Development and Validation of Machine Learning Algorithms for Prediction of Colorectal Polyps Based on Electronic Health Records
Abstract
:1. Introduction
2. Materials and Methods
2.1. Study Population
2.2. Data Collection and Progress
2.3. Variable Selection
2.4. Model Development and Comparison
2.5. Model Optimization and Evaluation
2.6. Statistical Analysis
3. Results
3.1. Basic Characteristics of the Study Population
3.2. Feature Analysis
3.3. Comparison of Multiple Classification Models
3.4. Model Optimization and External Validation
3.5. Model Interpretability and Application for Clinical Utility
4. Discussion
5. Strengths and Limitations
6. Conclusions
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Siegel, R.L.; Giaquinto, A.N.; Jemal, A. Cancer statistics, 2024. CA Cancer J. Clin. 2024, 74, 12–49. [Google Scholar] [CrossRef] [PubMed]
- Rex, D.K.; Boland, C.R.; Dominitz, J.A.; Giardiello, F.M.; Johnson, D.A.; Kaltenbach, T.; Levin, T.R.; Lieberman, D.; Robertson, D.J. Colorectal cancer screening: Recommendations for physicians and patients from the U.S. Multi-Society Task Force on Colorectal Cancer. Gastrointest. Endosc. 2017, 86, 18–33. [Google Scholar] [CrossRef]
- Keum, N.; Giovannucci, E. Global burden of colorectal cancer: Emerging trends, risk factors and prevention strategies. Nat. Rev. Gastroenterol. Hepatol. 2019, 16, 713–732. [Google Scholar] [CrossRef] [PubMed]
- Strum, W.B. Colorectal Adenomas. N. Engl. J. Med. 2016, 374, 1065–1075. [Google Scholar] [CrossRef]
- Siegel, R.L.; Wagle, N.S.; Cercek, A.; Smith, R.A.; Jemal, A. Colorectal cancer statistics, 2023. CA Cancer J. Clin. 2023, 73, 233–254. [Google Scholar] [CrossRef]
- Brenner, H.; Stock, C.; Hoffmeister, M. Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: Systematic review and meta-analysis of randomized controlled trials and observational studies. BMJ 2014, 348, g2467. [Google Scholar] [CrossRef] [PubMed]
- Brenner, H.; Altenhofen, L.; Kretschmann, J.; Rösch, T.; Pox, C.; Stock, C.; Hoffmeister, M. Trends in Adenoma Detection Rates During the First 10 Years of the German Screening Colonoscopy Program. Gastroenterology 2015, 149, 356–366.e1. [Google Scholar] [CrossRef]
- Levin, B.; Lieberman, D.A.; McFarland, B.; Smith, R.A.; Brooks, D.; Andrews, K.S.; Dash, C.; Giardiello, F.M.; Glick, S.; Levin, T.R.; et al. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: A joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008, 134, 1570–1595. [Google Scholar] [CrossRef]
- Cooper, G.S.; Kou, T.D.; Rex, D.K. Complications following colonoscopy with anesthesia assistance: A population-based analysis. JAMA Intern. Med. 2013, 173, 551–556. [Google Scholar] [CrossRef]
- Inadomi, J.M.; Vijan, S.; Janz, N.K.; Fagerlin, A.; Thomas, J.P.; Lin, Y.V.; Muñoz, R.; Lau, C.; Somsouk, M.; El-Nachef, N.; et al. Adherence to colorectal cancer screening: A randomized clinical trial of competing strategies. Arch. Intern. Med. 2012, 172, 575–582. [Google Scholar]
- Philipp, G.M.; Andrey, G.C.; Henrik, S.; Robert, H.; Niels, B.; Florian, J.; Ralf, M. Directedness, correlations, and daily cycles in springbok motion: From data via stochastic models to movement prediction. Phys. Rev. Res. 2023, 5, 043129. [Google Scholar]
- Ghani, A.; Aina, A.; See, C.H.; Yu, H.; Keates, S. Accelerated Diagnosis of Novel Coronavirus (COVID-19)—Computer Vision with Convolutional Neural Networks (CNNs). Electronics 2022, 11, 1148. [Google Scholar] [CrossRef]
- Ghani, A.; Hodeify, R.; See, C.H.; Keates, S.; Lee, D.-J.; Bouridane, A. Computer Vision-Based Kidney’s (HK-2) Damaged Cells Classification with Reconfigurable Hardware Accelerator (FPGA). Electronics 2022, 11, 4234. [Google Scholar] [CrossRef]
- Urban, G.; Tripathi, P.; Alkayali, T.; Mittal, M.; Jalali, F.; Karnes, W.; Baldi, P. Deep Learning Localizes and Identifies Polyps in Real Time with 96% Accuracy in Screening Colonoscopy. Gastroenterology 2018, 155, 1069–1078.e8. [Google Scholar] [CrossRef]
- Luo, Y.; Zhang, Y.; Liu, M.; Lai, Y.; Liu, P.; Wang, Z.; Xing, T.; Huang, Y.; Li, Y.; Li, A.; et al. Artificial Intelligence-Assisted Colonoscopy for Detection of Colon Polyps: A Prospective, Randomized Cohort Study. J. Gastrointest. Surg. 2021, 25, 2011–2018. [Google Scholar] [CrossRef] [PubMed]
- Grosu, S.; Wesp, P.; Graser, A.; Maurus, S.; Schulz, C.; Knösel, T.; Cyran, C.C.; Ricke, J.; Ingrisch, M.; Kazmierczak, P.M. Machine Learning-based Differentiation of Benign and Premalignant Colorectal Polyps Detected with CT Colonography in an Asymptomatic Screening Population: A Proof-of-Concept Study. Radiology 2021, 299, 326–335. [Google Scholar] [CrossRef]
- Benke, K.; Benke, G. Artificial Intelligence and Big Data in Public Health. Int. J. Environ. Res. Public Health 2018, 15, 2796. [Google Scholar] [CrossRef]
- Vera Cruz, G.; Bucourt, E.; Réveillère, C.; Martaillé, V.; Joncker-Vannier, I.; Goupille, P.; Mulleman, D.; Courtois, R. Machine learning reveals the most important psychological and social variables predicting the differential diagnosis of rheumatic and musculoskeletal diseases. Rheumatol. Int. 2022, 42, 1053–1062. [Google Scholar] [CrossRef]
- Nohara, Y.; Matsumoto, K.; Soejima, H.; Nakashima, N. Explanation of machine learning models using shapley additive explanation and application for real data in hospital. Comput. Methods Programs Biomed. 2022, 214, 106584. [Google Scholar] [CrossRef]
- Shi, D.; Zhou, F.; Mu, W.; Ling, C.; Mu, T.; Yu, G.; Li, R. Deep insights into the viscosity of deep eutectic solvents by an XGBoost-based model plus SHapley Additive exPlanation. Phys. Chem. Chem. Phys. 2022, 24, 26029–26036. [Google Scholar] [CrossRef]
- Sarker, I.H. Machine Learning: Algorithms, Real-World Applications and Research Directions. SN Comput. Sci. 2021, 2, 160. [Google Scholar] [CrossRef]
- Yan, J.; Xu, Y.; Cheng, Q.; Jiang, S.; Wang, Q.; Xiao, Y.; Ma, C.; Yan, J.; Wang, X. LightGBM: Accelerated genomically designed crop breeding through ensemble learning. Genome Biol. 2021, 22, 271. [Google Scholar] [CrossRef]
- Uddin, S.; Khan, A.; Hossain, M.E.; Moni, M.A. Comparing different supervised machine learning algorithms for disease prediction. BMC Med. Inform. Decis. Mak. 2019, 19, 281. [Google Scholar] [CrossRef] [PubMed]
- Zhuang, Z.; Qi, Y.; Yao, Y.; Yu, Y. A predictive model for disease severity among COVID-19 elderly patients based on IgG subtypes and machine learning. Front. Immunol. 2023, 14, 1286380. [Google Scholar] [CrossRef]
- Wei, W.; Wang, Y.; Ouyang, R.; Wang, T.; Chen, R.; Yuan, X.; Wang, F.; Wu, S.; Hou, H. Machine Learning for Early Discrimination Between Lung Cancer and Benign Nodules Using Routine Clinical and Laboratory Data. Ann. Surg. Oncol. 2024, 1–12. [Google Scholar] [CrossRef]
- Zhang, A.; Yin, Y.; Tian, J.; Wang, X.; Yue, Z.; Pei, L.; Liu, L.; Qin, L.; Jia, M.; Wang, H.; et al. The close association of micronutrients with COVID-19. Heliyon 2024, 10, e28629. [Google Scholar] [CrossRef]
- Guo, Q.H.; Xie, F.C.; Zhong, F.M.; Wen, W.; Zhang, X.R.; Yu, X.J.; Wang, X.L.; Huang, B.; Li, L.P.; Wang, X.Z. Application of interpretable machine learning algorithms to predict distant metastasis in ovarian clear cell carcinoma. Cancer Med. 2024, 13, e7161. [Google Scholar] [CrossRef] [PubMed]
- Shaukat, A.; Kahi, C.J.; Burke, C.A.; Rabeneck, L.; Sauer, B.G.; Rex, D.K. ACG Clinical Guidelines: Colorectal Cancer Screening 2021. Am. J. Gastroenterol. 2021, 116, 458–479. [Google Scholar] [CrossRef] [PubMed]
- Dave, D.; Lu, R.; Klair, J.S.; Kassim, T.; Ashfaq, M.; Chandra, S.; Murali, A. Prevalence and Predictors of Adenomas in Young Adults Undergoing Diagnostic Colonoscopy in a Multicenter Midwest U.S. Cohort. Am. J. Gastroenterol. 2019, 114, S76. [Google Scholar] [CrossRef]
- Lawrence, E.M.; Pickhardt, P.J.; Kim, D.H.; Robbins, J.B. Colorectal polyps: Stand-alone performance of computer-aided detection in a large asymptomatic screening population. Radiology 2010, 256, 791–798. [Google Scholar] [CrossRef]
- Wang, P.; Xiao, X.; Glissen Brown, J.R.; Berzin, T.M.; Tu, M.; Xiong, F.; Hu, X.; Liu, P.; Song, Y.; Zhang, D.; et al. Development and validation of a deep-learning algorithm for the detection of polyps during colonoscopy. Nat. Biomed. Eng. 2018, 2, 741–748. [Google Scholar] [CrossRef]
- Gargallo, C.J.; Aznar, R.; Carrera, P.; Lanas, Á.; Ferrández, Á.; Quintero, E.; Carrillo, M.; Alonso-Abreu, I.; Esteban, L.M.; de la Vega Rodrigálvarez, M.; Del Hoyo-Alonso, R.; García-González, M.A. Predictive Value of Genetic Risk Scores in the Development of Colorectal Adenomas. Dig. Dis. Sci. 2022, 67, 4049–4058. [Google Scholar] [CrossRef] [PubMed]
- Huang, Y.; Liu, Y.; Yin, X.; Zhang, T.; Hao, Y.; Zhang, P.; Yang, Y.; Gao, Z.; Liu, S.; Yu, S.; et al. Establishment of clinical predictive model based on the study of influence factors in patients with colorectal polyps. Front. Surg. 2023, 10, 1077175. [Google Scholar] [CrossRef] [PubMed]
- Lyu, Z.; Hang, D.; He, X.; Wu, K.; Cao, Y.; Rosner, B.; Chan, A.T.; Ogino, S.; Li, N.; Dai, M.; et al. Simple Prediction Model for Colorectal Serrated Polyps: Development and External Validation Study in U.S. Prospective Cohorts. Cancer Prev. Res. 2023, 16, 293–302. [Google Scholar] [CrossRef] [PubMed]
- Kaltenbach, T.; Rastogi, A.; Rouse, R.V.; McQuaid, K.R.; Sato, T.; Bansal, A.; Kosek, J.C.; Soetikno, R. Real-time optical diagnosis for diminutive colorectal polyps using narrow-band imaging: The VALID randomised clinical trial. Gut 2015, 64, 1569–1577. [Google Scholar] [CrossRef] [PubMed]
- Hussan, H.; Zhao, J.; Badu-Tawiah, A.K.; Stanich, P.; Tabung, F.; Gray, D.; Ma, Q.; Kalady, M.; Clinton, S.K. Utility of machine learning in developing a predictive model for early-age-onset colorectal neoplasia using electronic health records. PLoS ONE 2022, 17, e0265209. [Google Scholar] [CrossRef]
- Hao, Y.; Wang, Y.; Qi, M.; He, X.; Zhu, Y.; Hong, J. Risk Factors for Recurrent Colorectal Polyps. Gut Liver 2020, 14, 399–411. [Google Scholar] [CrossRef]
- US Preventive Services Task Force; Davidson, K.W.; Barry, M.J.; Mangione, C.M.; Cabana, M.; Caughey, A.B.; Davis, E.M.; Donahue, K.E.; Stevermer, J.; Tseng, C.W.; et al. ;Screening for Colorectal Cancer: US Preventive Services Task Force Recommendation Statement. JAMA 2021, 325, 1965–1977. [Google Scholar] [CrossRef]
- O’Sullivan, D.E.; Sutherland, R.L.; Town, S.; Chow, K.; Fan, J.; Forbes, N.; Heitman, S.J.; Hilsden, R.J.; Brenner, D.R. Risk Factors for Early-Onset Colorectal Cancer: A Systematic Review and Meta-analysis. Clin. Gastroenterol. Hepatol. 2022, 20, 1229–1240.e5. [Google Scholar] [CrossRef]
- Pendergrass, C.J.; Edelstein, D.L.; Hylind, L.M.; Phillips, B.T.; Iacobuzio-Donahue, C.; Romans, K.; Griffin, C.A.; Cruz-Correa, M.; Tersmette, A.C.; Offerhaus, G.J.; et al. Occurrence of colorectal adenomas in younger adults: An epidemiologic necropsy study. Clin. Gastroenterol. Hepatol. 2008, 6, 1011–1015. [Google Scholar] [CrossRef]
- May, F.P.; Shaukat, A. State of the Science on Quality Indicators for Colonoscopy and How to Achieve Them. Am. J. Gastroenterol. 2020, 115, 1183–1190. [Google Scholar] [CrossRef] [PubMed]
- Jang, S.; Han, H.; Oh, Y.; Kim, Y. Sex differences in inflammation correlated with estrogen and estrogen receptor-β levels in azoxymethane/dextran sodium sulfate-induced colitis-associated colorectal cancer mice. Heliyon 2024, 10, e28121. [Google Scholar] [CrossRef]
- Issa, J.P.; Ottaviano, Y.L.; Celano, P.; Hamilton, S.R.; Davidson, N.E.; Baylin, S.B. Methylation of the oestrogen receptor CpG island links ageing and neoplasia in human colon. Nat. Genet. 1994, 7, 536–540. [Google Scholar] [CrossRef] [PubMed]
- Ocvirk, S.; O’Keefe, S.J.D. Dietary fat, bile acid metabolism and colorectal cancer. Semin. Cancer Biol. 2021, 73, 347–355. [Google Scholar] [CrossRef] [PubMed]
- Hammarström, S. The carcinoembryonic antigen (CEA) family: Structures, suggested functions and expression in normal and malignant tissues. Semin. Cancer Biol. 1999, 9, 67–81. [Google Scholar] [CrossRef]
- Kong, X.; Zhang, Y.; Xiang, L.; You, Y.; Duan, Y.; Zhao, Y.; Li, S.; Wu, R.; Zhang, J.; Zhou, L.; et al. Fusobacterium nucleatum-triggered neutrophil extracellular traps facilitate colorectal carcinoma progression. J. Exp. Clin. Cancer Res. 2023, 42, 236. [Google Scholar] [CrossRef]
- Mignini, I.; Ainora, M.E.; Di Francesco, S.; Galasso, L.; Gasbarrini, A.; Zocco, M.A. Tumorigenesis in Inflammatory Bowel Disease: Microbiota-Environment Interconnections. Cancers 2023, 15, 3200. [Google Scholar] [CrossRef]
- Riondino, S.; Roselli, M.; Palmirotta, R.; Della-Morte, D.; Ferroni, P.; Guadagni, F. Obesity and colorectal cancer: Role of adipokines in tumor initiation and progression. World J. Gastroenterol. 2014, 20, 5177–5190. [Google Scholar] [CrossRef]
- Ottaviano, L.F.; Li, X.; Murray, M.; Frye, J.T.; Lung, B.E.; Zhang, Y.Y.; Yang, J.; Taub, E.M.; Bucobo, J.C.; Buscaglia, J.M.; et al. Type 2 diabetes impacts colorectal adenoma detection in screening colonoscopy. Sci. Rep. 2020, 10, 7793. [Google Scholar] [CrossRef]
- Ben, Q.; An, W.; Jiang, Y.; Zhan, X.; Du, Y.; Cai, Q.C.; Gao, J.; Li, Z. Body mass index increases risk for colorectal adenomas based on meta-analysis. Gastroenterology 2012, 142, 762–772. [Google Scholar] [CrossRef]
- Kahn, D.; Macias, E.; Zarini, S.; Garfield, A.; Zemski Berry, K.; MacLean, P.; Gerszten, R.E.; Libby, A.; Solt, C.; Schoen, J.; et al. Exploring Visceral and Subcutaneous Adipose Tissue Secretomes in Human Obesity: Implications for Metabolic Disease. Endocrinology 2022, 163, bqac140. [Google Scholar] [CrossRef] [PubMed]
Variables | Cohort 1: Training Dataset (n = 3798) | Cohort 2: Validation Dataset (n = 1628) | ||||||
---|---|---|---|---|---|---|---|---|
Colorectal Polyps (n = 1723) | Non-Colorectal Polyps (n = 2075) | χ2/t | p Value | Colorectal Polyps (n = 764) | Non-Colorectal Polyps (n = 864) | χ2/t | p Value | |
Sex (M/F), n (%) | ||||||||
male | 1339 (77.713) | 1205 (58.072) | 164.199 | <0.001 | 578 (75.654) | 506 (58.565) | 53.223 | <0.001 |
female | 384 (22.287) | 870 (41.928) | 186 (24.346) | 358 (41.435) | ||||
Age (years) | 51.000 [43.000, 58.000] | 47.000 [38.000, 54.000] | 11.608 | <0.001 | 52.000 [44.000, 58.000] | 47.000 [38.000, 54.000] | 9.265 | <0.001 |
BMI (kg/m2) | 25.267 ± 3.246 | 24.294 ± 3.480 | 8.674 | <0.001 | 25.130 ± 3.397 | 24.293 ± 3.457 | 4.771 | <0.001 |
SBP (mmHg) | 125.046 ± 16.467 | 120.701 ± 15.849 | 8.061 | <0.001 | 124.397 ± 16.460 | 121.398 ± 15.532 | 3.673 | <0.001 |
DBP (mmHg) | 79.052 ± 11.127 | 75.874 ± 10.797 | 8.686 | <0.001 | 78.434 ± 11.212 | 76.060 ± 10.861 | 4.213 | <0.001 |
Serum tumor markers | ||||||||
CEA (ng/mL) | 2.159 ± 1.344 | 1.829 ± 1.396 | 7.265 | <0.001 | 2.074 ± 1.225 | 1.911 ± 1.649 | 2.206 | 0.028 |
AFP (ng/mL) | 3.478 ± 3.740 | 3.582 ± 8.773 | −0.402 | 0.688 | 3.517 ± 1.763 | 3.310 ± 3.411 | 1.309 | 0.191 |
Glycolipid metabolites | ||||||||
TG (mmol/L) | 1.954 ± 2.030 | 1.576 ± 1.312 | 6.627 | <0.001 | 1.889 ± 1.884 | 1.669 ± 1.463 | 2.638 | 0.008 |
TC (mmol/L) | 4.825 ± 0.967 | 4.762 ± 0.889 | 2.08 | 0.038 | 4.767 ± 0.909 | 4.781 ± 0.909 | −0.307 | 0.759 |
LDL-C (mmol/L) | 2.977 ± 0.805 | 2.945 ± 0.780 | 1.241 | 0.215 | 2.915 ± 0.810 | 2.935 ± 0.837 | −0.489 | 0.625 |
HDL-C (mmol/L) | 1.187 ± 0.292 | 1.278 ± 0.310 | −9.243 | <0.001 | 1.212 ± 0.306 | 1.269 ± 0.313 | −3.671 | <0.001 |
FINS (μIU/mL) | 8.699 ± 5.916 | 7.971 ± 5.082 | 3.474 | <0.001 | 8.686 ± 5.264 | 7.897 ± 5.938 | 2.395 | 0.017 |
FBG (mmol/L) | 5.570 ± 1.299 | 5.281 ± 0.955 | 7.633 | <0.001 | 5.483 ± 1.011 | 5.295 ± 0.804 | 4.067 | <0.001 |
HbA1c (%) | 5.868 ± 0.766 | 5.689 ± 0.571 | 7.741 | <0.001 | 5.802 ± 0.675 | 5.685 ± 0.483 | 3.858 | <0.001 |
Renal function | ||||||||
UREA (mmol/L) | 5.172 ± 1.335 | 4.966 ± 1.255 | 4.878 | <0.001 | 5.189 ± 1.241 | 4.911 ± 1.223 | 4.524 | <0.001 |
Scr (μmol/L) | 78.903 ± 15.761 | 74.299 ± 15.843 | 8.914 | <0.001 | 78.092 ± 14.684 | 74.256 ± 15.677 | 5.079 | <0.001 |
UA (μmol/L) | 368.301 ± 91.944 | 343.079 ± 95.507 | 8.239 | <0.001 | 365.022 ± 94.235 | 346.272 ± 101.939 | 3.831 | <0.001 |
Liver function | ||||||||
TBIL (μmol/L) | 12.325 ± 5.252 | 12.047 ± 5.127 | 1.643 | 0.1 | 12.407 ± 5.483 | 12.137 ± 5.339 | 1.003 | 0.316 |
ALT (U/L) | 26.368 ± 22.886 | 23.508 ± 20.073 | 4.093 | <0.001 | 24.562 ± 19.530 | 24.811 ± 20.877 | −0.248 | 0.804 |
AST (U/L) | 23.201 ± 13.150 | 22.182 ± 12.247 | 2.456 | 0.014 | 22.063 ± 8.994 | 22.569 ± 10.694 | −1.021 | 0.307 |
GLB (g/L) | 28.496 ± 3.498 | 28.849 ± 3.575 | −3.056 | 0.002 | 28.337 ± 3.313 | 28.837 ± 3.442 | −2.968 | 0.003 |
ALB (g/L) | 46.145 ± 2.446 | 46.382 ± 2.492 | −2.937 | 0.003 | 46.110 ± 2.539 | 46.416 ± 2.489 | −2.448 | 0.014 |
TP (g/L) | 74.641 ± 3.919 | 75.228 ± 3.930 | −4.58 | <0.001 | 74.441 ± 3.853 | 75.254 ± 4.052 | −4.123 | <0.001 |
Blood routine | ||||||||
WBC (×109/L) | 6.008 ± 1.531 | 5.736 ± 1.549 | 5.402 | <0.001 | 5.988 ± 1.589 | 5.733 ± 1.612 | 3.194 | 0.001 |
PLT (×109/L) | 232.150 ± 54.401 | 236.957 ± 58.383 | −2.62 | 0.009 | 229.818 ± 53.757 | 235.027 ± 57.765 | −1.874 | 0.061 |
RBC (×1012/L) | 4.931 ± 0.474 | 4.837 ± 0.524 | 5.755 | <0.001 | 4.943 ± 0.513 | 4.850 ± 0.544 | 3.539 | <0.001 |
NEUT# (×109/L) | 3.484 ± 1.169 | 3.309 ± 1.200 | 4.525 | <0.001 | 3.515 ± 1.250 | 3.310 ± 1.200 | 3.362 | <0.001 |
EO# (×109/L) | 0.155 ± 0.142 | 0.133 ± 0.123 | 4.97 | <0.001 | 0.151 ± 0.130 | 0.135 ± 0.124 | 2.514 | 0.012 |
BASO# (×109/L) | 0.030 ± 0.018 | 0.027 ± 0.017 | 4.128 | <0.001 | 0.030 ± 0.018 | 0.027 ± 0.016 | 4.008 | <0.001 |
LYMPH# (×109/L) | 1.944 ± 0.556 | 1.894 ± 0.555 | 2.753 | 0.006 | 1.898 ± 0.550 | 1.892 ± 0.586 | 0.209 | 0.834 |
MONO# (×109/L) | 0.396 ± 0.126 | 0.373 ± 0.127 | 5.472 | <0.001 | 0.393 ± 0.131 | 0.368 ± 0.126 | 3.909 | <0.001 |
HGB (g/L) | 150.450 ± 14.792 | 145.389 ± 16.595 | 9.924 | <0.001 | 150.018 ± 15.538 | 145.502 ± 17.077 | 5.581 | <0.001 |
MCH (pg) | 30.572 ± 2.001 | 30.120 ± 2.244 | 6.484 | <0.001 | 30.422 ± 2.097 | 30.074 ± 2.301 | 3.172 | 0.002 |
MCHC (g/L) | 333.333 ± 10.987 | 331.056 ± 11.602 | 6.164 | <0.001 | 333.304 ± 12.162 | 330.689 ± 12.365 | 4.288 | <0.001 |
MCV (fL) | 91.701 ± 5.195 | 90.939 ± 5.648 | 4.292 | <0.001 | 91.247 ± 5.159 | 90.892 ± 5.715 | 1.308 | 0.191 |
MPV | 10.510 ± 1.162 | 10.659 ± 1.169 | −3.886 | <0.001 | 10.598 ± 1.164 | 10.693 ± 1.210 | −1.601 | 0.11 |
PCT | 0.242 ± 0.051 | 0.250 ± 0.054 | −4.88 | <0.001 | 0.241 ± 0.051 | 0.248 ± 0.054 | −2.611 | 0.009 |
novel composite indices | ||||||||
TyG index | 9113.408 ± 12,985.090 | 6810.454 ± 6279.994 | 6.695 | <0.001 | 8463.506 ± 9463.168 | 7254.801 ± 7205.916 | 2.89 | 0.004 |
SIRI | 0.770 ± 0.466 | 0.714 ± 0.525 | 3.447 | <0.001 | 0.807 ± 0.589 | 0.706 ± 0.488 | 3.749 | <0.001 |
LHR | 1.745 ± 0.693 | 1.593 ± 0.688 | 6.742 | <0.001 | 1.678 ± 0.685 | 1.599 ± 0.707 | 2.264 | 0.024 |
NHR | 3.140 ± 1.393 | 2.788 ± 1.346 | 7.848 | <0.001 | 3.107 ± 1.386 | 2.807 ± 1.362 | 4.367 | <0.001 |
NPR | 0.016 ± 0.006 | 0.015 ± 0.006 | 5.7 | <0.001 | 0.016 ± 0.006 | 0.015 ± 0.006 | 4.191 | <0.001 |
LMR | 5.228 ± 1.755 | 5.429 ± 1.831 | −3.454 | <0.001 | 5.174 ± 1.778 | 5.487 ± 1.855 | −3.464 | <0.001 |
PLR | 127.316 ± 41.929 | 133.279 ± 45.883 | −4.143 | <0.001 | 128.693 ± 40.971 | 133.208 ± 47.315 | −2.043 | 0.041 |
Models | AUC (SD) | Cutoff (SD) | Accuracy (SD) | Sensitivity (SD) | Specificity (SD) | PPV (SD) | NPV (SD) | F1 Score (SD) | Kappa (SD) |
---|---|---|---|---|---|---|---|---|---|
Training set | |||||||||
LR | 0.682 (0.006) | 0.472 (0.020) | 0.645 (0.006) | 0.613 (0.047) | 0.672 (0.044) | 0.609 (0.017) | 0.677 (0.013) | 0.610 (0.019) | 0.284 (0.010) |
XGBoost | 0.881 (0.008) | 0.453 (0.020) | 0.797 (0.007) | 0.800 (0.040) | 0.795 (0.028) | 0.765 (0.016) | 0.828 (0.023) | 0.781 (0.014) | 0.592 (0.016) |
LightGBM | 0.697 (0.003) | 0.455 (0.008) | 0.646 (0.008) | 0.645 (0.052) | 0.650 (0.056) | 0.612 (0.021) | 0.678 (0.017) | 0.626 (0.016) | 0.286 (0.013) |
RF | 0.826 (0.005) | 0.458 (0.020) | 0.748 (0.007) | 0.745 (0.045) | 0.752 (0.044) | 0.716 (0.025) | 0.781 (0.021) | 0.729 (0.012) | 0.494 (0.013) |
AdaBoost | 0.727 (0.005) | 0.498 (0.001) | 0.668 (0.005) | 0.640 (0.041) | 0.692 (0.039) | 0.634 (0.020) | 0.698 (0.014) | 0.636 (0.014) | 0.330 (0.009) |
DT | 0.730 (0.006) | 0.464 (0.010) | 0.668 (0.007) | 0.694 (0.047) | 0.643 (0.047) | 0.631 (0.020) | 0.705 (0.017) | 0.659 (0.014) | 0.334 (0.011) |
GBDT | 0.730 (0.006) | 0.454 (0.004) | 0.668 (0.007) | 0.694 (0.047) | 0.643 (0.047) | 0.631 (0.020) | 0.705 (0.017) | 0.659 (0.014) | 0.334 (0.011) |
GNB | 0.662 (0.005) | 0.315 (0.046) | 0.632 (0.006) | 0.633 (0.043) | 0.632 (0.045) | 0.589 (0.014) | 0.675 (0.009) | 0.609 (0.016) | 0.263 (0.009) |
MLP | 0.535 (0.070) | 0.548 (0.305) | 0.554 (0.020) | 0.527 (0.366) | 0.557 (0.336) | NaN (NaN) | 0.581 (0.047) | NaN (NaN) | 0.057 (0.085) |
Internal validation set | |||||||||
LR | 0.688 (0.023) | 0.472 (0.020) | 0.643 (0.020) | 0.660 (0.048) | 0.656 (0.052) | 0.614 (0.034) | 0.670 (0.020) | 0.635 (0.028) | 0.282 (0.039) |
XGBoost | 0.680 (0.019) | 0.453 (0.020) | 0.630 (0.019) | 0.586 (0.107) | 0.708 (0.090) | 0.597 (0.033) | 0.662 (0.021) | 0.588 (0.064) | 0.258 (0.036) |
LightGBM | 0.678 (0.018) | 0.455 (0.008) | 0.633 (0.015) | 0.565 (0.065) | 0.711 (0.072) | 0.607 (0.029) | 0.658 (0.020) | 0.583 (0.039) | 0.261 (0.027) |
RF | 0.688 (0.021) | 0.458 (0.020) | 0.640 (0.016) | 0.637 (0.056) | 0.665 (0.050) | 0.607 (0.024) | 0.672 (0.029) | 0.619 (0.024) | 0.276 (0.034) |
AdaBoost | 0.694 (0.022) | 0.498 (0.001) | 0.644 (0.015) | 0.579 (0.071) | 0.734 (0.073) | 0.614 (0.027) | 0.669 (0.017) | 0.593 (0.039) | 0.282 (0.033) |
DT | 0.653 (0.024) | 0.464 (0.010) | 0.613 (0.023) | 0.617 (0.080) | 0.630 (0.084) | 0.577 (0.029) | 0.648 (0.027) | 0.594 (0.037) | 0.223 (0.047) |
GBDT | 0.652 (0.023) | 0.454 (0.004) | 0.613 (0.023) | 0.617 (0.080) | 0.629 (0.084) | 0.577 (0.029) | 0.647 (0.027) | 0.593 (0.037) | 0.222 (0.046) |
GNB | 0.665 (0.012) | 0.315 (0.046) | 0.627 (0.009) | 0.652 (0.044) | 0.633 (0.049) | 0.591 (0.025) | 0.665 (0.026) | 0.619 (0.028) | 0.253 (0.019) |
MLP | 0.535 (0.074) | 0.548 (0.305) | 0.547 (0.028) | 0.443 (0.373) | 0.646 (0.316) | NaN (NaN) | 0.570 (0.045) | NaN (NaN) | 0.053 (0.077) |
Sets | AUC (95% CI) | Cutoff (95% CI) | Accuracy (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) | F1 Score (95% CI) |
---|---|---|---|---|---|---|---|---|
Training set | 0.732 (0.713–0.751) | 0.498 (0.497–0.499) | 0.670 (0.667–0.674) | 0.662 (0.619–0.706) | 0.678 (0.636–0.720) | 0.637 (0.619–0.656) | 0.706 (0.694–0.719) | 0.646 (0.632–0.661) |
Internal validation set | 0.687 (0.626–0.749) | 0.498 (0.497–0.499) | 0.632 (0.618–0.646) | 0.635 (0.550–0.721) | 0.674 (0.591–0.758) | 0.593 (0.576–0.611) | 0.673 (0.654–0.691) | 0.608 (0.560–0.655) |
test set | 0.66 | 0.498 | 0.627 | 0.447 | 0.801 | 0.584 | 0.66 | 0.507 |
Outcome | AUC | Cutoff | Accuracy | Sensitivity | Specificity | PPV | NPV | F1 Score | Kappa |
---|---|---|---|---|---|---|---|---|---|
Colorectal polyps | 0.675 | 0.498 | 0.631 | 0.73 | 0.544 | 0.61 | 0.648 | 0.665 | 0.259 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Ba, Q.; Yuan, X.; Wang, Y.; Shen, N.; Xie, H.; Lu, Y. Development and Validation of Machine Learning Algorithms for Prediction of Colorectal Polyps Based on Electronic Health Records. Biomedicines 2024, 12, 1955. https://doi.org/10.3390/biomedicines12091955
Ba Q, Yuan X, Wang Y, Shen N, Xie H, Lu Y. Development and Validation of Machine Learning Algorithms for Prediction of Colorectal Polyps Based on Electronic Health Records. Biomedicines. 2024; 12(9):1955. https://doi.org/10.3390/biomedicines12091955
Chicago/Turabian StyleBa, Qinwen, Xu Yuan, Yun Wang, Na Shen, Huaping Xie, and Yanjun Lu. 2024. "Development and Validation of Machine Learning Algorithms for Prediction of Colorectal Polyps Based on Electronic Health Records" Biomedicines 12, no. 9: 1955. https://doi.org/10.3390/biomedicines12091955