Dataset Information

Computer aided detection for laterally spreading tumors and sessile serrated adenomas during colonoscopy.

ABSTRACT: BACKGROUND:Evidence has shown that deep learning computer aided detection (CADe) system achieved high overall detection accuracy for polyp detection during colonoscopy. AIM:The detection performance of CADe system on non-polypoid laterally spreading tumors (LSTs) and sessile serrated adenomas/polyps (SSA/Ps), with higher risk for malignancy transformation and miss rate, has not been exclusively investigated. METHODS:A previously validated deep learning CADe system for polyp detection was tested exclusively on LSTs and SSA/Ps. 1451 LST images from 184 patients were collected between July 2015 and January 2019, 82 SSA/Ps videos from 26 patients were collected between September 2018 and January 2019. The per-frame sensitivity and per-lesion sensitivity were calculated. RESULTS:(1) For LSTs image dataset, the system achieved an overall per-image sensitivity and per-lesion sensitivity of 94.07% (1365/1451) and 98.99% (197/199) respectively. The per-frame sensitivity for LST-G(H), LST-G(M), LST-NG(F), LST-NG(PD) was 93.97% (343/365), 98.72% (692/701), 85.71% (324/378) and 85.71% (6/7) respectively. The per-lesion sensitivity of each subgroup was 100.00% (71/71), 100.00% (64/64), 98.31% (58/59) and 80.00% (4/5). (2) For SSA/Ps video dataset, the system achieved an overall per-frame sensitivity and per-lesion sensitivity of 84.10% (15883/18885) and 100.00% (42/42), respectively. CONCLUSIONS:This study demonstrated that a local-feature-prioritized automatic CADe system could detect LSTs and SSA/Ps with high sensitivity. The per-frame sensitivity for non-granular LSTs and small SSA/Ps should be further improved.

SUBMITTER: Zhou G

PROVIDER: S-EPMC7173785 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Computer aided detection for laterally spreading tumors and sessile serrated adenomas during colonoscopy.

Zhou Guanyu G Xiao Xun X Tu Mengtian M Liu Peixi P Yang Dan D Liu Xiaogang X Zhang Renyi R Li Liangping L Lei Shan S Wang Han H Song Yan Y Wang Pu P

PloS one 20200421 4

<h4>Background</h4>Evidence has shown that deep learning computer aided detection (CADe) system achieved high overall detection accuracy for polyp detection during colonoscopy.<h4>Aim</h4>The detection performance of CADe system on non-polypoid laterally spreading tumors (LSTs) and sessile serrated adenomas/polyps (SSA/Ps), with higher risk for malignancy transformation and miss rate, has not been exclusively investigated.<h4>Methods</h4>A previously validated deep learning CADe system for polyp ...[more]

PMID: 32315365

Similar Datasets

Project description:Background & aimsMany cancers in the proximal colon develop via from sessile serrated adenomas (SSAs), which have flat, subtle features that are difficult to detect with conventional white-light colonoscopy. Many SSA cells have the V600E mutation in BRAF. We investigated whether this feature could be used with imaging methods to detect SSAs in patients.MethodsWe used phage display to identify a peptide that binds specifically to SSAs, using subtractive hybridization with HT29 colorectal cancer cells containing the V600E mutation in BRAF and Hs738.St/Int cells as a control. Binding of fluorescently labeled peptide to colorectal cancer cells was evaluated with confocal fluorescence microscopy. Rats received intra-colonic 0.0086 mg/kg, 0.026 mg/kg, or 0.86 mg/kg peptide or vehicle and morbidity, mortality, and injury were monitored twice daily to assess toxicity. In the clinical safety study, fluorescently labeled peptide was topically administered, using a spray catheter, to the proximal colon of 25 subjects undergoing routine outpatient colonoscopies (3 subjects were given 2.25 μmol/L and 22 patients were given 76.4 μmol/L). We performed blood cell count, chemistry, liver function, and urine analyses approximately 24 hours after peptide administration. In the clinical imaging study, 38 subjects undergoing routine outpatient colonoscopies, at high risk for colorectal cancer, or with a suspected unresected proximal colonic polyp, were first evaluated by white-light endoscopy to identify suspicious regions. The fluorescently labeled peptide (76.4 μmol/L) was administered topically to proximal colon, unbound peptide was washed away, and white-light, reflectance, and fluorescence videos were recorded digitally. Fluorescence intensities of SSAs were compared with those of normal colonic mucosa. Endoscopists resected identified lesions, which were analyzed histologically by gastrointestinal pathologists (reference standard). We also analyzed the ability of the peptide to identify SSAs vs adenomas, hyperplastic polyps, and normal colonic mucosa in specimens obtained from the tissue bank at the University of Michigan.ResultsWe identified the peptide sequence KCCFPAQ and measured an apparent dissociation constant of Kd = 72 nM and an apparent association time constant of K = 0.174 min-1 (5.76 minutes). During fluorescence imaging of patients during endoscopy, regions of SSA had 2.43-fold higher mean fluorescence intensity than that for normal colonic mucosa. Fluorescence labeling distinguished SSAs from normal colonic mucosa with 89% sensitivity and 92% specificity. The peptide had no observed toxic effects in animals or patients. In the analysis of ex vivo specimens, peptide bound to SSAs had significantly higher mean fluorescence intensity than to hyperplastic polyps.ConclusionsWe have identified a fluorescently labeled peptide that has no observed toxic effects in animals or humans and can be used for wide-field imaging of lesions in the proximal colon. It distinguishes SSAs from normal colonic mucosa with 89% sensitivity and 92% specificity. This targeted imaging method might be used in early detection of premalignant serrated lesions during routine colonoscopies. ClinicalTrials.gov ID: NCT02156557.

Project description:Sessile serrated adenomas are now recognized as precursor lesions of a substantial subset of colorectal cancers arising via a so-called “serrated pathway”. However, their biological markers remain to be defined. The aim of our study was to identify differentially expressed genes in sessile serrated adenomas, hyperplastic polyps and tubular adenomas. Gene expression analysis demonstrated molecular differences between polyp types. Further studies using QRT-PCR on Cathepsin E demonstrated a significantly (p< 0.05) higher expression in sessile serrated adenomas as compared to both other polyp types. Trefoil Factor 1, showed the same trend of expression for sessile serrated adenomas as compared to hyperplastic polyps, and was significantly higher in both polyps compared to tubular adenomas. Immunohistochemistry for both proteins demonstrated strong cytoplasmic staining of abnormal crypts in all sessile serrated adenomas while staining in tubular adenomas and hyperplastic polyps was weak and focal. BRAF and KRAS mutation analysis were employed to further validate polyp discrimination. The findings demonstrated the positive association of the BRAF mutation, V600E, with sessile serrated adenomas and KRAS mutations with tubular adenomas (P<0.05). This study demonstrates CTSE and TFF1 over-expression in sessile serrated adenomas compared to both hyperplastic polyps and tubular adenomas. Keywords: colonic polyp tissue comparison, linear modelling, SSA Microarray analysis was performed on 13 SSAs and 11 TAs. SSAs were from 4 males and 9 females (mean age of 75) and TAs were from 5 males and 6 females (mean age of 72). Samples were directly hybridised to each other (SSA versus TA) or to a pooled normal control (SSA versus control). A linear model (Smyth 2004) was fitted to the data bringing together the three contrasts: SSA versus control, TA versus control and SSA versus TA. Human OligoLibrary (Compugen Human Oligo Library (v1) containing 18861 60-mer oligonucleotides, representing approximately 16,000 unique genes) by the Adelaide Microarray Centre (Australia) was used. Slides were scanned using a GenePix 3000B scanner (Axon Instruments, Sunnydale, CA), and the Spot package (CSIRO, Australia) was used to identify spots and estimate fore- and background intensities (using a morphological opening background estimator) (Yang, Buckley et al. 2001; Ritchie 2004). Data analysis was performed in R (www.r-project.org) using the Limma package of Bioconductor (Gentleman, Carey et al. 2004; Smyth 2004). Loess print tip method was used to correct for dye-bias and intensity within each group of adjacent spots printed by one pin (Yang, Dudoit et al. 2002). Linear modelling was performed with the Limma package of Bioconductor (Smyth 2004).

Project description:BACKGROUND : We aimed to compare the accuracy of the optical diagnosis of diminutive colorectal polyps, including sessile serrated lesions (SSLs), between a computer-aided diagnosis (CADx) system and endoscopists during real-time colonoscopy. METHODS : We developed the POLyp Artificial Recognition (POLAR) system, which was capable of performing real-time characterization of diminutive colorectal polyps. For pretraining, the Microsoft-COCO dataset with over 300 000 nonpolyp object images was used. For training, eight hospitals prospectively collected 2637 annotated images from 1339 polyps (i. e. publicly available online POLAR database). For clinical validation, POLAR was tested during colonoscopy in patients with a positive fecal immunochemical test (FIT), and compared with the performance of 20 endoscopists from eight hospitals. Endoscopists were blinded to the POLAR output. Primary outcome was the comparison of accuracy of the optical diagnosis of diminutive colorectal polyps between POLAR and endoscopists (neoplastic [adenomas and SSLs] versus non-neoplastic [hyperplastic polyps]). Histopathology served as the reference standard. RESULTS : During clinical validation, 423 diminutive polyps detected in 194 FIT-positive individuals were included for analysis (300 adenomas, 41 SSLs, 82 hyperplastic polyps). POLAR distinguished neoplastic from non-neoplastic lesions with 79 % accuracy, 89 % sensitivity, and 38 % specificity. The endoscopists achieved 83 % accuracy, 92 % sensitivity, and 44 % specificity. The optical diagnosis accuracy between POLAR and endoscopists was not significantly different (P = 0.10). The proportion of polyps in which POLAR was able to provide an optical diagnosis was 98 % (i. e. success rate). CONCLUSIONS : We developed a CADx system that differentiated neoplastic from non-neoplastic diminutive polyps during endoscopy, with an accuracy comparable to that of screening endoscopists and near-perfect success rate.

Project description:BackgroundSessile serrated adenomas (SSAs) are common polyps which give rise to 20-30% of colorectal cancer (CRC). SSAs display clinicopathologic features which present challenges in surveillance, including overrepresentation in young patients, proclivity for the proximal colon and rarity of histologic dysplasia (referred to then as SSAs with dysplasia, SSADs). Once dysplasia develops, there is rapid progression to CRC, even at a small size. There is therefore a clinical need to separate the "advanced" SSAs at high risk of progression to SSAD and cancer from ordinary SSAs. Since SSAs are known to accumulate methylation over time prior to the development of dysplasia, SSAD backgrounds (the remnant SSA present within an SSAD) likely harbour additional methylation events compared with ordinary SSAs. We therefore performed MethyLight and comprehensive methylation array (Illumina MethylationEPIC) on 40 SSAD backgrounds and 40 matched ordinary SSAs, and compared the methylation results with CRC methylation, CRC expression and immunohistochemical data.ResultsSSAD backgrounds demonstrated significant hypermethylation of CpG islands compared with ordinary SSAs, and the proportion of hypermethylated probes decreased progressively in the shore, shelf and open sea regions. Hypomethylation occurred in concert with hypermethylation, which showed a reverse pattern, increasing progressively away from the island regions. These methylation changes were also identified in BRAF-mutant hypermethylated CRCs. When compared with CRC expression data, SV2B, MLH1/EPM2AIP1, C16orf62, RCOR3, BAIAP3, OGDHL, HDHD3 and ATP1B2 demonstrated both promoter hypermethylation and decreased expression. Although SSAD backgrounds were histologically indistinguishable from ordinary SSAs, MLH1 methylation was detectable via MethyLight in 62.9% of SSAD backgrounds, and focal immunohistochemical MLH1 loss was seen in 52.5% of SSAD backgrounds.ConclusionsSignificant hyper- and hypomethylation events occur during SSA progression well before the development of histologically identifiable changes. Methylation is a heterogeneous process within individual SSAs, as typified by MLH1, where both MLH1 methylation and focal immunohistochemical MLH1 loss can be seen in the absence of dysplasia. This heterogeneity is likely a generalised phenomenon and should be taken into account in future methylation-based studies and the development of clinical methylation panels.

Project description:BACKGROUND & AIMS:Little is known about the genetic factors that contribute to the development of sessile serrated adenomas (SSAs). SSAs contain somatic mutations in BRAF or KRAS early in development. However, evidence from humans and mouse models indicates that these mutations result in oncogene-induced senescence (OIS) of intestinal crypt cells. Progression to serrated neoplasia requires cells to escape OIS via inactivation of tumor suppressor pathways. We investigated whether subjects with multiple SSAs carry germline loss-of function mutations (nonsense and splice site) in genes that regulate OIS: the p16-Rb and ATM-ATR DNA damage response pathways. METHODS:Through a bioinformatic analysis of the literature, we identified a set of genes that function at the main nodes of the p16-Rb and ATM-ATR DNA damage response pathways. We performed whole-exome sequencing of 20 unrelated subjects with multiple SSAs; most had features of serrated polyposis. We compared sequences with those from 4300 subjects matched for ethnicity (controls). We also used an integrative genomics approach to identify additional genes involved in senescence mechanisms. RESULTS:We identified mutations in genes that regulate senescence (ATM, PIF1, TELO2,XAF1, and RBL1) in 5 of 20 subjects with multiple SSAs (odds ratio, 3.0; 95% confidence interval, 0.9–8.9; P =.04). In 2 subjects,we found nonsense mutations in RNF43, indicating that it is also associated with multiple serrated polyps (odds ratio, 460; 95% confidence interval, 23.1–16,384; P = 6.8 x 10(-5)). In knockdown experiments with pancreatic duct cells exposed to UV light, RNF43 appeared to function as a regulator of ATMATRDNA damage response. CONCLUSIONS:We associated germline loss-of-function variants in genes that regulate senescence pathways with the development of multiple SSAs.We identified RNF43 as a regulator of the DNA damage response and associated nonsense variants in this gene with a high risk of developing SSAs.

Dataset Information

Computer aided detection for laterally spreading tumors and sessile serrated adenomas during colonoscopy.

Publications

Computer aided detection for laterally spreading tumors and sessile serrated adenomas during colonoscopy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets