Neural network-based computer-aided diagnosis in distinguishing malignant from benign solitary pulmonary nodules by computed tomography

( views:84, downloads:0 )
Author:
CHEN Hui()
WANG Xiao-hua()
MA Da-qing()
MA Bin-rong()
Journal Title:
CHINESE MEDICAL JOURNAL
Issue:
Volume 120, Issue 14, 2007
DOI:
Key Word:
diagnosis,computer-assisted;neural networks(computer);solitary pulmonary nodules;computed tomography;ROC curve

Abstract: Background Computer-aided diagnosis (CAD) of lung cancer is the subject of many current researches. Statistical methods and artificial neural networks have been applied to more quantitatively characterize solitary pulmonary nodules (SPNs). In this study, we developed a CAD scheme based on an artificial neural network to distinguish malignant from benign SPNs on thin-section computed tomography (CT) images, and investigated how the CAD scheme can help radiologists with different levels of experience make diagnostic decisions.Methods Two hundred thin-section CT images of SPNs with proven diagnoses (135 small peripheral lung cancers and 65 benign nodules) were analyzed. Three clinical features and nine CT signs of each case were studied by radiologists,and the indices of qualitative diagnosis were quantified. One hundred and forty nodules were selected randomly to form training samples, on which the neural network model was built. The remaining 60 nodules, forming test samples, were presented to 9 radiologists with 3-20 years of clinical experience, accompanied by standard reference images. The radiologists were asked to determine whether a nodule was malignant or benign first without and then with CAD output.Diagnostic performance was evaluated by receiver operating characteristic (ROC) analysis.Results CAD outputs on test samples had higher agreement with pathological diagnoses (Kappa=0.841, P<0.001).Compared with diagnostic results without CAD output, the average area under the ROC curve with CAD output was 0.96(P<0.001) for junior radiologists, 0.94 (P=0.014) for secondary radiologists and 0.96 (P=0.221) for senior radiologists,respectively. The differences in diagnostic performance with CAD output among the three levels of radiologists were not statistically significant (P=0.584, 0.920 and 0.707, respectively).Conclusions This CAD scheme based on an artificial neural network could improve diagnostic performance and assist radiologists in distinguishing malignant from benign SPNs on thin-section CT images.

  • [1]Dickenson BT,Baumert B.Multidetector-row CT of the solitary pulmonary nodule.Semin Roentgenol 2003; 38:158-167.
  • [2]Iwano S,Makino N,Ikeda M,Itoh S,Tadokoro M,Satake H,et al.Solitary Pulmonary differentiating malignant from benign.Clin Imaging 2004; 28:322-328.
  • [3]Henschke CI,McCauley DI,Yankelevitz DF,Naidich DP,McGuinness G,Miettinen OS,et al.Early lung cancer action project:overall design and findings from baseline screening.Lancet 1999; 354:99-105.
  • [4]Stephenson SM,Mech KF,Sardi A.Lung cancer screening with low-dose spiral computed tomography.Am Surgeon 2005;71:1015-1017.
  • [5]Leef JL 3rd,Klein IS.The solitary pulmonary nodule.Radiol Clin North Am 2002; 40:123-143.
  • [6]Li HM,Xiao XS.CT evaluation of pulmonary nodules.Chin Comput Med Imag (Chin) 2001; 7:30-41.
  • [7]Zwirewich CV,Vedal S,Miller RR,Muller NL.Solitary pulmonary nodule:high-resolution CT radiologic-pathologic correlation.Radiology 1991; 179:469-476.
  • [8]Wang MM,Wu JL,Wang YY,Li GJ,Li W.Research on HRCT-pathologic correlation in benign pulmonary nodules and peripheral lung caner.J Dalian Med Univ (Chin) 2004; 26:66-71.
  • [9]Shiraishi J,Abe H,Engelmann R,Aoyama M,Doi K.Computer-aided diagnosis to distinguish benign from malignant solitary pulmonary nodules on radiographs:Roc analysis of radiologists' performance.Radiology 2003;227:469-474.
  • [10]Mori K,Niki N,Kondo T,Kamiyama Y,Kodama T,Kawada Y,et al.Development of a novel computer-aided diagnosis system for automatic discriminant of malignant from benign solitary nodules on thin-section dynamic computed tomography.J Comput Assist Tomogr 2005; 29:215-222.
  • [11]Li F,Aoyama M,Shiraishi J,Abe H,Li Q,Suzuki K,et al.Radiologists' performance for differentiating benign from malignant lung nodules on high resolution CT using computer-estimated likelihood of malignancy.Am J Roentgenol 2004; 183:1209-1215.
  • [12]Aoyama M,Li Q,Katsuragawa S,Li F,Sone S,Doi K.Computerized scheme for determination of the likelihood measure of malignancy for pulmonary nodules on low-dose CT images.Med Phys 2003; 30:387-394.
  • [13]Matsuki Y,Nakamura K,Watanabe H,Aoki T,Nakata H,Katsuragawa S,et al.Usefulness of an artificial neural network for differentiating benign from malignant pulmonary nodules on high-resolution CT:evaluation with receiver operating characteristic analysis.Am J Roentgenol 2002; 178:657-663.
  • [14]Lee J W,Goo JM,Lee HJ,Kim JH,Kim S,Kim YT.The potential contribution of a computer-aided detection system for lung nodule detection in multidetector row computed tomography.Invest Radiol 2004; 39:649-655.
  • [15]Gurney JW,Swensn SJ.Solitary pulmonary nodules:determining the likelihood of malignancy with artificial neural analysis.Radiology 1995; 196:823-829.
  • [16]Jiang XT,Luo LM,Wang JW,Pan XM,Luo YG.A qualitative research of lung tumor based on quantitative three dimensional shape analysis.J Appl Sci (Chin) 2004; 22:217-222.
  • [17]Zhou ZH,Li N,Yang YB,Chen SF.Early stage lung cancer diagnosis based on neural network ensemble.J Comput Res Dev (Chin) 2002; 39:1248-1253.
  • [18]Chen H,Wang XH,Ma DQ.Preliminary application of neural network in differentiating benign from malignant solitary pulmonary nodule on HRCT.Beijing Biomed Eng (Chin)2005; 24:436-439.
  • [19]Zhang MM,Zhou H,Zou Y.Quantitative investigation of solitary pulmonary nodules with dynamic contrast enhanced functional CT.Chin J Radiol (Chin) 2004; 38:263-267.
  • [20]Li XS,Xiao XS,Zhang WS,Xu JX,Li HM,Liu SY,et al.The value of the time-intensity curves in diagnosis of peripheral lung cancer.Chin J Med Imaging (Chin) 2002; 10:11-14.
  • [21]Jennings SG,Winer-Muram HT,Tarver RD.Lung tumor growth:assessment with CT-comparison of diameter and cross-sliceal area with volume measurements.Radiology 2004;231:866-871.
  • [22]Reeves AP,Kressler BM.Computer-aided diagnostics.Thorac Surg Clin 2004; 14:125-133.
WanfangData CO.,Ltd All Rights Reserved
About WanfangData | Contact US
Healthcare Department, Fuxing Road NO.15, Haidian District Beijing, 100038 P.R.China
Tel:+86-010-58882616 Fax:+86-010-58882615 Email:yiyao@wanfangdata.com.cn