Deep Learning-Based Real-time Ureter Identification in Laparoscopic Colorectal Surgery.

BACKGROUND: Iatrogenic ureteral injury is a serious complication of abdominopelvic surgery. Identifying the ureters intraoperatively is essential to avoid iatrogenic ureteral injury. Here, we developed a model that may minimize this complication. IMPACT OF INNOVATION: We applied a deep learning-based semantic segmentation algorithm to the ureter recognition task and developed a deep learning model called UreterNet. This study aimed to verify whether the ureters could be identified in videos of laparoscopic colorectal surgery. TECHNOLOGY MATERIALS AND METHODS: Semantic segmentation of the ureter area was performed using a convolutional neural network-based approach. Feature Pyramid Networks were used as the convolutional neural network architecture for semantic segmentation. Precision, recall, and the dice coefficient were used as the evaluation metrics in this study. PRELIMINARY RESULTS: We created 14,069 annotated images from 304 videos, with 9537, 2266, and 2266 images in the training, validation, and test datasets, respectively. Concerning ureter recognition performance, precision, recall, and the Dice coefficient for the test data were 0.712, 0.722, and 0.716, respectively. Regarding the real-time performance on recorded videos, it took 71 ms for UreterNet to infer all pixels corresponding to the ureter from a single still image and 143 ms to output and display the inferred results as a segmentation mask on the laparoscopic monitor. CONCLUSIONS AND FUTURE DIRECTIONS: UreterNet is a noninvasive method for identifying the ureter in videos of laparoscopic colorectal surgery and can potentially improve surgical safety. Although this could lead to the development of an image-navigated surgical system, it is necessary to verify whether UreterNet reduces the occurrence of iatrogenic ureteral injury.

Automatic surgical phase recognition-based skill assessment in laparoscopic distal gastrectomy using multicenter videos.

BACKGROUND: Gastric surgery involves numerous surgical phases; however, its steps can be clearly defined. Deep learning-based surgical phase recognition can promote stylization of gastric surgery with applications in automatic surgical skill assessment. This study aimed to develop a deep learning-based surgical phase-recognition model using multicenter videos of laparoscopic distal gastrectomy, and examine the feasibility of automatic surgical skill assessment using the developed model. METHODS: Surgical videos from 20 hospitals were used. Laparoscopic distal gastrectomy was defined and annotated into nine phases and a deep learning-based image classification model was developed for phase recognition. We examined whether the developed model's output, including the number of frames in each phase and the adequacy of the surgical field development during the phase of supra-pancreatic lymphadenectomy, correlated with the manually assigned skill assessment score. RESULTS: The overall accuracy of phase recognition was 88.8%. Regarding surgical skill assessment based on the number of frames during the phases of lymphadenectomy of the left greater curvature and reconstruction, the number of frames in the high-score group were significantly less than those in the low-score group (829 vs. 1,152, P < 0.01; 1,208 vs. 1,586, P = 0.01, respectively). The output score of the adequacy of the surgical field development, which is the developed model's output, was significantly higher in the high-score group than that in the low-score group (0.975 vs. 0.970, P = 0.04). CONCLUSION: The developed model had high accuracy in phase-recognition tasks and has the potential for application in automatic surgical skill assessment systems.

Predictive risk factors for resection surface-related complications after laparoscopic simultaneous resection of primary colorectal tumor and synchronous liver metastases: a single-center retrospective study.

BACKGROUND: Laparoscopic simultaneous resection (LSR) of primary colorectal tumors and synchronous colorectal liver metastases (sCRLM) has been recently performed. This study aimed to evaluate the postoperative outcomes after LSR and determine the risk factors for resection surface-related complications (RSRC), such as postoperative biliary fistula and liver-transection surface abscess. METHODS: Between 2009 and 2022, consecutive patients with sCRLM who underwent LSR were included. We retrospectively analyzed clinicopathological data, including intraoperative factors and postoperative outcomes. The difficulty level of all liver resections was classified according to the IWATE difficulty scoring system (DSS). We then performed univariate and multivariate analyses to identify the risk factors for RSRC. RESULTS: Of the 112 patients, 94 (83.9%) underwent partial hepatectomy and colorectal surgery. The median DSS score was 5 points (1-11), with 12 (10.7%) patients scoring ≥ 7 points. Postoperative complications were observed in 41 (36.6%) patients, of whom 16 (14.3%) experienced severe complications classified as Clavien-Dindo grade IIIa or higher. There was no postoperative mortality. The most common complication was RSRC (19 patients, 17.0%). Multivariate analysis identified American Society of Anesthesiologists (ASA) classification ≥ 3 [odds ratio (OR) 10.3, 95% confidence interval (CI) 1.37-77.8; P = 0.023], DSS score ≥ 7 points (OR 5.08, 95% CI 1.17-20.0; P = 0.030), and right-sided colectomy (OR 4.67, 95% CI 1.46-15.0; P = 0.009) as independent risk factors for RSRC. Postoperative hospital stays were significantly longer for patients with RSRC than for those without RSRC (22 days vs. 11 days; P < 0.001). CONCLUSION: Short-term outcomes of LSR for patients with sCRLM were acceptable in an experienced center. RSRC was the most common complication, and high-difficulty hepatectomy, right-sided colectomy, and ASA classification ≥ 3 were independent risk factors for RSRC.

Deep learning-based vessel automatic recognition for laparoscopic right hemicolectomy.

BACKGROUND: In laparoscopic right hemicolectomy (RHC) for right-sided colon cancer, accurate recognition of the vascular anatomy is required for appropriate lymph node harvesting and safe operative procedures. We aimed to develop a deep learning model that enables the automatic recognition and visualization of major blood vessels in laparoscopic RHC. MATERIALS AND METHODS: This was a single-institution retrospective feasibility study. Semantic segmentation of three vessel areas, including the superior mesenteric vein (SMV), ileocolic artery (ICA), and ileocolic vein (ICV), was performed using the developed deep learning model. The Dice coefficient, recall, and precision were utilized as evaluation metrics to quantify the model performance after fivefold cross-validation. The model was further qualitatively appraised by 13 surgeons, based on a grading rubric to assess its potential for clinical application. RESULTS: In total, 2624 images were extracted from 104 laparoscopic colectomy for right-sided colon cancer videos, and the pixels corresponding to the SMV, ICA, and ICV were manually annotated and utilized as training data. SMV recognition was the most accurate, with all three evaluation metrics having values above 0.75, whereas the recognition accuracy of ICA and ICV ranged from 0.53 to 0.57 for the three evaluation metrics. Additionally, all 13 surgeons gave acceptable ratings for the possibility of clinical application in rubric-based quantitative evaluations. CONCLUSION: We developed a DL-based vessel segmentation model capable of achieving feasible identification and visualization of major blood vessels in association with RHC. This model may be used by surgeons to accomplish reliable navigation of vessel visualization.

Intraoperative artificial intelligence system identifying liver vessels in laparoscopic liver resection: a retrospective experimental study.

BACKGROUND: The precise recognition of liver vessels during liver parenchymal dissection is the crucial technique for laparoscopic liver resection (LLR). This retrospective feasibility study aimed to develop artificial intelligence (AI) models to recognize liver vessels in LLR, and to evaluate their accuracy and real-time performance. METHODS: Images from LLR videos were extracted, and the hepatic veins and Glissonean pedicles were labeled separately. Two AI models were developed to recognize liver vessels: the "2-class model" which recognized both hepatic veins and Glissonean pedicles as equivalent vessels and distinguished them from the background class, and the "3-class model" which recognized them all separately. The Feature Pyramid Network was used as a neural network architecture for both models in their semantic segmentation tasks. The models were evaluated using fivefold cross-validation tests, and the Dice coefficient (DC) was used as an evaluation metric. Ten gastroenterological surgeons also evaluated the models qualitatively through rubric. RESULTS: In total, 2421 frames from 48 video clips were extracted. The mean DC value of the 2-class model was 0.789, with a processing speed of 0.094 s. The mean DC values for the hepatic vein and the Glissonean pedicle in the 3-class model were 0.631 and 0.482, respectively. The average processing time for the 3-class model was 0.097 s. Qualitative evaluation by surgeons revealed that false-negative and false-positive ratings in the 2-class model averaged 4.40 and 3.46, respectively, on a five-point scale, while the false-negative, false-positive, and vessel differentiation ratings in the 3-class model averaged 4.36, 3.44, and 3.28, respectively, on a five-point scale. CONCLUSION: We successfully developed deep-learning models that recognize liver vessels in LLR with high accuracy and sufficient processing speed. These findings suggest the potential of a new real-time automated navigation system for LLR.

Surgical step recognition in laparoscopic distal gastrectomy using artificial intelligence: a proof-of-concept study.

PURPOSE: Laparoscopic distal gastrectomy (LDG) is a difficult procedure for early career surgeons. Artificial intelligence (AI)-based surgical step recognition is crucial for establishing context-aware computer-aided surgery systems. In this study, we aimed to develop an automatic recognition model for LDG using AI and evaluate its performance. METHODS: Patients who underwent LDG at our institution in 2019 were included in this study. Surgical video data were classified into the following nine steps: (1) Port insertion; (2) Lymphadenectomy on the left side of the greater curvature; (3) Lymphadenectomy on the right side of the greater curvature; (4) Division of the duodenum; (5) Lymphadenectomy of the suprapancreatic area; (6) Lymphadenectomy on the lesser curvature; (7) Division of the stomach; (8) Reconstruction; and (9) From reconstruction to completion of surgery. Two gastric surgeons manually assigned all annotation labels. Convolutional neural network (CNN)-based image classification was further employed to identify surgical steps. RESULTS: The dataset comprised 40 LDG videos. Over 1,000,000 frames with annotated labels of the LDG steps were used to train the deep-learning model, with 30 and 10 surgical videos for training and validation, respectively. The classification accuracies of the developed models were precision, 0.88; recall, 0.87; F1 score, 0.88; and overall accuracy, 0.89. The inference speed of the proposed model was 32 ps. CONCLUSION: The developed CNN model automatically recognized the LDG surgical process with relatively high accuracy. Adding more data to this model could provide a fundamental technology that could be used in the development of future surgical instruments.

Preoperative risk factors for 90-day postoperative mortality outcome in patients with non-occlusive mesenteric ischemia.

PURPOSE: The mortality rate for non-occlusive mesenteric ischemia remains high even after patients survive the acute postoperative period with tremendous treatment efforts, including emergency surgery, which is challenging. The aim of this study was to explore the preoperative risk factors for 90-day postoperative mortality in patients with non-occlusive mesenteric ischemia. METHODS: This single-center, retrospective cohort study included patients diagnosed with non-occlusive mesenteric ischemia who underwent emergency surgery between August 2014 and January 2023. All patients were divided into survival-to-discharge and mortality outcome groups at the 90-day postoperative follow-up. Preoperative factors, including comorbidities, preoperative status of vital signs and consciousness, blood gas analysis, blood test results, and computed tomography, were compared between the two groups. RESULTS: Twenty patients were eligible, and 90-day mortality was observed in 10 patients (50%). The mortality outcome group had significantly lower HCO3- (20.9 vs. 14.6, p = 0.006) and higher lactate (4.4 vs. 9.4, p = 0.023) levels than did the survival outcome group. The median postoperative time to death was 19 [2-69] days, and five patients (50%) died after postoperative day 30, mainly because hemodialysis was discontinued because of hemodynamic instability in patients requiring hemodialysis. CONCLUSION: Low preoperative HCO3- and high lactate levels may be preoperative risk factors for 90-day postoperative mortality in patients with non-occlusive mesenteric ischemia. However, patients on hemodialysis die from discontinuing hemodialysis even after surviving the acute postoperative phase. Therefore, indications for emergency surgery in patients with risk factors for postoperative mortality should be carefully determined.

Machine learning-based Automatic Evaluation of Tissue Handling Skills in Laparoscopic Colorectal Surgery: A Retrospective Experimental Study.

OBJECTIVE: To develop a machine learning model that automatically quantifies the spread of blood in the surgical field using intraoperative videos of laparoscopic colorectal surgery and evaluate whether the index measured with the developed model can be used to assess tissue handling skill. BACKGROUND: Although skill evaluation is crucial in laparoscopic surgery, existing evaluation systems suffer from evaluator subjectivity and are labor-intensive. Therefore, automatic evaluation using machine learning is potentially useful. MATERIALS AND METHODS: In this retrospective experimental study, we used training data with annotated labels of blood or non-blood pixels on intraoperative images to develop a machine learning model to classify pixel RGB values into blood and non-blood. The blood pixel count per frame (the total number of blood pixels throughout a surgery divided by the number of frames) was compared among groups of surgeons with different tissue handling skills. RESULTS: The overall accuracy of the machine learning model for the blood classification task was 85.7%. The high tissue handling skill group had the lowest blood pixel count per frame, and the novice surgeon group had the highest count (mean [SD]: high tissue handling skill group 20972.23 [19287.05] vs. low tissue handling skill group 34473.42 [28144.29] vs. novice surgeon group 50630.04 [42427.76], P <0.01). The difference between any 2 groups was significant. CONCLUSIONS: We developed a machine learning model to measure blood pixels in laparoscopic colorectal surgery images using RGB information. The blood pixel count per frame measured with this model significantly correlated with surgeons' tissue handling skills.

Two-team lateral lymph node dissection assisted by the transanal approach for locally advanced lower rectal cancer: comparison with the conventional transabdominal approach.

BACKGROUND: An optimal surgical approach to lateral lymph node dissection (LLND) remains controversial. With the recent popularity of transanal total mesorectal excision, a two-team procedure combining the transabdominal and transanal approaches was established as a novel approach to LLND. This study aimed to clarify the safety and feasibility of two-team LLND (2team-LLND) and compare its short-term outcomes with those of conventional transabdominal LLND (Conv-LLND). METHODS: Between April 2013 and March 2020, 463 patients diagnosed with primary locally advanced rectal cancer underwent a transanal total mesorectal excision; among them, 93 patients who underwent bilateral prophylactic LLND were included in this single-center, retrospective study. Among these patients, 50 and 43 patients underwent Conv-LLND (the Conv-LLND group) and 2team-LLND (the 2team-LLND group), respectively. The short-term outcomes, including the operation time, blood loss volume, number of complications, and number of harvested lymph nodes, were compared between the two groups. RESULTS: The intraoperative and postoperative complications in the 2team-LLND group were equivalent to those in the Conv-LLND group; furthermore, the incidence of postoperative urinary retention in the 2team-LLND group was acceptably low (9%). Compared with the Conv-LLND group, the 2team-LLND group had a significantly shorter operation time (P = 0.003), lower median blood loss (P = 0.02), and higher number of harvested lateral lymph nodes (P = 0.0005). CONCLUSION: The intraoperative and postoperative complications of 2team-LLND were comparable with those of Conv-LLND. Thus, 2team-LLND was safe and feasible for advanced lower rectal cancer. Moreover, it was superior to Conv-LLND in terms of the operation time, blood loss volume, and number of harvested lateral lymph nodes. Therefore, it can be a promising LLND approach.

Universal meta-competencies of operative performances: a literature review and qualitative synthesis.

BACKGROUND: Prioritizing patient health is essential, and given the risk of mortality, surgical techniques should be objectively evaluated. However, there is no comprehensive cross-disciplinary system that evaluates skills across all aspects among surgeons of varying levels. Therefore, this study aimed to uncover universal surgical competencies by decomposing and reconstructing specific descriptions in operative performance assessment tools, as the basis of building automated evaluation system using computer vision and machine learning-based analysis. METHODS: The study participants were primarily expert surgeons in the gastrointestinal surgery field and the methodology comprised data collection, thematic analysis, and validation. For the data collection, participants identified global operative performance assessment tools according to detailed inclusion and exclusion criteria. Thereafter, thematic analysis was used to conduct detailed analyses of the descriptions in the tools where specific rules were coded, integrated, and discussed to obtain high-level concepts, namely, "Skill meta-competencies." "Skill meta-competencies" was recategorized for data validation and reliability assurance. Nine assessment tools were selected based on participant criteria. RESULTS: In total, 189 types of skill performances were extracted from the nine tool descriptions and organized into the following five competencies: (1) Tissue handling, (2) Psychomotor skill, (3) Efficiency, (4) Dissection quality, and (5) Exposure quality. The evolutionary importance of these competences' different evaluation targets and purpose over time were assessed; the results showed relatively high reliability, indicating that the categorization was reproducible. The inclusion of basic (tissue handling, psychomotor skill, and efficiency) and advanced (dissection quality and exposure quality) skills in these competencies enhanced the tools' comprehensiveness. CONCLUSIONS: The competencies identified to help surgeons formalize and implement tacit knowledge of operative performance are highly reproducible. These results can be used to form the basis of an automated skill evaluation system and help surgeons improve the provision of care and training, consequently, improving patient prognosis.

Single-center comparative study of short-term outcomes of transanal and laparoscopic total mesorectal excisions for low and middle rectal cancers.

BACKGROUND: Transanal total mesorectal excision is a promising surgical treatment for rectal cancer. However, evidence regarding the differences in outcomes between the transanal and laparoscopic total mesorectal excisions is scarce. We compared the short-term outcomes of transanal and laparoscopic total mesorectal excisions for low and middle rectal cancers. METHODS: This retrospective study included patients who underwent low anterior or intersphincteric resection for middle (5-10 cm) or low (< 5 cm) rectal cancer at the National Cancer Center Hospital East, Japan, from May 2013 to March 2020. Primary rectal adenocarcinoma was confirmed histologically. Circumferential resection margins (CRMs) of resected specimens were measured; margins ≤ 1 mm were considered positive. The operative time, blood loss, hospitalization length, postoperative readmission rate, and short-term treatment results were compared. RESULTS: Four hundred twenty-nine patients were divided into two mesorectal excision groups: transanal (n = 295) and laparoscopic (n = 134). Operative times were significantly shorter in the transanal group than in the laparoscopic group (p < 0.001). The pathological T stage and N status were not significantly different. The transanal group had significantly lower positive CRM rates (p = 0.04), and significantly lower incidence of the Clavien-Dindo grade III (p = 0.02) and IV (p = 0.03) complications. Both groups had distal margin positivity rates of 0%. CONCLUSIONS: Compared to laparoscopic, transanal total mesorectal excision for low and middle rectal cancers has lower incident postoperative complication and CRM-positivity rates, demonstrating the safety and usefulness of local curability for middle and low rectal cancers.

A comparative study of stoma-related complications from diverting loop ileostomy or colostomy after colorectal surgery.

PURPOSE: Even though minor, stoma-related complications significantly impact quality of life, they are often excluded from clinical analyses that compare short-term postoperative outcomes of loop ileostomy and loop colostomy. This study compares stoma-related complications between loop ileostomy and loop colostomy after rectal resection, including minor complications, and discusses the characteristics of diverting stoma types. METHODS: A retrospective review was conducted in patients who underwent diverting stoma construction after rectal resection. Data on patient background and postoperative short-term outcomes, including stoma-related complications and morbidity after stoma closure, were collected and compared between loop ileostomy and loop colostomy groups. Morbidities of all severity grades were targeted for analysis. RESULTS: A total of 47 patients (27 loop ileostomy, 20 loop colostomy) underwent diverting stoma construction following rectal resection. Overall stoma-related complications, incidence of skin irritation, high-output stoma, and outlet obstruction were significantly higher in the loop ileostomy group but high-output stoma and outlet obstruction were absent in the loop colostomy group. Regarding morbidity after stoma closure, operation times and surgical site infections were significantly higher in the loop colostomy group while anastomotic leakage after diverting stoma closure occurred (2 cases; 15%) in the loop colostomy group but not the loop ileostomy group. CONCLUSION: Because stoma-related complications were significantly higher in the loop ileostomy group, and even these minor complications may impair QOL, early loop ileostomy closure is recommended. For loop colostomy, stoma-related morbidities are lower but post-closure leakage is a calculated risk.

Artificial Intelligence-Based Total Mesorectal Excision Plane Navigation in Laparoscopic Colorectal Surgery.

BACKGROUND: Total mesorectal excision is the standard surgical procedure for rectal cancer because it is associated with low local recurrence rates. To the best of our knowledge, this is the first study to use an image-guided navigation system with total mesorectal excision. IMPACT OF INNOVATION: The impact of innovation is the development of a deep learning-based image-guided navigation system for areolar tissue in the total mesorectal excision plane. Such a system might be helpful to surgeons because areolar tissue can be used as a landmark for the appropriate dissection plane. TECHNOLOGY, MATERIALS, AND METHODS: This was a single-center experimental feasibility study involving 32 randomly selected patients who had undergone laparoscopic left-sided colorectal resection between 2015 and 2019. Deep learning-based semantic segmentation of areolar tissue in the total mesorectal excision plane was performed. Intraoperative images capturing the total mesorectal excision scene extracted from left colorectal laparoscopic resection videos were used as training data for the deep learning model. Six hundred annotation images were created from 32 videos, with 528 images in the training and 72 images in the test data sets. The experimental feasibility study was conducted at the Department of Colorectal Surgery, National Cancer Center Hospital East, Chiba, Japan. Dice coefficient was used to evaluate semantic segmentation accuracy for areolar tissue. PRELIMINARY RESULTS: The developed semantic segmentation model helped locate and highlight the areolar tissue area in the total mesorectal excision plane. The accuracy and generalization performance of deep learning models depend mainly on the quantity and quality of the training data. This study had only 600 images; thus, more images for training are necessary to improve the recognition accuracy. CONCLUSION AND FUTURE DIRECTIONS: We successfully developed a total mesorectal excision plane image-guided navigation system based on an areolar tissue segmentation approach with high accuracy. This may aid surgeons in recognizing the total mesorectal excision plane for dissection.

Glycan expression profile of signet ring cell gastric cancer cells and potential applicability of rBC2LCN-targeted lectin drug conjugate therapy.

BACKGROUND: Signet ring cell carcinoma (SRC) is a distinct subtype of gastric cancer (GC); however, the specific characteristics of cancer cell surface glycans and glycosylation remain unclear. In this study, we investigated SRC-specific glycans using lectin microarray and evaluated the potential applicability of a glycan-targeting therapy. METHODS: SRC cell lines (NUGC-4 and KATO-III) and non-SRC (NSRC) cell lines (NCI-N87, SNU-1, and MKN-45) were subjected to lectin microarray analysis to identify the SRC-specific glycans. Additionally, we performed immunohistochemical lectin staining and evaluated the anti-tumor effects of lectin drug conjugates (LDCs) using high-affinity lectins for SRC. RESULTS: Among the 96 lectins tested, 11 high-affinity and 8 low-affinity lectins were identified for SRC. Glycan-binding motifs varied in the high-affinity lectins, but 5 (62.5%) low-affinity lectins bound the same glycan structure, α2-6-linked sialic acids. The ratio of signal intensity in SRC to NSRC (SRC/NSRC) was highest in the rBC2LCN lectin (1.930-fold), followed by the BPL lectin (1.786-fold). rBC2LCN lectin showed high affinity for both SRC cell lines and one of the three NSRC cell lines (NCI-N87). The therapeutic effects of the LDC, rBC2LCN-PE38 (rBC2LCN, and Pseudomonas exotoxin A), showed cytocidal effects in vitro and tumor regression in in vivo mouse xenograft models. CONCLUSION: We reported specific glycan profiles in SRC cells, showing reduced α2-6-linked sialic acids. Additionally, we found a targeted therapy using rBC2LCN lectin might be applicable as an alternative treatment option for patients with SRC.

Real-time vascular anatomical image navigation for laparoscopic surgery: experimental study.

BACKGROUND: Recognition of the inferior mesenteric artery (IMA) during colorectal cancer surgery is crucial to avoid intraoperative hemorrhage and define the appropriate lymph node dissection line. This retrospective feasibility study aimed to develop an IMA anatomical recognition model for laparoscopic colorectal resection using deep learning, and to evaluate its recognition accuracy and real-time performance. METHODS: A complete multi-institutional surgical video database, LapSig300 was used for this study. Intraoperative videos of 60 patients who underwent laparoscopic sigmoid colon resection or high anterior resection were randomly extracted from the database and included. Deep learning-based semantic segmentation accuracy and real-time performance of the developed IMA recognition model were evaluated using Dice similarity coefficient (DSC) and frames per second (FPS), respectively. RESULTS: In a fivefold cross-validation conducted using 1200 annotated images for the IMA semantic segmentation task, the mean DSC value was 0.798 (± 0.0161 SD) and the maximum DSC was 0.816. The proposed deep learning model operated at a speed of over 12 FPS. CONCLUSION: To the best of our knowledge, this is the first study to evaluate the feasibility of real-time vascular anatomical navigation during laparoscopic colorectal surgery using a deep learning-based semantic segmentation approach. This experimental study was conducted to confirm the feasibility of our model; therefore, its safety and usefulness were not verified in clinical practice. However, the proposed deep learning model demonstrated a relatively high accuracy in recognizing IMA in intraoperative images. The proposed approach has potential application in image navigation systems for unfixed soft tissues and organs during various laparoscopic surgeries.

Deep learning-based automatic surgical step recognition in intraoperative videos for transanal total mesorectal excision.

BACKGROUND: Dividing a surgical procedure into a sequence of identifiable and meaningful steps facilitates intraoperative video data acquisition and storage. These efforts are especially valuable for technically challenging procedures that require intraoperative video analysis, such as transanal total mesorectal excision (TaTME); however, manual video indexing is time-consuming. Thus, in this study, we constructed an annotated video dataset for TaTME with surgical step information and evaluated the performance of a deep learning model in recognizing the surgical steps in TaTME. METHODS: This was a single-institutional retrospective feasibility study. All TaTME intraoperative videos were divided into frames. Each frame was manually annotated as one of the following major steps: (1) purse-string closure; (2) full thickness transection of the rectal wall; (3) down-to-up dissection; (4) dissection after rendezvous; and (5) purse-string suture for stapled anastomosis. Steps 3 and 4 were each further classified into four sub-steps, specifically, for dissection of the anterior, posterior, right, and left planes. A convolutional neural network-based deep learning model, Xception, was utilized for the surgical step classification task. RESULTS: Our dataset containing 50 TaTME videos was randomly divided into two subsets for training and testing with 40 and 10 videos, respectively. The overall accuracy obtained for all classification steps was 93.2%. By contrast, when sub-step classification was included in the performance analysis, a mean accuracy (± standard deviation) of 78% (± 5%), with a maximum accuracy of 85%, was obtained. CONCLUSIONS: To the best of our knowledge, this is the first study based on automatic surgical step classification for TaTME. Our deep learning model self-learned and recognized the classification steps in TaTME videos with high accuracy after training. Thus, our model can be applied to a system for intraoperative guidance or for postoperative video indexing and analysis in TaTME procedures.

Computer-assisted real-time automatic prostate segmentation during TaTME: a single-center feasibility study.

BACKGROUND: Urethral injuries (UIs) are significant complications pertaining to transanal total mesorectal excision (TaTME). It is important for surgeons to identify the prostate during TaTME to prevent UI occurrence; intraoperative image navigation could be considered useful in this regard. This study aims at developing a deep learning model for real-time automatic prostate segmentation based on intraoperative video during TaTME. The proposed model's performance has been evaluated. METHODS: This was a single-institution retrospective feasibility study. Semantic segmentation of the prostate area was performed using a convolutional neural network (CNN)-based approach. DeepLab v3 plus was utilized as the CNN model for the semantic segmentation task. The Dice coefficient (DC), which is calculated based on the overlapping area between the ground truth and predicted area, was utilized as an evaluation metric for the proposed model. RESULTS: Five hundred prostate images were randomly extracted from 17 TaTME videos, and the prostate area was manually annotated on each image. Fivefold cross-validation tests were performed, and as observed, the average DC value equaled 0.71 ± 0.04, the maximum value being 0.77. Additionally, the model operated at 11 fps, which provides acceptable real-time performance. CONCLUSIONS: To the best of the authors' knowledge, this is the first effort toward realization of computer-assisted TaTME, and results obtained in this study suggest that the proposed deep learning model can be utilized for real-time automatic prostate segmentation. In future endeavors, the accuracy and performance of the proposed model will be improved to enable its use in practical applications, and its capability to reduce UI risks during TaTME will be verified.

Lectin drug conjugate therapy for colorectal cancer.

Drug resistance represents an obstacle in colorectal cancer (CRC) treatment because of its association with poor prognosis. rBC2LCN is a lectin isolated from Burkholderia that binds cell surface glycans that have fucose moieties. Because fucosylation is enhanced in many types of cancers, this lectin could be an efficient drug carrier if CRC cells specifically present such glycans. Therefore, we examined the therapeutic efficacy and toxicity of lectin drug conjugate therapy in CRC mouse xenograft models. The affinity of rBC2LCN for human CRC cell lines HT-29, LoVo, LS174T, and DLD-1 was assessed in vitro. The cytocidal efficacy of a lectin drug conjugate, rBC2LCN-38 kDa domain of pseudomonas exotoxin A (PE38) was evaluated by MTT assay. The therapeutic effects and toxicity for each CRC cell line-derived mouse xenograft model were compared between the intervention and control groups. LS174T and DLD-1 cell lines showed a strong affinity for rBC2LCN. In the xenograft model, the tumor volume in the rBC2LCN-PE38 group was significantly reduced compared with that using control treatment alone. However, the HT-29 cell line showed weak affinity and poor therapeutic efficacy. No significant toxicities or adverse responses were observed. In conclusion, we demonstrated that rBC2LCN lectin binds CRC cells and that rBC2LCN-PE38 significantly suppresses tumor growth in vivo. In addition, the efficacy of the drug conjugate correlated with its binding affinity for each CRC cell line. These results suggest that lectin drug conjugate therapy has potential as a novel targeted therapy for CRC cell surface glycans.

Real-time automatic surgical phase recognition in laparoscopic sigmoidectomy using the convolutional neural network-based deep learning approach.

BACKGROUND: Automatic surgical workflow recognition is a key component for developing the context-aware computer-assisted surgery (CA-CAS) systems. However, automatic surgical phase recognition focused on colorectal surgery has not been reported. We aimed to develop a deep learning model for automatic surgical phase recognition based on laparoscopic sigmoidectomy (Lap-S) videos, which could be used for real-time phase recognition, and to clarify the accuracies of the automatic surgical phase and action recognitions using visual information. METHODS: The dataset used contained 71 cases of Lap-S. The video data were divided into frame units every 1/30 s as static images. Every Lap-S video was manually divided into 11 surgical phases (Phases 0-10) and manually annotated for each surgical action on every frame. The model was generated based on the training data. Validation of the model was performed on a set of unseen test data. Convolutional neural network (CNN)-based deep learning was also used. RESULTS: The average surgical time was 175 min (± 43 min SD), with the individual surgical phases also showing high variations in the duration between cases. Each surgery started in the first phase (Phase 0) and ended in the last phase (Phase 10), and phase transitions occurred 14 (± 2 SD) times per procedure on an average. The accuracy of the automatic surgical phase recognition was 91.9% and those for the automatic surgical action recognition of extracorporeal action and irrigation were 89.4% and 82.5%, respectively. Moreover, this system could perform real-time automatic surgical phase recognition at 32 fps. CONCLUSIONS: The CNN-based deep learning approach enabled the recognition of surgical phases and actions in 71 Lap-S cases based on manually annotated data. This system could perform automatic surgical phase recognition and automatic target surgical action recognition with high accuracy. Moreover, this study showed the feasibility of real-time automatic surgical phase recognition with high frame rate.

Exploring optimal examination to detect occult anastomotic leakage after rectal resection in patients with diverting stoma.

BACKGROUND: When considering "early stoma closure", both standardized inclusion/exclusion criteria and standardized methods to assess anastomosis are necessary to reduce the risk of occult anastomotic leakage (AL). However, in the immediate postoperative period, neither have the incidence and risk factors of occult AL in patients with diverting stoma (DS) been clarified nor have methods to assess anastomosis been standardized. The aim of this study was to elucidate the incidence and risk factors of occult AL in patients who had undergone rectal resection with DS and to evaluate the significance of computed tomography (CT) following water-soluble contrast enema (CE) to detect occult anastomotic leakage. METHODS: This was a single institutional prospective observational study of patients who had undergone rectal resection with the selective use of DS between May and October 2019. Fifteen patients had undergone CE and CT to assess for AL on postoperative day (POD) 7, and CT was performed just after CE. Univariate analysis was performed to assess the relationship between preoperative variables and the incidence of occult AL on POD 7. RESULTS: The incidence of occult AL on postoperative day 7 was 6 of 15 (40%). Hand-sewn anastomosis, compared with stapled anastomosis, was a significant risk factor. Five more cases with occult AL that could not be detected with CE could be detected on CT following CE; CE alone had a 33% false-negative radiological result rate. CONCLUSIONS: Hand-sewn anastomosis appeared to be a risk factor for occult AL, and CE alone had a high false-negative radiological result rate. When considering the introduction of early stoma closure, stapled anastomosis and CT following CE could be an appropriate inclusion criterion and preoperative examination, respectively.