Prediction model for neoplastic gallbladder polyps based on random forest algorithm

doi:10.3877/cma.j.issn.2095-3232.2025013

Abstract

Abstract:

Objective

To explore the risk factors of neoplastic gallbladder polyps, and construct a prediction model for neoplastic gallbladder polyps based on random forest algorithm.

Methods

Clinicopathological data of 745 patients with gallbladder polyps who underwent cholecystectomy in 11 medical centers from January 2015 to August 2023 were collected.The informed consents of all patients were obtained and the local ethical committee approval was received.Among them, 286 patients were male and 459 female, aged from 18 to 80 years, with a median age of 46 years.The maximum diameter of gallbladder polyps was ranged from 10 to 15 mm, and the median diameter was 11 mm.Univariate analysis of the risk factors of neoplastic gallbladder polyps was conducted by Chi-square test or Mann Whitney U test.According to the admission date, they were divided into the training set (n=588) and test set (n=157).The training set was used to construct the random forest prediction model, and the test set was utilized to validate the prediction model.The prediction performance of this model was assessed by the area under the ROC curve (AUC) and confusion matrix.

Results

In this study, non-neoplastic gallbladder polyps patients accounted for 87.2%(650/745), including 518 cases of cholesterol polyps, 55 cases of inflammatory polyps and 47 cases of adenomatous hyperplasia.The proportion of neoplastic gallbladder polyps was 12.8%(95/745), including 83 cases of gallbladder adenomas and 12 cases of T1 gallbladder carcinomas.Univariate analysis showed that the number of polyp, maximum and minimum diameter of polyp, polyp short diameter, polyp basal status, polyp location and echo intensity were correlated with the incidence of neoplastic gallbladder polyps (χ²=20.675, Z=-4.694, Z=-2.595, χ²=6.692, Z=3.935, Z=-2.690; P＜0.05).Based on the risk factors of neoplastic gallbladder polyps and the ranking of importance, a random forest prediction model was constructed.The AUC of the training and test sets was 0.79 and 0.69, with a sensitivity of 0.74 and 0.63 and a specificity of 0.75 and 0.68, respectively.Based on the random forest prediction model and confusion matrix analysis of neoplastic gallbladder polyps, the accuracy of the training and test sets was 75% and 68%,respectively.

Conclusions

The incidence of neoplastic gallbladder polyps is significantly correlated with the number of polyp, polyp size, polyp basal status, polyp location and echo intensity, etc.The prediction model based on random forest algorithm contributes to identifying neoplastic gallbladder polyps and providing decision support for surgical diagnosis, treatment and follow-up strategy for patients with gallbladder polyps.

Key words: Benign gallbladder disease, Gallbladder polyps, Neoplastic polyps, Cholecystectomy, Random forest algorithm

Qi Li, Minghui Dou, Pengbo Jia, Xintuan Wang, Da Lei, Junhui Li, Rui Yang, Chenglin Yang, Xiaodi Zhang, Qiwei Hao, Xilin Geng, Yu Zhang, Yimin Liu, Zhihua Guo, Chunhe Yao, Tianyi Wang, Zhiqiang Cai, Shubin Si, Wenbin Yang, Zhimin Geng, Dong Zhang. Prediction model for neoplastic gallbladder polyps based on random forest algorithm[J]. Chinese Journal of Hepatic Surgery(Electronic Edition), 2025, 14(01): 46-52.

/ / Recommend

Add to citation manager EndNote|Ris|BibTeX

URL: https://zhgzwkssxdzzz.cma-cmc.com.cn/EN/10.3877/cma.j.issn.2095-3232.2025013

https://zhgzwkssxdzzz.cma-cmc.com.cn/EN/Y2025/V14/I01/46

Figures/Tables 5

References 27

[1]	中华医学会外科学分会胆道外科学组, 中国医师协会外科医师分会胆道外科医师委员会.胆囊良性疾病外科治疗的专家共识(2021版)[J].中华外科杂志, 2022, 60(1):4-9.
[2]	Riddell ZC, Corallo C, Albazaz R, et al.Gallbladder polyps and adenomyomatosis[J].Br J Radiol, 2023, 96(1142):20220115.
[3]	Zhang X, Wang J, Wu B, et al.A nomogram-based model to predict neoplastic risk for patients with gallbladder polyps[J].J Clin Transl Hepatol, 2022, 10(2):263-272.
[4]	Yu MH, Kim YJ, Park HS, et al.Benign gallbladder diseases: imaging techniques and tips for differentiating with malignant gallbladder diseases[J].World J Gastroenterol, 2020, 26(22):2967-2986.
[5]	Foley KG, Lahaye MJ, Thoeni RF, et al.Management and follow-up of gallbladder polyps: updated joint guidelines between the ESGAR,EAES, EFISDS and ESGE[J].Eur Radiol, 2022, 32(5):3358-3368.
[6]	Li Q, Dou M, Zhang J, et al.A Bayesian network model to predict neoplastic risk for patients with gallbladder polyps larger than 10 mm based on preoperative ultrasound features[J].Surg Endosc, 2023,37(7):5453-5463.
[7]	Wennmacker SZ, van Dijk AH, Raessens JHJ, et al.Polyp size of 1 cm is insufficient to discriminate neoplastic and non-neoplastic gallbladder polyps[J].Surg Endosc, 2019, 33(5):1564-1571.
[8]	Liu K, Lin N, You Y, et al.Risk factors to discriminate neoplastic polypoid lesions of gallbladder: a large-scale case-series study[J].Asian J Surg, 2021, 44(12):1515-1519.
[9]	Park HY, Oh SH, Lee KH, et al.Is cholecystectomy a reasonable treatment option for simple gallbladder polyps larger than 10 mm?[J].World J Gastroenterol, 2015, 21(14):4248-4254.
[10]	Chattopadhyay D, Lochan R, Balupuri S, et al.Outcome of gall bladder polypoidal lesions detected by transabdominal ultrasound scanning:a nine year experience[J].World J Gastroenterol, 2005, 11(14):2171-2173.
[11]	陈少华, 王成方, 李会星, 等.胆囊息肉恶变的危险因素分析[J].中华消化外科杂志, 2019, 18(2):146-151.
[12]	Sun Y, Yang Z, Lan X, et al.Neoplastic polyps in gallbladder: a retrospective study to determine risk factors and treatment strategy for gallbladder polyps[J].Hepatobiliary Surg Nutr, 2019, 8(3):219-227.
[13]	Wang X, Zhu JA, Liu YJ, et al.Conventional ultrasound combined with contrast-enhanced ultrasound in differential diagnosis of gallbladder cholesterol and adenomatous polyps (1-2 cm)[J].J Ultrasound Med, 2022, 41(3):617-626.
[14]	Babu BI, Dennison AR, Garcea G.Management and diagnosis of gallbladder polyps: a systematic review[J].Langenbecks Arch Surg,2015, 400(4):455-462.
[15]	Szpakowski JL, Tucker LY.Outcomes of gallbladder polyps and their association with gallbladder cancer in a 20-year cohort[J].JAMA Netw Open, 2020, 3(5): e205143.
[16]	Fujiwara K, Abe A, Masatsugu T, et al.Effect of gallbladder polyp size on the prediction and detection of gallbladder cancer[J].Surg Endosc,2021, 35(9):5179-5185.
[17]	张东, 李起, 张小弟, 等.胆囊息肉手术适应证的多中心回顾性研究(附2 272例报告)[J].中华消化外科杂志, 2020, 19(8):824-834.
[18]	Mellnick VM, Menias CO, Sandrasegaran K, et al.Polypoid lesions of the gallbladder: disease spectrum with pathologic correlation[J].Radiographics, 2015, 35(2):387-399.
[19]	Liu J, Qian Y, Yang F, et al.Value of prediction model in distinguishing gallbladder adenoma from cholesterol polyp[J].J Gastroenterol Hepatol, 2022, 37(10):1893-1900.
[20]	McCain RS, Diamond A, Jones C, et al.Current practices and future prospects for the management of gallbladder polyps: a topical review[J].World J Gastroenterol, 2018, 24(26):2844-2852.
[21]	Zhang D, Li Q, Zhang X, et al.Establishment of a nomogram prediction model for long diameter 10-15 mm gallbladder polyps with malignant tendency[J].Surgery, 2021, 170(3):664-672.
[22]	Yuan HX, Yu QH, Zhang YQ, et al.Ultrasound radiomics effective for preoperative identification of true and pseudo gallbladder polyps based on spatial and morphological features[J].Front Oncol, 2020, 10:1719.
[23]	Zhang X, Wang J, Wu B, et al.A nomogram-based model and ultrasonic radiomic features for gallbladder polyp classification[J].J Gastroenterol Hepatol, 2022, 37(7):1380-1388.
[24]	石镇维, 刘再毅.影像组学研究的困境和出路[J].中华放射学杂志,2022, 56(1):9-11.
[25]	李起, 金哲川, 李孟柯, 等.影像组学在胆管恶性肿瘤诊疗中的应用现状及展望[J].中国普外基础与临床杂志, 2022, 29(12):1546-1553.
[26]	Huang T, Liu H, Lin Z, et al.Preoperative prediction of intrahepatic cholangiocarcinoma lymph node metastasis by means of machine learning: a multicenter study in China[J].BMC Cancer, 2022,22(1):931.
[27]	Alaimo L, Lima HA, Moazzam Z, et al.Development and validation of a machine-learning model to predict early recurrence of intrahepatic cholangiocarcinoma[J].Ann Surg Oncol, 2023, 30(9):5406-5415.

术前超声参数	非肿瘤性息肉	肿瘤性息肉	统计值	P值
息肉数量			χ²=20.675	＜0.001
单发	291	22
多发	221	54
息肉长径(mm)			Z=-4.694	＜0.001
10	228	15
11	97	16
12	86	17
13	48	9
14	28	5
15	25	14
息肉短径(mm)			Z=-2.595	0.009
＜6	65	8
6～9	390	49
＞9	57	19
息肉位置			χ²=6.692	0.035
颈部	39	3
体部	350	63
底部	123	10
基底			Z=3.935	0.047
带蒂	385	49
宽基底	127	27
≥4 mm	172	25
回声强度			Z=-2.690	0.007
弱	26	4
中等	220	46
强	226	26

术前超声参数	非肿瘤性息肉	肿瘤性息肉	统计值	P值
息肉数量			χ²=20.675	＜0.001
单发	291	22
多发	221	54
息肉长径(mm)			Z=-4.694	＜0.001
10	228	15
11	97	16
12	86	17
13	48	9
14	28	5
15	25	14
息肉短径(mm)			Z=-2.595	0.009
＜6	65	8
6～9	390	49
＞9	57	19
息肉位置			χ²=6.692	0.035
颈部	39	3
体部	350	63
底部	123	10
基底			Z=3.935	0.047
带蒂	385	49
宽基底	127	27
≥4 mm	172	25
回声强度			Z=-2.690	0.007
弱	26	4
中等	220	46
强	226	26

项目	训练集			测试集
项目	预测肿瘤性息肉	非肿瘤性息肉	肿瘤性息肉	预测肿瘤性息肉	非肿瘤性息肉	肿瘤性息肉
混淆矩阵（例）	否（404例）	384	20	否（101例）	94	7
	是（184例）	128	56	是（56例）	44	12
可靠性（%）	否（404例）	95	5	否（101例）	93	7
	是（184例）	70	30	是（56例）	79	21
准确率（%）	否（404例）	75	26	否（101例）	68	37
	是（184例）	25	74	是（56例）	32	63

项目	训练集			测试集
项目	预测肿瘤性息肉	非肿瘤性息肉	肿瘤性息肉	预测肿瘤性息肉	非肿瘤性息肉	肿瘤性息肉
混淆矩阵（例）	否（404例）	384	20	否（101例）	94	7
	是（184例）	128	56	是（56例）	44	12
可靠性（%）	否（404例）	95	5	否（101例）	93	7
	是（184例）	70	30	是（56例）	79	21
准确率（%）	否（404例）	75	26	否（101例）	68	37
	是（184例）	25	74	是（56例）	32	63

Please choose a citation manager

Content to export