Colorectal cancer is a major health problem, where advances towards computer-aided diagnosis (CAD) systems to assist the endoscopist can be a promising path to improvement. Here, a deep learning model for real-time polyp detection based on a pre-trained YOLOv3 (You Only Look Once) architecture and complemented with a post-processing step based on an object-tracking algorithm to reduce false positives is reported. The base YOLOv3 network was fine-tuned using a dataset composed of 28,576 images labelled with locations of 941 polyps that will be made public soon. In a frame-based evaluation using isolated images containing polyps, a general F₁ score of 0.88 was achieved (recall = 0.87, precision = 0.89), with lower predictive performance in flat polyps, but higher for sessile, and pedunculated morphologies, as well as with the usage of narrow band imaging, whereas polyp size < 5 mm does not seem to have significant impact. In a polyp-based evaluation using polyp and normal mucosa videos, with a positive criterion defined as the presence of at least one 50-frames-length (window size) segment with a ratio of 75% of frames with predicted bounding boxes (frames positivity), 72.61% of sensitivity (95% CI 68.99–75.95) and 83.04% of specificity (95% CI 76.70–87.92) were achieved (Youden = 0.55, diagnostic odds ratio (DOR) = 12.98). When the positive criterion is less stringent (window size = 25, frames positivity = 50%), sensitivity reaches around 90% (sensitivity = 89.91%, 95% CI 87.20–91.94; specificity = 54.97%, 95% CI 47.49–62.24; Youden = 0.45; DOR = 10.76). The object-tracking algorithm has demonstrated a significant improvement in specificity whereas maintaining sensitivity, as well as a marginal impact on computational performance. These results suggest that the model could be effectively integrated into a CAD system.

We aimed to assess the risk of cancer in patients with abdominal symptoms after a complete colonoscopy without colorectal cancer (CRC), according to the carcinoembryonic antigen (CEA) concentration, as well as its diagnostic accuracy. For this purpose, we performed a post-hoc analysis within a cohort of 1431 patients from the COLONPREDICT study, prospectively designed to assess the fecal immunochemical test accuracy in detecting CRC. Over 36.5 ± 8.4 months, cancer was detected in 115 (8%) patients. Patients with CEA values higher than 3 ng/mL revealed an increased risk of cancer (HR 2.0, 95% CI 1.3–3.1), CRC (HR 4.4, 95% CI 1.1–17.7) and non-gastrointestinal cancer (HR 1.7, 95% CI 1.0–2.8). A new malignancy was detected in 51 (3.6%) patients during the first year and three variables were independently associated: anemia (OR 2.8, 95% CI 1.3–5.8), rectal bleeding (OR 0.3, 95% CI 0.1–0.7) and CEA level >3 ng/mL (OR 3.4, 95% CI 1.7–7.1). However, CEA was increased only in 31.8% (95% CI, 16.4–52.7%) and 50% (95% CI, 25.4–74.6%) of patients with and without anemia, respectively, who would be diagnosed with cancer during the first year of follow-up. On the basis of this information, CEA should not be used to assist in the triage of patients presenting with lower bowel symptoms who have recently been ruled out a CRC.

Background: PolyDeep is a computer-aided detection and characterization system that has demonstrated a high diagnostic yield for in vitro detection of colorectal polyps. Our objective is to compare the diagnostic performance of expert endoscopists and PolyDeep for colorectal polyp detection. Methods: PolyDeep Advance 1 (NCT05514301) is an unicentric diagnostic test study with a second observer design. Endoscopists performed colonoscopy blinded to PolyDeep’s detection results. The main endpoint was the sensitivity for colorectal polyp (adenoma, serrated or hyperplastic lesion) detection. The secondary endpoints were the diagnostic performance for diminutive lesions (≤5 mm), neoplasia (adenoma, serrated lesion) and adenoma detection. Results: We included 205 patients (55.1% male, 63.0 ± 6.2 years of age) referred to colonoscopy (positive faecal immunochemical occult blood test = 60.5%, surveillance colonoscopy = 39.5%). We excluded eight patients due to incomplete colonoscopy. Endoscopists detected 384 lesions, of which 39 were not detected by PolyDeep. In contrast, PolyDeep predicted 410 possible additional lesions, 26 of these predictions confirmed by endoscopists as lesions, resulting in a potential 6.8% detection increase with respect to the 384 lesions detected by the endoscopists. In total, 410 lesions were detected, 20 were not retrieved, five were colorectal adenocarcinoma, 343 were colorectal polyps (231 adenomas, 39 serrated and 73 hyperplastic polyps), 42 were normal mucosa and 289 were ≤5 mm. We did not find statistically significant differences between endoscopists and PolyDeep for colorectal polyp detection (Sensitivity = 94.2%, 91.5%, p = 0.2; Specificity = 9.5%, 14.3%, p = 0.7), diminutive lesions (Sensitivity = 92.3%, 89.5%, p = 0.4; Specificity = 9.8%, 14.6%, p = 0.7), neoplasia (Sensitivity = 95.2%, 92.9%, p = 0.3; Specificity = 9.6%, 13.9%, p = 0.4) and adenoma detection (Sensitivity = 94.4%, 92.6%, p = 0.5; Specificity = 7.2%, 11.8%, p = 0.2). Conclusions: Expert endoscopists and PolyDeep have similar diagnostic performance for colorectal polyp detection.

Background/Objectives: Computer-aided detection (CADe) systems are increasingly used in endoscopy to enhance lesion recognition. PolyDeep is a CADe/x tool previously assessed in an observational study. The aim of our study is to determine if PolyDeep-assisted colonoscopy reduces the adenoma miss rate (AMR) compared with conventional colonoscopy. Methods: We carried out a multicenter randomized controlled trial with a tandem colonoscopy design in participants from a colorectal cancer screening program (positive fecal immunochemical test-FIT or surveillance). Expert endoscopists performed all colonoscopies, and patients were allocated to groups by a computer-generated sequence. The primary endpoint was AMR; secondary endpoints included polyp miss rate (PMR), serrated lesion miss rate (SLMR) and advanced polyp miss rate (APMR). Results: From May to November 2023, we recruited 260 patients and excluded 20, leaving 240 for analysis. Baseline characteristics were balanced between groups (62.1% male; mean age 62.3 ± 6.5 years; 65.8% FIT-positive; mean first withdrawal time 13:38 ± 08:07 min; mean second withdrawal time 07:50 ± 03:38 min; lesion detection rate 76.6%; mean polyps per patient 3.4 ± 3.1). We did not find statistically significant differences between PolyDeep-assisted and conventional colonoscopy groups in AMR (21.3% vs. 18.1%, p = 0.5), PMR (21.8% vs. 20.3%, p = 0.7), SLMR (23.4% vs. 25.6%, p = 0.9) or APMR (7.3% vs. 11.3%, p = 0.5). In the subgroup analysis according to indication, we did not find any statistically significant differences. Conclusions: In the context of a CRC screening program, PolyDeep-assisted colonoscopy did not reduce AMR.

Background: Most colorectal cancer patients diagnosed are candidates for surgical resection with curative intent, although colorectal surgery is associated with some complications that could be life-threatening. Antibiotic prophylaxis is commonly used for the prevention of infectious postoperative complications. However, this intervention can change the composition of intestinal microbiota and promote adverse inflammatory outcomes in colorectal cancer patients. The combination of different fungal extracts could be beneficial because of their role in gut microbiota modulation and their anti-inflammatory activity. Objective: Based on this hypothesis, we have designed a double-bind, randomized clinical trial to evaluate the effect of the nutraceutical fungal extract Micodigest 2.0 on complications of surgery for colorectal cancer resection. Methods: Colorectal cancer candidates for surgery will be considered for inclusion in the study. After evaluation by the multidisciplinary tumor board, patients who meet selection criteria will be screened, stratified according to tumor location, and randomly allocated to be treated with Micodigest 2.0 or placebo. Treatment will be continued until admission for surgery (4-6 weeks). Participants will undergo a medical and clinical evaluation including baseline and preadmission quality of life, microbiome composition, inflammatory and nutritional status, adverse events, and adherence assessments. The main end point of the study is the surgery complication rate. We will evaluate complications using the Clavien-Dindo classification. It will be necessary to recruit 144 patients to find a relevant clinical difference. We will also evaluate the effect of the nutraceutical on microbiome composition, inflammatory response, nutritional status, and quality of life, as well as the effect of these variables on surgical complications. Results: This study was funded in 2020 by the Center for Industrial Technology Development. Recruitment began in September 2021 and is expected to be completed in September 2022. Data will be analyzed and the results will be disseminated in 2023. Conclusions: The results of this protocol study could help to reduce surgery complications in patients with colorectal cancer using the new treatment Micodigest. This study could also identify new features associated with colorectal surgery complications. In summary, this study trial could improve the management of colorectal cancer patients. Trial Registration: Clinical Trials.gov NCT04821258; https://clinicaltrials.gov/ct2/show/NCT04821258 International Registered Report Identifier (IRRID): DERR1-10.2196/34292