Adenoma detection rate metrics in colorectal cancer surveillance colonoscopy

Efficient multi-stage feedback attention for diverse lesion in cancer image segmentation

In the domain of Computer-Aided Diagnosis (CAD) systems, the accurate identification of cancer lesions is paramount, given the life-threatening nature of cancer and the complexities inherent in its manifestation. This task is particularly arduous due to the often vague boundaries of cancerous regions, compounded by the presence of noise and the heterogeneity in the appearance of lesions, making precise segmentation a critical yet challenging endeavor. This study introduces an innovative, an iterative feedback mechanism tailored for the nuanced detection of cancer lesions in a variety of medical imaging modalities, offering a refining phase to adjust detection results. The core of our approach is the elimination of the need for an initial segmentation mask, a common limitation in iterative-based segmentation methods. Instead, we utilize a novel system where the feedback for refining segmentation is derived directly from the encoder-decoder architecture of our neural network model. This shift allows for more dynamic and accurate lesion identification. To further enhance the accuracy of our CAD system, we employ a multi-scale feedback attention mechanism to guide and refine predicted mask subsequent iterations. In parallel, we introduce a sophisticated weighted feedback loss function. This function synergistically combines global and iteration-specific loss considerations, thereby refining parameter estimation and improving the overall precision of the segmentation. We conducted comprehensive experiments across three distinct categories of medical imaging: colonoscopy, ultrasonography, and dermoscopic images. The experimental results demonstrate that our method not only competes favorably with but also surpasses current state-of-the-art methods in various scenarios, including both standard and challenging out-of-domain tasks. This evidences the robustness and versatility of our approach in accurately identifying cancer lesions across a spectrum of medical imaging contexts. Our source code can be found at https://github.com/dewamsa/EfficientFeedbackNetwork.

RA-DENet: Reverse Attention and Distractions Elimination Network for polyp segmentation

To address the problems of polyps of different shapes, sizes, and colors, low-contrast polyps, various noise distractions, and blurred edges on colonoscopy, we propose the Reverse Attention and Distraction Elimination Network, which includes Improved Reverse Attention, Distraction Elimination, and Feature Enhancement. First, we input the images in the polyp image set, and use the five levels polyp features and the global polyp feature extracted from the Res2Net-based backbone as the input of the Improved Reverse Attention to obtain augmented representations of salient and non-salient regions to capture the different shapes of polyp and distinguish low-contrast polyps from background. Then, the augmented representations of salient and non-salient areas are fed into the Distraction Elimination to obtain the refined polyp feature without false positive and false negative distractions for eliminating noises. Finally, the extracted low-level polyp feature is used as the input of the Feature Enhancement to obtain the edge feature for supplementing missing edge information of polyp. The polyp segmentation result is output by connecting the edge feature with the refined polyp feature. The proposed method is evaluated on five polyp datasets and compared with the current polyp segmentation models. Our model improves the mDice to 0.760 on the most challenge dataset (ETIS).

Boundary Constraint Network with Cross Layer Feature Integration for Polyp Segmentation

Clinically, proper polyp localization in endoscopy images plays a vital role in the follow-up treatment (e.g., surgical planning). Deep convolutional neural networks (CNNs) provide a favoured prospect for automatic polyp segmentation and evade the limitations of visual inspection, e.g., subjectivity and overwork. However, most existing CNNs-based methods often provide unsatisfactory segmentation performance. In this paper, we propose a novel boundary constraint network, namely BCNet, for accurate polyp segmentation. The success of BCNet benefits from integrating cross-level context information and leveraging edge information. Specifically, to avoid the drawbacks caused by simple feature addition or concentration, BCNet applies a cross-layer feature integration strategy (CFIS) in fusing the features of the top-three highest layers, yielding a better performance. CFIS consists of three attention-driven cross-layer feature interaction modules (ACFIMs) and two global feature integration modules (GFIMs). ACFIM adaptively fuses the context information of the top-three highest layers via the self-attention mechanism instead of direct addition or concentration. GFIM integrates the fused information across layers with the guidance from global attention. To obtain accurate boundaries, BCNet introduces a bilateral boundary extraction module that explores the polyp and non-polyp information of the shallow layer collaboratively based on the high-level location information and boundary supervision. Through joint supervision of the polyp area and boundary, BCNet is able to get more accurate polyp masks. Experimental results on three public datasets show that the proposed BCNet outperforms seven state-of-the-art competing methods in terms of both effectiveness and generalization.

Effect of oral simethicone on the quality of colonoscopy: A systematic review and meta-analysis of randomized controlled trials

OBJECTIVE

In this systematic review and meta-analysis, we aimed to investigate the effect of oral simethicone (SIM), an antifoaming agent, on the quality of colonoscopy in terms of bowel preparation quality, adenoma or polyp detection rate (ADR/PDR) and cecal intubation rate (CIR).

METHODS

All randomized controlled trials (RCTs) on the use of SIM during bowel preparation for colonoscopy published up to 17 March 2021 were identified from the PubMed, EMBASE and Cochrane Library databases. Bowel preparation quality, ADR/PDR/CIR, cecal intubation time (CIT), withdrawal time (WT), patients' tolerability, acceptability and volume of foam and bubbles were compared between the SIM and non-SIM groups.

RESULTS

Thirty-eight RCTs with 10 505 patients were included. Oral SIM significantly increased the rate of total Boston bowel preparation scale (BBPS) score ≥6 (risk ratio [RR]  1.13, P < 0.0001), acceptability (RR  1.15, P = 0.01) and the rate of no or minimal foam and bubbles (RR  1.28, P < 0.00001) and decreased abdominal distension (RR  0.64, P < 0.0001). However, it had no significant impact on overall ADR, overall PDR, CIR, CIT or WT. The rate of total BBPS score ≥6 remained significantly higher in the SIM group when a single-dose laxative regimen or a SIM dosage of ≥320 mg was employed; and ADR, PDR and CIR were significantly increased in the SIM group among colonoscopy clinicians who achieved an ADR <31%, PDR <45% and CIR <96%, respectively.

CONCLUSIONS

Oral SIM can improve bowel preparation quality, especially in patients receiving a SIM dosage of ≥320 mg or a single-dose laxative regimen. SIM may be preferred by junior colonoscopy physicians/trainees with a lower ADR/PDR or CIR.

Colonoscopy quality indicators in patients with and without prior colonic resection: A single-centre prospective comparative study

AIM

Effective colonoscopy relies on meeting rigorous quality control thresholds. Some earlier studies evaluating colonoscopy key performance indicators (KPIs) have excluded patients who have previously undergone colonic resection (i.e., they have a nonintact colon); such patients also deserve high-quality colonoscopy. This study aimed to compare colonoscopy KPIs between patients with nonintact and intact colons.

METHOD

Consecutive colonoscopies performed at Whanganui Hospital (New Zealand) between September 2016 and March 2020 were included. The primary outcome was the caecal or ileal intubation rate (CIIR). Secondary outcomes were the adenoma detection rate (ADR), polyp detection rate (PDR), colonoscope withdrawal time (CWT) and caecal or ileal intubation time (CIIT).

RESULTS

In total, 3017 colonoscopies were performed: 322 in nonintact colons and 2695 in intact colons. CIIR was significantly higher in nonintact than in intact colons (98.4% vs. 95.0%; P = 0.0086). When all colonoscopies were included, the CIIR was 95.4%; this value decreased to 95.0% when nonintact colonoscopies were excluded. However, the ADR (39.9% vs. 38.8%; P = 0.77) and PDR (58.4% vs. 59.1%; P = 0.86) were similar for both nonintact and intact colons. CWT (P < 0.0001) and CIIT (P < 0.0001) were significantly shorter in participants with nonintact colons.

CONCLUSION

The CIIR exceeded recommended targets and was 3.4% higher in patients with nonintact than intact colons. Patients with nonintact colons comprise a small proportion of the overall colonoscopy cohort and it is unlikely that this small difference is relevant for most endoscopists or endoscopy units. The ADR and PDR were similar among patients with nonintact and intact colons, despite nonintact colonoscopies being significantly quicker. Patients with nonintact colons deserve high-quality colonoscopy and therefore their KPIs should be included in colonoscopy performance reports.

The incidence and risk factors of sessile serrated adenomas in left side colon cancer patients after curative surgery

Sessile serrated adenomas (SSAs) are precursors of colorectal cancer (CRC). However, there are limited data on detection rates of this premalignant lesion during colonoscopy surveillance in patients with a history of left side colonic resection for cancer. We aimed to identify the incidence and risk factors of SSAs in post-left side colectomy patients.We retrospectively reviewed the medical records of patients who had undergone left side colectomy for colon and rectal cancer between September 2009 and September 2016 and had at least 1 follow-up colonoscopy. Patient baseline characteristics, SSA diagnoses and characteristics, and colonoscopy information were collected.In total, 539 patients were enrolled. At the first follow-up (mean duration 11.5 months), 98 SSAs were identified (22.2%). At the second follow-up (mean duration 25.8 months), 51 SSAs were identified in 212 patients (24.0%). Multivariate analysis showed that alcohol intake (hazard ratio [HR] 1.524; 95% confidence interval [CI] .963-2.411, P = .041), excellent bowel preparation (HR 2.081; 95% CI 1.214-3.567, P = .049), and use of a transparent cap (HR 1.702; 95% CI 1.060-2.735, P = .013) were associated with higher SSA incidence in the first surveillance colonoscopy, while body mass index (BMI) ≥ 25.0 (HR 1.602; 95% CI 1.060-2.836) was associated with a significantly increased risk of SSAs in the second surveillance.Considering the endoscopic appearance of SSAs, adequate bowel preparation and use of transparent caps during postoperative surveillance colonoscopy can increase the diagnosis rate. Modification of alcohol intake and BMI may reduce the incidence of SSAs in left side colon cancer patients.

Serum piRNA-54265 is a New Biomarker for early detection and clinical surveillance of Human Colorectal Cancer

Background: Our previous study has demonstrated an oncogenic role of PIWI-interacting RNA-54265 (piR-54265) in colorectal cancer (CRC). Here, we investigate whether it can be a blood biomarker for population screening and clinical applications.

Methods: Serum piR-54265 levels were determined by a digital PCR method in 209 cancer-free healthy controls, 725 patients with CRC, 1303 patients with other types of digestive cancer and 192 patients with benign colorectal tumors. A prospective case-control analysis was conducted to assess the predictive value of serum piR-54265 for future CRC diagnosis. Receiver operating characteristic (ROC) curve was constructed to quantify the diagnostic performance of serum piR-54265 levels by assessing its sensitivity, specificity and respective areas under curve (AUC). The odds ratios (ORs) were computed using multivariate logistic regression models.

Results: Serum piR-54265 levels were significantly elevated only in patients with CRC compared with controls and patients with other cancer types. The AUC for recognizing CRC was 0.896 (95% CI, 0.874-0.914), with a sensitivity and specificity being 85.7% and 65.1% at 1500 copies/µL as a cut-off value. The serum piR-54265 levels in patients declined substantially after surgery but increased significantly again when tumor relapses. The prediagnostic serum piR-54265 levels were significantly associated with future CRC diagnosis, with the ORs of 7.23, 2.80, 2.45, and 1.24 for those whose CRC was diagnosed within 1, 2, 3 and >3 years. Serum piR-54265 test is more sensitive than other blood CRC markers.

Conclusion: Serum piR-54265 may serve as a valuable biomarker for CRC screening, early detection and clinical surveillance.

Quality of Preoperative Colonoscopy Affects Missed Postoperative Adenoma Detection in Colorectal Cancer Patients

BACKGROUND

Despite thorough preoperative screening, 19-30% of synchronous polyps or adenomas are detected after colon cancer surgery. Remnant synchronous lesions require additional colonoscopy procedures or surgery.

AIM

To investigate factors of preoperative colonoscopy potentially affecting the detection of missed lesions in patients subjected to colorectal cancer surgery.

METHODS

Of 1147 colorectal cancer patients subjected to curative open or laparoscopic colectomy and colonoscopy at the Chungnam National University Hospital from January 2012 to December 2016, 518 patients underwent pre- and postoperative colonoscopy. The index colonoscopy was defined as the last preoperative endoscopy performed. We analyzed pre- and postoperative medical charts for colonoscopy and pathological data. The effects of patient, procedure, and tumor factors on the postoperative adenoma detection rate, advanced adenoma detection rate, and adenoma miss rate (AMR) were analyzed.

RESULTS

The overall AMR was 25.7% (95% confidence interval, 22.2-29.8%). Comparing optimal and non-optimal bowel preparation groups, the latter had greater postoperative polyp missed rate (PMR), AMR (p < 0.01), and AAMR (p = 0.272). The optimal preparation group allowed identification of more synchronous adenomas than in the fair (OR 5.72) and poor (OR 11.39) preparation groups. On univariate analysis, patient age and left-sided colectomy (p < 0.01) influenced AMR. Multivariate analysis showed that age, preoperative bowel preparation, and left colon resection influenced postoperative AMR.

CONCLUSION

A better quality of index colonoscopy had a positive effect on lowering the detection rate of postoperative adenoma. Older age and suboptimal bowel preparation at the index colonoscopy and left-sided colectomy had negative effects on lowering the postoperative AMR.