative adverse events were observed in two patients: one developed a stoma outlet obstruction, which resolved with temporary placement of a decompression tube via the stoma, and the other required a tracheostomy for recurrent laryngeal nerve palsy. Neither of these events was considered directly related to the study procedures. The median postoperative hospital stay was 9 days (6–37 days). Inter-observer reproducibility of blood flow boundary measurements with superb microvascular imaging (SMI). Observer A and Observer B are both colorectal surgeons, each with 19 years of clinical experience. Exploratory Outcomes The exploratory outcomes are summarized in Table 2. The median time required to determine the blood flow boundary was 35 s (24, 70) for ICG-FI and 43 s (30, 90) for SMI. In ICG-FI, the delayed fluorescence boundary was located at a median of 10 mm (−10, 26) from the baseline (Figure 4), resulting in a median attenuation width of 10 mm (0, 16). In SMI, the boundary of submucosal flow signal absence was located at a median of 14 mm (−60, 24) from the baseline (Figure 4), and the gap in the observed flow signals between the demarcation lines for the muscularis propria and submucosal layers, representing the attenuation width, exhibited a median of 18 mm (10, 52). The results of the quantitative assessment of blood flow using ICG-FI are presented in Table 2. For Case 4, the data could not be preserved and were therefore unavailable for analysis. Although VISERA ELITE II and the da Vinci Surgical System Xi with Firefly™ imaging were permitted by the protocol, the procedure was performed using VISERA ELITE II only in Case 4. Accordingly, the quantitative assessment of blood flow with ICG-FI was conducted in six analyzable procedures, all of which were performed with the da Vinci Surgical System Xi with Firefly™ imaging. Table 2.Comparison of Blood Flow Boundaries and Quantitative Assessment of Blood Flow Using Superb Microvascular Imaging (SMI) and Indocyanine Green Fluorescence Imaging (ICG-FI).Variable Case1234567median (range)Primary OutcomeDifferences in the blood flow boundaries between ICG-FI and SMI (mm)-40-140-4-20-60-4 (-60, 0)Exploratory OutcomesAttenuation width in ICG-FI (mm)1061610610010 (0, 16)Attenuation width in SMI (mm)1824201010521018 (10, 52)Time required to determine the blood flow boundary using ICG-FI (s)2424257035506535 (24, 70)Time required to determine the blood flow boundary using SMI (s)5834434030509043 (30, 90)Quantitative assessment of blood flowFT (s)213328M/V36716034.5 (21, 71)T0 (s)131914M/V26452522.0 (13, 45)Tmax (s)81414M/V10263514.0 (8, 35)Fmax155130150M/V805030105.0 (30, 155)Slope19.59.511.1M/V7.71.90.98.6 (0.9, 19.5)Attenuation width in ICG-FI: distance between the delayed fluorescence boundary and blood flow boundary in ICG-FI. Attenuation width in SMI: distance between the boundary where submucosal flow signals are completely absent and blood flow boundary in SMI. M/V: missing value Comparison of Blood Flow Boundaries and Quantitative Assessment of Blood Flow Using Superb Microvascular Imaging (SMI) and Indocyanine Green Fluorescence Imaging (ICG-FI). Attenuation width in ICG-FI: distance between the delayed fluorescence boundary and blood flow boundary in ICG-FI. Attenuation width in SMI: distance between the boundary where submucosal flow signals are completely absent and blood flow boundary in SMI. M/V: missing value Discussion In this pilot study, we quantified the difference in perfusion boundary between Doppler ultrasonography by a high-frequency probe with SMI and ICG-FI for intraoperative assessment of intestinal perfusion. A key finding was that, using SMI, a more proximal boundary of blood flow was identified compared with ICG-FI. In this pilot study, analyzable SMI data in seven of eight patients, high measurement reproducibility, and the absence of procedure-related adverse events supported the technical feasibility of intraoperative SMI acquisition and boundary identification. However, this study was not designed to determine clinical benefit or superiority over ICG-FI. Further studies are needed to determine the clinical utility of intraoperative SMI. The blood flow boundary using SMI was consistently more proximal compared with that using ICG-FI (median difference: −4 mm) (Figure 4). At least two factors could explain this discrepancy. First, SMI and ICG-FI might represent perfusion in different intramural layers. ICG fluorescence can be detected from tissues up to 12 mm deep under experimental conditions[20]. Therefore, in the colon wall, which is approximately 5 mm thick, ICG-FI is likely influenced by transmural perfusion. However, in our study, the SMI boundary was defined as the line where vertical vessels penetrating the muscularis propria were no longer visible. In our layer-specific observations, flow signals in the submucosa extended more distally compared with those in the muscularis propria (Figure 4). This suggests that ICG-FI might have partly reflected submucosal perfusion and thereby indicated a more distal boundary other than the muscularis-based SMI boundary. Second, SMI might fail to detect extremely low-flow microcirculation because of signal-to-noise limitations, specifically using high-frequency probes. Although SMI suppresses tissue-motion clutter and improves sensitivity to low-velocity microvascular flow compared with conventional color and power Doppler methods[21,22], high-frequency probes are more susceptible to frequency-dependent attenuation. In soft tissue, the attenuation coefficient is commonly described as approximately 0.5 dB/cm/MHz[23]. Thus, at 33 MHz, attenuation over only a few millimeters can markedly reduce echo amplitude and degrade the signal-to-noise ratio. As a result, weak microvascular flow signals may fall below the noise floor and also be removed by clutter-reduction filters, even when some perfusion is present. In future work, newer SMI implementations such as angio mode[24] might help improve visualization of very low-flow signals. Notably, in Cases 6 and 7, the SMI boundary shifted markedly proximally from baseline (−32 mm and −70 mm), resulting in large differences between ICG-FI and SMI of −20 mm and −60 mm, respectively. In these cases, the ICG fluorescence slope was lower than in the other cases (Table 2); moreover, a lower slope has been associated with reduced intestinal blood flow[25]. This finding suggests that perfusion might have been extremely low and close to the detection limit of SMI under these conditions. Nevertheless, both patients recovered without anastomotic leakage. Therefore, the clinical significance of a proximal SMI boundary remains unclear. Since the transection line was ultimately determined by ICG-FI, further studies are needed to clarify how SMI findings would relate to functional perfusion adequacy and clinical outcomes. A key advantage of our SMI method is its ability to visualize the individual layers of the bowel wall and independently assess blood flow in the muscularis and submucosal layers. Several alternatives to ICG-FI for perfusion assessment have been developed, including laser speckle contrast imaging[26,27], near-infrared spectroscopy[28], and tissue oxygen saturation (StO2) imaging[29]. However, none can directly evaluate the distinct blood flow in the submucosal and muscularis layers. Although a previous study assessed blood flow using the Doppler method, it focused on the mesenteric vessels and did not directly measure the intramural perfusion of the bowel wall itself[30]. In contrast, our SMI method enabled direct and separate evaluation of blood flow signals within these bowel wall layers and revealed a consistent gap between their respective perfusion boundaries. Specifically, the submucosal flow signals consistently extended more distally than those of the muscularis propria (Figure 4). This finding corresponds to the established vascular anatomy of the bowel wall[31,32] and suggests that blood flow reaches further distally via the submucosal vascular arcade, as shown in Figure 6. Intramural vascular anatomy based on layer-specific perfusion of the bowel wall assessed by superb microvascular imaging (SMI). (a) Schematic of the bowel wall layers and mural vasculature, including the marginal artery and vasa recta. The arrow indicates the presumed blood flow pathway after mesenteric division. (b) Intraoperative monochrome SMI (mSMI) mode shows layer-specific intramural flow signals. Submucosal flow signals extend more distally than those from the muscularis propria after mesenteric division. Our findings on layer-specific perfusion may also help to address a known limitation of ICG-FI in determining the optimal timing for assessment. The fluorescence boundary has often been observed to gradually shift distally over time. In our study, this “delayed fluorescence” extended the perfusion boundary by a median of 10 mm (range, 0–16 mm) within 90 s (Table 2). Many clinicians recognize that this delayed signal does not necessarily indicate adequate perfusion, thus posing the risk of overestimating the well-perfused tissue area[33]. In this study, attenuation width was observed in ICG-FI and SMI. However, the relationship between them was not consistent across cases; its clinical significance remains unclear. Nevertheless, since the submucosal perfusion boundary consistently extended more distally than did the muscularis-based boundary in all cases using SMI, we hypothesized that the delayed fluorescence reflects reduced and slow blood flow in the submucosal vascular arcade. Verifying this hypothesis is an important direction for future studies to clarify the physiological blood flow dynamics within the intestinal wall. As the next step, applying SMI to live animal models is necessary to explore this hypothesis. This study has certain limitations. First, as a pilot study, our sample size was small; therefore, our results are preliminary. Second, as this was a single-center study, our results may not be fully generalizable, and larger multicenter studies are needed to validate our observations. Third, a potential learning curve exists for the technical application and interpretation of SMI images, which could influence outcomes. SMI requires skill in probe handling, image interpretation, and avoidance of compression artifacts. In this pilot study, intraoperative SMI acquisition and boundary identification were performed by a single trained examiner. Therefore, the generalizability of these findings to other operators and institutions remains to be established. Additionally, intra-observer variability was not quantified in this pilot study. Future studies should include repeated assessments to evaluate within-observer measurement consistency. Fourth, in one of the eight patients, valid SMI data could not be obtained, and the case was excluded from the efficacy analysis, suggesting that technical or patient-related factors can present challenges, highlighting the need to standardize the procedure. Despite these limitations, this study provides important initial insights into this novel perfusion assessment method. In conclusion, this pilot study demonstrated the technical feasibility of intraoperative SMI acquisition and boundary identification, based on analyzable perfusion boundary data obtained in most enrolled patients, relatively high inter-observer reproducibility, and no procedure-related adverse events. However, the unsuccessful acquisition in one case and variable agreement with ICG-FI across patients indicated technical limitations and measurement variability. This study was not designed to determine the clinical utility or superiority over ICG-FI. Therefore, standardized protocols for image acquisition and boundary interpretation, along with further prospective validation, are required before clinical implementation. Author Contributions Atsushi Hirata designed the study, conducted data analysis and interpretation, and drafted the initial manuscript. Satoko Arai provided expert technical advice on the ultrasound examination methods and contributed to the revision of the manuscript. The remaining authors critically revised the manuscript for important intellectual content. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work. Approval by Institutional Review Board (IRB) The study protocol was approved by the Certified Review Board of Chiba University Hospital (Approval number: CRB0095-24). Trial registration Japan Registry of Clinical Trials, jRCTs032240428. Conflicts of Interest The authors have no conflicts of interest to declare. Data Sharing Statement De-identified individual participant data (IPD) underlying the results will be available, along with the study protocol, statistical analysis plan, informed consent form, clinical study report, and analytic code. Data will be accessible from 12 to 36 months after publication to researchers with a methodologically sound proposal, subject to review and a data access agreement. Data may be shared for validation of the reported findings and related secondary analyses. Requests should be sent to the corresponding author (atsushirata0403@gmail.com). The trial was registered with the Japan Registry of Clinical Trials (jRCT; jRCTs032240428); trial registration details are available at the provided jRCT registration link (https://jrct.mhlw.go.jp/latest-detail/jRCTs032240428).