前列腺癌人工神经网络诊断模型的应用研究

邱志磊; 徐立柱; 贾魁; 梁鑫; 胡伟; 郑帅帅

doi:10.13201/j.issn.1001-1420.2015.10.011

前列腺癌人工神经网络诊断模型的应用研究

- 1. 青岛市市立医院泌尿外科(青岛, 266071);
- 2. 海信集团国家重点实验室

详细信息

通讯作者: 徐立柱,E-mail:qiuzhilei2006@163.com

中图分类号: R737.25

收稿日期: 2015-02-27

Study on application of artificial neural network in the diagnosis model of prostate cancer

- 1. Department of Urology, Qingdao Municipal Hospital, Qingdao, 266071, China;
- 2. Hisense National Key Laboratory

More Information

Corresponding author: XU Lizhu, E-mail: qiuzhilei2006@163.com

Received Date: 27 February 2015

摘要

摘要: 目的:结合患者临床信息和前列腺肿瘤标记物进行数据挖掘,建立基于人工神经网络的前列腺癌诊断模型,评估其准确性,并为前列腺癌的临床诊断提供客观的参考信息。方法:回顾分析2010年1月~2014年6月在我院就诊的前列腺病患310例,其中前列腺癌组100例,非前列腺癌组210例。患者均接受经直肠前列腺穿刺活检。用210例样本(前列腺癌组70例,非前列腺癌组140例)建立人工神经网络(ANN)模型,并用100例样本(前列腺癌组30例,非前列腺癌组70例)盲法测试和评估此模型。结果:纳入分析的指标有年龄、前列腺特异性抗原(TPSA、FPSA/TPSA、PSAD)、直肠指检、前列腺体积、经直肠超声和前列腺核磁共振结果等,经过反复训练建立的ANN模型对测试样本预测的特异度为80.0%,灵敏度为93.3%。结论:ANN在计算机辅助前列腺癌的诊断,评估患者患有前列腺癌的风险,以及对穿刺检测的指导等方面具有广阔的应用前景。
- 前列腺癌 /
- 人工神经网络 /
- 诊断模型
Abstract: Objective:To establish diagnostic model for prostate cancer based on artificial neural network (ANN) by combining the clinical information and tumor makers of prostate cancer, and evaluate the performance of ANN in order to provide references for clinical diagnosis of prostate cancer.Method:Clinical data of 310 patients with prostatic disease in our institution from January 2010 to June 2014 were collected and analyzed. The cases were divided into prostate cancer group (100 cases) and non-prostate cancer group (210 cases). All the patients underwent transrectal prostate biopsy. A total of 210 cases (70 cases from prostate cancer group and 140 cases from non-prostate cancer group) were used to establish the diagnostic model with ANN and 100 samples (30 samples from prostate cancer group and 70 samples from non-prostate cancer group) were used to test and evaluate this model.Result:Age, PSA (TPSA, F/TPSA, PSAD), rectal examination, prostatic volume, the results of prostate transrectal ultrasound and magnetic resonance were used for the analysis. After repeated training of the ANN model, the specificity and sensitivity of this model were 80.0% and 93.3%.Conclusion:The use of ANN shows great promise in computer-aided diagnosis, risk assessment of prostate cancer and prostate biopsy indication.
- prostate cancer /
- artificial neural network /
- diagnosis model

HTML全文

参考文献(15)

[1]	Center M M, Jemal A, Lortet-Tieulent J,et al. International variation in prostate cancer incidence and mortality rates[J]. Eur Urol, 2012, 61(6):1079-1092.
[2]	孙颖浩. 我国前列腺癌的研究现状[J].中华泌尿外科杂志, 2004, 25(2):77-80.
[3]	Manjunath G, Jaeger H. Echo state property linked to an input:exploring a fundamental characteristic of recurrent neural networks[J]. Neural Comput,2013, 25(3):671-696.
[4]	Lawrentschuk N, Lockwood G, Davies P, et al. Predicting prostate biopsy outcome:artificial neural networks and polychotomous regression are equivalent models[J]. Int Urol Nephrol,2011,43(1):23-30.
[5]	Krogh A. What are artificial neural networks[J]? Nat Biotechnol, 2008, 26(2):195-197.
[6]	Cammann H, Jung K, Meyer H A, et al. Avoiding pit- falls in applying prediction models, as illustrated by the example of prostate cancer diagnosis[J]. Clin Chem, 2011, 57(11):1490-1498.
[7]	Patten S B. Problems encountered with the use of simulation in an attempt to enhance interpretation of a secondary data source in epidemiologic mental health research[J]. BMC Res Notes, 2010, 3:231.
[8]	Panebianco V, Sciarra A, Marcantonio A, et al. Conventional imaging and multiparametric magnetic resonance (MRI, MRS, DWI, MRP) in the diagnosis of prostate cancer[J]. Q J Nucl Med Mol Imaging, 2012, 56(4):331-342.
[9]	Regnier-Coudert O, McCall J, Lothian R, et al. Machine learning for improved pathological staging of prostate cancer:a performance comparison on a range of classifiers[J]. Artif Intell Med, 2012, 55(1):25-35.
[10]	Stephan C, Vincendeau S, Houlgatte A, et al. Multicenter evaluation of[-2] proprostate-specific antigen and the prostate health index for detecting prostate cancer[J]. Clin Chem, 2013, 59(1):306-314.
[11]	Stephan C, Cammann H, Deger S, et al. Benign prostatic hyperplasia-associated free prostate-specific antigen improves detection of prostate cancer in an artificial neural network[J].Urology, 2009, 74(4):873-877.
[12]	Porter C R, Crawford E D. Combining artificial neural networks and transrectal ultrasound in the diagnosis of prostate cancer[J]. Oncology (Williston Park), 2003, 17(10):1395-1399; discussion 1399, 1403-1406. Review.
[13]	Matsui Y, Utsunomiya N, Ichioka K, et al. The use of artificial neural network analysis to improve the predictive accuracy of pros- tate biopsy in the Japanese population[J]. Jpn J Clin Oncol, 2004, 34(10):602-607.
[14]	Remzi M, Anagnostou T, Ravery V, et al. An artificial neural network to predict the outcome of repeat prostate biopsies[J]. Urology, 2003, 62(3):456-460.
[15]	Meijer R P, Gemen E F, van Onna I E, et al. The value of an artificial neural network in the decision making for prostate biopsies[J]. World J Urol, 2009, 27(5):593-598.