Deep Neural Network-Based Prediction of the Risk of Advanced Colorectal Neoplasia
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Background/Aims: Risk prediction models using a deep neural network (DNN) have not been reported to predict the risk of advanced colorectal neoplasia (ACRN). The aim of this study was to compare DNN models with simple clinical score models to predict the risk of ACRN in colorectal cancer screening. Methods: Databases of screening colonoscopy from Kangbuk Samsung Hospital (n=121,794) and Kyung Hee University Hospital at Gangdong (n=3,728) were used to develop DNN-based prediction models. Two DNN models, the Asian-Pacific Colorectal Screening (APCS) model and the Korean Colorectal Screening (KCS) model, were developed and compared with two simple score models using logistic regression methods to predict the risk of ACRN. The areas under the receiver operating characteristic curves (AUCs) of the models were compared in internal and external validation databases. Results: In the internal validation set, the AUCs of DNN model 1 and the APCS score model were 0.713 and 0.662 (p<0.001), respectively, and the AUCs of DNN model 2 and the KCS score model were 0.730 and 0.667 (p<0.001), respectively. However, in the external validation set, the prediction performances were not significantly different between the two DNN models and the corresponding APCS and KCS score models (both p>0.1). Conclusions: Simple score models for the risk prediction of ACRN are as useful as DNN-based models when input variables are limited. However, further studies on this issue are warranted to predict the risk of ACRN in colorectal cancer screening because DNN-based models are currently under improvement.
Abstract
Background/Aims: Risk prediction models using a deep neural network (DNN) have not been reported to predict the risk of advanced colorectal neoplasia (ACRN). The aim of this study was to compare DNN models with simple clinical score models to predict the risk of ACRN in colorectal cancer screening. Methods: Databases of screening colonoscopy from Kangbuk Samsung Hospital (n=121,794) and Kyung Hee University Hospital at Gangdong (n=3,728) were used to develop DNN-based prediction models. Two DNN models, the Asian-Pacific Colorectal Screening (APCS) model and the Korean Colorectal Screening (KCS) model, were developed and compared with two simple score models using logistic regression methods to predict the risk of ACRN. The areas under the receiver operating characteristic curves (AUCs) of the models were compared in internal and external validation databases. Results: In the internal validation set, the AUCs of DNN model 1 and the APCS score model were 0.713 and 0.662 (p<0.001), respectively, and the AUCs of DNN model 2 and the KCS score model were 0.730 and 0.667 (p<0.001), respectively. However, in the external validation set, the prediction performances were not significantly different between the two DNN models and the corresponding APCS and KCS score models (both p>0.1). Conclusions: Simple score models for the risk prediction of ACRN are as useful as DNN-based models when input variables are limited. However, further studies on this issue are warranted to predict the risk of ACRN in colorectal cancer screening because DNN-based models are currently under improvement.
참고문헌 (32)
- Ferlay J, "Cancer incidence and mortality worldwide : sources, methods and major patterns in GLOBOCAN 2012", Int J Cancer, 136, pp.E359~ (2015)
- Shaukat A, "Long-term mortality after screening for colorectal cancer", N Engl J Med, 369, pp.1106~1114 (2013)
- Nishihara R, "Long-term colorectalcancer incidence and mortality after lower endoscopy", N Engl J Med, 369, pp.1095~1105 (2013)
- Liang PS, "Adherence to competing strategies for colorectal cancer screening over 3 years", Am J Gastroenterol, 111, pp.105~114 (2016)
- Imperiale TF, "Derivation and validation of a scoring system to stratify risk for advanced colorectal neoplasia in asymptomatic adults : a cross-sectional study", Ann Intern Med, 163, pp.339~346 (2015)
- Yeoh KG, "The Asia-Pacific Colorectal Screening score : a validated tool that stratifies risk for colorectal advanced neoplasia in asymptomatic Asian subjects", Gut, 60, pp.1236~1241 (2011)
- Kaminski MF, "A score to estimate the likelihood of detecting advanced colorectal neoplasia at colonoscopy", Gut, 63, pp.1112~1119 (2014)
- Lin OS, "Risk stratification for colon neoplasia : screening strategies using colonoscopy and computerized tomographic colonography", Gastroenterology, 131, pp.1011~1019 (2006)
- Schroy PC 3rd, "A risk prediction index for advanced colorectal neoplasia at screening colonoscopy", Am J Gastroenterol, 110, pp.1062~1071 (2015)
- Yang HJ, "Derivation and validation of a risk scoring model to predict advanced colorectal neoplasm in adults of all ages", J Gastroenterol Hepatol, 32, pp.1328~1335 (2017)
- LeCun Y, "Deep learning", Nature, 521, pp.436~444 (2015)
- Min S, "Deep learning in bioinformatics", Brief Bioinform, 18, pp.851~869 (2017)
- Esteva A, "Dermatologist-level classification of skin cancer with deep neural networks", Nature, 542, pp.115~118 (2017)
- Ting DSW, "Development and validation of a deep learning system for diabetic retinopathy and related eye diseases using retinal images from multiethnic populations with diabetes", JAMA, 318, pp.2211~2223 (2017)
- Gulshan V, "Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs", JAMA, 316, pp.2402~2410 (2016)
- Ehteshami Bejnordi B, "Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer", JAMA, 318, pp.2199~2210 (2017)
- Byrne MF, "Real-time differentiation of adenomatous and hyperplastic diminutive colorectal polyps during analysis of unaltered videos of standard colonoscopy using a deep learning model", Gut, 68, pp.94~100 (2019)
- Chen PJ, "Accurate classification of diminutive colorectal polyps using computer-aided analysis", Gastroenterology, 154, pp.568~575 (2018)
- Kim DH, "Development and validation of a risk stratification-based screening model for predicting colorectal advanced neoplasia in Korea", J Clin Gastroenterol, 49, pp.41~49 (2015)
- Rhee EJ, "Baseline glycemic status and mortality in 241, 499 Korean metropolitan subjects : a Kangbuk Samsung Health Study", Metabolism, 65, pp.68~77 (2016)
- Wen CP, "Are Asians at greater mortality risks for being overweight than Caucasians? Redefining obesity for Asians", Public Health Nutr, 12, pp.497~506 (2009)
- Tahmasebi P, "Application of a modular feedforward neural network for grade estimation", Nat Resour Res, 20, pp.25~32 (2011)
- (2016)[URL]
- Aksoy S, "Feature normalization and likelihood-based similarity measures for image retrieval", Pattern Recognit Lett, 22, pp.563~582 (2001)
- (2014)[URL]
- Glorot X, "Understanding the difficulty of training deep feedforward neural networks", J Mach Learn Res, 9, pp.249~256 (2010)
- (2015)[URL]
- Srivastava N, "Dropout : a simple way to prevent neural networks from overfitting", J Mach Learn Res, 15, pp.1929~1958 (2014)
- DeLong ER, "Comparing the areas under two or more correlated receiver operating characteristic curves : a nonparametric approach", Biometrics, 44, pp.837~845 (1988)
- Hanley JA, "A method of comparing the areas under receiver operating characteristic curves derived from the same cases", Radiology, 148, pp.839~843 (1983)
- Sarfaty M, "How to increase colorectal cancer screening rates in practice", CA Cancer J Clin, 57, pp.354~366 (2007)
- Hosmer D, "Applied logistic regression" (2000), Wiley