Search

We have designed a deep-learning model, an “Artificial Intelligent Endoscopist (a.k.a. AI-doscopist)”, to localise colonic neoplasia during colonoscopy. This study aims to evaluate the agreement between endoscopists and AI-doscopist for colorectal neoplasm localisation. AI-doscopist was pre-trained by 1.2 million non-medical images and fine-tuned by 291,090 colonoscopy and non-medical images. The colonoscopy images were obtained from six databases, where the colonoscopy images were classified into 13 categories and the polyps’ locations were marked image-by-image by the smallest bounding boxes. Seven categories of non-medical images, which were believed to share some common features with colorectal polyps, were downloaded from an online search engine. Written informed consent were obtained from 144 patients who underwent colonoscopy and their full colonoscopy videos were prospectively recorded for evaluation. A total of 128 suspicious lesions were resected or biopsied for histological confirmation. When evaluated image-by-image on the 144 full colonoscopies, the specificity of AI-doscopist was 93.3%. AI-doscopist were able to localise 124 out of 128 polyps (polyp-based sensitivity = 96.9%). Furthermore, after reviewing the suspected regions highlighted by AI-doscopist in a 102-patient cohort, an endoscopist has high confidence in recognizing four missed polyps in three patients who were not diagnosed with any lesion during their original colonoscopies. In summary, AI-doscopist can localise 96.9% of the polyps resected by the endoscopists. If AI-doscopist were to be used in real-time, it can potentially assist endoscopists in detecting one more patient with polyp in every 20–33 colonoscopies.

Background Granulin epithelin precursor (GEP) is reported to function as a growth factor stimulating proliferation and migration, and conferring chemoresistance in many cancer types. However, the expression and functional roles of GEP in colorectal cancer (CRC) remain elusive. The aim of this study was thus to investigate the clinical significance of GEP in CRC and reveal the molecular mechanism of GEP in CRC initiation and progression. Methods The mRNA expression of GEP in CRC cell lines were detected by qRT-PCR. The GEP protein expression was validated by immunohistochemistry in tissue microarray (TMA) including 190 CRC patient samples. The clinicopathological correlation analysis were achieved by GEP expression on TMA. Functional roles of GEP were determined by MTT proliferation, monolayer colony formation, cell invasion and migration and in vivo studies through siRNA/shRNA mediated knockdown assays. The cancer signaling pathway identification was acquired by flow cytometry, western blot and luciferase activity assays. Results The mRNA expression of GEP in CRC was significantly higher than it in normal colon tissues. GEP protein was predominantly localized in the cytoplasm and most of the CRC cases demonstrated abundant GEP protein compared with non-tumorous tissues. GEP overexpression was associated with non-rectal location, advanced AJCC stage, regional lymph node and distant metastasis. By Kaplan–Meier survival analysis, GEP abundance served as a prognostic marker for worse survival in CRC patients. GEP knockdown exhibited anti-cancer effect such as inhibiting cell proliferation, monolayer colony formation, cell invasion and migration in DLD-1 and HCT 116 cells and decelerating xenograft formation in nude mice. siGEP also induced G1 cell cycle arrest and apoptosis. Luciferase activity assays further demonstrated GEP activation was involved in MAPK/ERK signaling pathway. Conclusion In summary, we compressively delineate the oncogenic role of GEP in colorectal tumorigenesis by activating MAPK/ERK signaling pathway. GEP might serve as a useful prognostic biomarker and therapeutic target for CRC. Electronic supplementary material The online version of this article (10.1186/s12967-018-1530-7) contains supplementary material, which is available to authorized users.

Colorectal cancer (CRC) is a leading cause of cancer deaths worldwide. Although polypectomy at early stage reduces CRC incidence, 90% of the polyps are small and diminutive, where removal of them poses risks to patients that may outweigh the benefits. Correctly detecting and predicting polyp type during colonoscopy allows endoscopists to resect and discard the tissue without submitting it for histology, saving time, and costs. Nevertheless, human visual observation of early stage polyps varies. Therefore, this paper aims at developing a fully automatic algorithm to detect and classify hyperplastic and adenomatous colorectal polyps. Adenomatous polyps should be removed, whereas distal diminutive hyperplastic polyps are considered clinically insignificant and may be left in situ . A novel transfer learning application is proposed utilizing features learned from big nonmedical datasets with 1.4-2.5 million images using deep convolutional neural network. The endoscopic images we collected for experiment were taken under random lighting conditions, zooming and optical magnification, including 1104 endoscopic nonpolyp images taken under both white-light and narrowband imaging (NBI) endoscopy and 826 NBI endoscopic polyp images, of which 263 images were hyperplasia and 563 were adenoma as confirmed by histology. The proposed method identified polyp images from nonpolyp images in the beginning followed by predicting the polyp histology. When compared with visual inspection by endoscopists, the results of this study show that the proposed method has similar precision (87.3% versus 86.4%) but a higher recall rate (87.6% versus 77.0%) and a higher accuracy (85.9% versus 74.3%). In conclusion, automatic algorithms can assist endoscopists in identifying polyps that are adenomatous but have been incorrectly judged as hyperplasia and, therefore, enable timely resection of these polyps at an early stage before they develop into invasive cancer.

Background: Overactive bladder is prevalent in adults over 40 years of age and its prevalence increases with age. The use of acupuncture in adults with overactive bladder is increasing globally. However, its effectiveness/efficacy and safety have not yet been confirmed. Objective: To determine the effectiveness/efficacy and safety of acupuncture and to identify the pattern of commonly used traditional acupuncture points in the management of overactive bladder in adults. Methods: Ten electronic English and Chinese language databases were systematically searched and two English together with four Chinese journals relevant to acupuncture were manually searched in libraries for randomised controlled trials (RCTs) of acupuncture for overactive bladder in adults from their inception to March 2017. The Cochrane risk of bias tool was used to assess the methodological quality of the RCTs. RevMan v.5.3 software was employed for data analysis. Results: Seven eligible trials involving 695 participants were included. Meta-analysis showed that acupuncture was comparable to drugs (tolterodine tartrate/solifenacin) in the reduction of micturition episodes over 24 hours (pooled standardised mean difference (SMD) 0.36, 95% CI −0.23 to 0.95; I2=83%), increase in voided volume of each micturition episode (pooled SMD −0.15, 95% CI −0.36 to 0.05; I2=0%), and reduction of overactive bladder symptom score (pooled SMD −0.23, 95% CI −1.30 to 0.85; I2=91%). In addition, acupuncture was not significantly different compared with placebo in the reduction of overactive bladder symptom score (pooled SMD −2.36, 95% CI −5.64 to 0.93; I2=97%). No serious adverse events were reported. Conclusions: No significant differences in effectiveness or efficacy were found between acupuncture and drug or between verum and sham acupuncture, respectively. Further high-quality studies are required.

Psychometric evaluation with tools such as the faecal incontinence quality of life scale is an essential component of clinical assessment. Currently, there is no translated Traditional Chinese version of the scale. A validated translated version may help to improve the quality of healthcare received in countries with Chinese minorities (0.5, 1.2 and 4.0 % of the UK, USA and Australian population, respectively) as well as local population of Hong Kong. The purpose of this study is to validate the Traditional Chinese version of the faecal incontinence quality of life scale (FIQL). The FIQL questionnaire was translated into Traditional Chinese Language followed by linguistic validation. It was then tested on 55 patients with faecal incontinence and 93 controls in the colorectal outpatients clinic. Faecal incontinence severity index was also used for the assessment of disease severity. Internal consistency was good/excellent for all scales (Cronbach’s alpha >0.70, between 0.71 and 0.93). The intra-class correlation indicated a high stability over time with coefficients ranging between 0.78 and 0.90. Test and retest of all four scales found no significant differences of mean scores between baseline and retest. The mean faecal incontinence quality of life scale scores of all four domains improved significantly after treatment of 10 patients whose faecal incontinence severity index scores decreased by 50 % of their pre-treatment scores, hence indicating good sensitivity. This study demonstrates the linguistic and psychometric validity of the traditional Chinese version of the faecal incontinence quality of life scale.

Background Patients with colorectal cancer (CRC) have a high incidence of regional and distant metastases. Although metastasis is the main cause of CRC-related death, its molecular mechanisms remain largely unknown. Methods Using array-CGH and expression microarray analyses, changes in DNA copy number and mRNA expression levels were investigated in human CRC samples. The mRNA expression level of RASAL2 was validated by qRT-PCR, and the protein expression was evaluated by western blot as well as immunohistochemistry in CRC cell lines and primary tumors. The functional role of RASAL2 in CRC was determined by MTT proliferation assay, monolayer and soft agar colony formation assays, cell cycle analysis, cell invasion and migration and in vivo study through siRNA/shRNA mediated knockdown and overexpression assays. Identification of RASAL2 involved in hippo pathway was achieved by expression microarray screening, double immunofluorescence staining and co-immunoprecipitation assays. Results Integrated genomic analysis identified copy number gains and upregulation of RASAL2 in metastatic CRC. RASAL2 encodes a RAS-GTPase-activating protein (RAS-GAP) and showed increased expression in CRC cell lines and clinical specimens. Higher RASAL2 expression was significantly correlated with lymph node involvement and distant metastasis in CRC patients. Moreover, we found that RASAL2 serves as an independent prognostic marker of overall survival in CRC patients. In vitro and in vivo functional studies revealed that RASAL2 promoted tumor progression in both KRAS/NRAS mutant and wild-type CRC cells. Knockdown of RASAL2 promoted YAP1 phosphorylation, cytoplasm retention and ubiquitination, therefore activating the hippo pathway through the LATS2/YAP1 axis. Conclusions Our findings demonstrated the roles of RASAL2 in CRC tumorigenesis as well as metastasis, and RASAL2 exerts its oncogenic property through LATS2/YAP1 axis of hippo signaling pathway in CRC. Electronic supplementary material The online version of this article (10.1186/s12943-018-0853-6) contains supplementary material, which is available to authorized users.

Algorithms for localising colorectal polyps have been studied extensively; however, they were often trained and tested using the same database. In this study, we present a new application of a unified and real-time object detector based on You-Only-Look-Once (YOLO) convolutional neural network (CNN) for localizing polyps with bounding boxes in endoscopic images. The model was first pre-trained with non-medical images and then fine-tuned with colonoscopic images from three different databases, including an image set we collected from 106 patients using narrow-band (NB) imaging endoscopy. YOLO was tested on 196 white light (WL) images of an independent public database. YOLO achieved a precision of 79.3% and sensitivity of 68.3% with time efficiency of 0.06 sec/frame in the localization task when trained by augmented images from multiple WL databases. In conclusion, YOLO has great potential to be used to assist endoscopists in localising colorectal polyps during endoscopy. CNN features of WL and NB endoscopic images are different and should be considered separately. A large-scale database that covers different scenarios, imaging modalities and scales is lacking but crucial in order to bring this research into reality.

Body sensors are now commonly used as ingestible, wearable and implantable devices for clinical diagnosis and continuous physiological monitoring. Nevertheless, they usually have limited resources. Recent advancements in technologies provide a possible solution to mitigate the resource limitation of these devices by connecting them with mobile devices and cloud services. To evaluate the feasibility of the cloud-enabled body sensor networks, this paper presents simulation results on testing the feasibility of 24-hour operating time and concurrent user support for the cloud-enabled applications.

KRAS mutational status has been shown to be a predictive biomarker of resistance to anti-EGFR monoclonal antibody (mAb) therapy in patients with metastatic colorectal cancer. We report the spectrum of KRAS mutation in 1506 patients with colorectal cancer and the identification and characterization of rare insertion mutations within the functional domain of KRAS. KRAS mutations are found in 44.5% (670/1506) of the patients. Two cases are found to harbor double mutations involving both codons 12 and 13. The frequencies of KRAS mutations at its codons 12, 13, 61, and 146 are 75.1%, 19.3%, 2.5%, and 2.7%, respectively. The most abundant mutation of codon 12 is G12D, followed by G12V and G12C while G13D is the predominant mutation in codon 13. Mutations in other codons are rare. The KRAS mutation rate is significantly higher in women (48%, 296/617) than in men (42.1%, 374/889, P = 0.023). Tumors on the right colon have a higher frequency of KRAS mutations than those on the left (57.3% vs. 40.4%, P < 0.0001). Two in-frame insertion mutations affect the phosphate-binding loop (codon 10–16) of KRAS are identified. One of them has never been reported before. Compared with wild-type protein, the insertion variants enhance the cellular accumulation of active RAS (RAS-GTP) and constitutively activate the downstream signaling pathway. NIH3T3 cells transfected with the insertion variants show enhanced anchorage-independent growth and in vivo tumorigenicity. Potentially these mutations contribute to primary resistance to anti-EGFR mAb therapy but the clinical implication requires further validation.

Objective To compare long-term oncologic outcomes between laparoscopic and open surgery for rectal cancer and to identify independent predictors of survival. Background Few randomized trials comparing laparoscopic and open surgery for rectal cancer have reported long-term survival data. Methods Data from the 3 randomized controlled trials comparing curative laparoscopic (n=136) and open surgery (n=142) for upper, mid, and low rectal cancer conducted at the Prince of Wales Hospital, Hong Kong, between September 1993 and August 2007 were pooled together for this analysis. Survival and disease status were updated to February 2012. Survival was calculated using the Kaplan-Meier method, and independent predictors of survival were determined using the Cox regression analysis. Results The demographic data of the 2 groups were comparable. The median follow-up time of living patients was 124.5 months in the laparoscopic group and 136.6 months in the open group. At 10 years, there were no significant differences in locoregional recurrence (5.5% vs. 9.3%; P=0.296), cancer-specific survival (82.5% vs. 77.6%; P=0.443), and overall survival (63.0% vs. 61.1%; P=0.505) between the laparoscopic and open groups. There was a trend toward lower recurrence rate at 10 years in the laparoscopic group than in the open group among patients with stage III cancer (P=0.078). The Cox regression analysis showed that stage III cancer, lymphovascular permeation, and blood transfusion, but not the operative approach, were independent predictors of poorer cancer-specific survival. Conclusions This pooled analysis with a follow-up of more than 10 years confirms the long-term oncologic safety of laparoscopic surgery for rectal cancer.

The aim of this study was to evaluate a structured training programme for laparoscopic colorectal surgery in a university colorectal unit over a 6-year period. Data on patients who underwent laparoscopic colectomy between November 2004 and October 2010 were analyzed. Operations were performed either by the consultant colorectal surgeons or colorectal fellows. The effectiveness and safety of our structured training programme were evaluated. During the study period, 813 patients (478 men) with a median age 69 years (range 22–93) underwent laparoscopic colectomy. A total of 370 cases (45.5 %) were performed by four colorectal fellows. Overall, 674 patients (82.9 %) were classified as ASA I or II. The conversion rate was 3.7 %. The conversion rate, intra-operative blood loss, number of lymph nodes retrieved and post-operative recovery were similar between the two groups. When comparing with consultant group, the patients operated by fellows were: (1) significantly older; (2) more were operated on as emergency cases; (3) had pathologically less advanced tumours; (4) less patients with low rectal cancers. There were two surgical mortalities in this series. The morbidities between the two groups were similar. At the end of 3 years of training, the fellows had performed more than 85 cases of laparoscopic colectomies. The level of supervision decreased with increased experience. Finally, experienced fellows were able to supervise more junior colleagues on laparoscopic colectomies. Our results confirmed a structured training programme for laparoscopic colectomy is safe and effective. Reasonable results were achieved even though a high volume of cases were performed by surgical fellows.

Colorectal cancer (CRC) is the third most commonly diagnosed cancer around the world. In the United States, there is an annual incidence of 143,000 cases with approximately 52,000 cancer deaths each year. Colorectal cancer death also represents around 9% of all cancer mortality [1]. The most recent cancer data from Hong Kong has even shown that the annual incidence of colorectal cancer has surpassed the incidence of lung cancer [2] (Fig. 9.1). While population screening for colorectal cancer is currently being developed in Hong Kong, strategies for effective surgical treatments are also required.

This single-center, prospective, randomized trial was designed to compare the short-term clinical outcome between laparoscopic-assisted versus open total mesorectal excision (TME) with anal sphincter preservation (ASP) in patients with mid and low rectal cancer. Long-term morbidity and survival data also were recorded and compared between the two groups. Between August 2001 and August 2007, 80 patients with mid and low rectal cancer were randomized to receive either laparoscopic-assisted (40 patients) or open (40 patients) TME with ASP. The median follow-up time for all patients was 75.7 (range 16.9–115.7) months for the laparoscopic-assisted group and 76.1 (range 4.7–126.6) months for the open group. The primary endpoint of the study was short-term clinical outcome. Secondary endpoints included long-term morbidity rate and survival. Data were analyzed by intention-to-treat principle. The demographic data of the two groups were comparable. Postoperative recovery was better after laparoscopic surgery, with less analgesic requirement (P

We investigated the efficacy of electroacupuncture in reducing the duration of postoperative ileus and hospital stay after laparoscopic surgery for colorectal cancer.We performed a prospective study of 165 patients undergoing elective laparoscopic surgery for colonic and upper rectal cancer, enrolled from October 2008 to October 2010. Patients were assigned randomly to groups that received electroacupuncture (n = 55) or sham acupuncture (n = 55), once daily from postoperative days 1-4, or no acupuncture (n = 55). The acupoints Zusanli, Sanyinjiao, Hegu, and Zhigou were used. The primary outcome was time to defecation. Secondary outcomes included postoperative analgesic requirement, time to ambulation, and length of hospital stay.Patients who received electroacupuncture had a shorter time to defecation than patients who received no acupuncture (85.9 ± 36.1 vs 122.1 ± 53.5 h; P.001) and length of hospital stay (6.5 ± 2.2 vs 8.5 ± 4.8 days; P = .007). Patients who received electroacupuncture also had a shorter time to defecation than patients who received sham acupuncture (85.9 ± 36.1 vs 107.5 ± 46.2 h; P = .007). Electroacupuncture was more effective than no or sham acupuncture in reducing postoperative analgesic requirement and time to ambulation. In multiple linear regression analysis, an absence of complications and electroacupuncture were associated with a shorter duration of postoperative ileus and hospital stay after the surgery.In a clinical trial, electroacupuncture reduced the duration of postoperative ileus, time to ambulation, and postoperative analgesic requirement, compared with no or sham acupuncture, after laparoscopic surgery for colorectal cancer. ClinicalTrials.gov number, NCT00464425.

Endoscopic submucosal dissection (ESD) is a technically demanding procedure, and exposure of the submucosa depends on the action of gravity and submucosal injection. The aim of the study was to investigate the effectiveness of the Endolifter(®) as a traction device for enhancing submucosal visualization during ESD.This was a prospective ex vivo comparative study conducted between September 2010 and March 2011 in the Prince of Wales Hospital. Consecutive ESDs were performed by four experienced endoscopists in an ex vivo ESD model with or without the Endolifter(®). The Endolifter(®) allows simultaneous grasping, retracting and lifting of the mucosa during ESD, resulting in exposure of the submucosa. Each of the procedures were recorded and reviewed later by two independent assessors. The outcome measures included the proportion of time that the submucosa was visualized during the procedures (SM ratio), procedural times, perforation rates, amount of submucosal injections, and the difficulty of the procedure.Forty-eight gastric ESD procedures were performed on the model. The SM ratio was higher in the Endolifter(®) group (P = 0.007), particularly for lesions located at the antrum (P0.001). The time required for submucosal dissection and the total procedural time were also less in the Endolifter(®) group. The endoscopists rated the ESD procedures in the Endolifter(®) group as less difficult (P = 0.033).The Endolifter(®) improved submucosal visualization during gastric ESD and reduces the difficulty of performing the procedures. The device may improve the ease of performing ESD in low-volume centers or large mucosal lesions.

AIM: To give a comprehensive review of current literature on robotic rectal cancer surgery. METHODS: A systematic review of current literature via PubMed and Embase search engines was performed to identify relevant articles from january 2007 to november 2013. The keywords used were: “robotic surgery”, “surgical robotics”, “laparoscopic computer-assisted surgery”, “colectomy” and “rectal resection”. RESULTS: After the initial screen of 380 articles, 20 papers were selected for review. A total of 1062 patients (male 64.0%) with a mean age of 61.1 years and body mass index of 24.9 kg/m2 were included in the review. Out of 1062 robotic-assisted operations, 831 (78.2%) anterior and low anterior resections, 132 (12.4%) intersphincteric resection with coloanal anastomosis, 98 (9.3%) abdominoperineal resections and 1 (0.1%) Hartmann’s operation were included in the review. Robotic rectal surgery was associated with longer operative time but with comparable oncological results and anastomotic leak rate when compared with laparoscopic rectal surgery. CONCLUSION: Robotic colorectal surgery has continued to evolve to its current state with promising results; feasible surgical option with low conversion rate and comparable short-term oncological results. The challenges faced with robotic surgery are for more high quality studies to justify its cost.