BACKGROUND AND OBJECTIVE: The aim of this study was to develop a simple prediction model for the underlying diagnosis of solitary pulmonary nodules (SPN) based on clinical characteristics and thin-section CT findings. METHODS: Retrospective analysis was carried out on 452 patients with SPN (113 benign and 339 malignant) smaller than 30 mm, who underwent thin-section CT followed by surgical resection and histological diagnosis. The clinical characteristics were collected from medical records, and radiographic characteristics from thin-section CT findings. The prediction model was determined using multivariate logistic analysis. The prediction model was validated in 148 consecutive patients with undiagnosed SPN, and the diagnostic accuracy of the model was compared with that of an experienced chest radiologist. RESULTS: The prediction model comprised the level of serum CRP, the level of carcinoembryonic antigen, the presence or absence of calcification, spiculation and CT bronchus sign. The areas under the receiver-operating characteristic curve in training and validation sets were 0.966 and 0.840, respectively. The diagnostic accuracies of the prediction model and the experienced chest radiologist for the validation set were 0.858 and 0.905, respectively. CONCLUSION: The simple prediction model consisted of two biochemical and three radiographic characteristics. The diagnostic accuracy of an experienced chest radiologist was higher compared with the prediction model.
BACKGROUND AND OBJECTIVE: The aim of this study was to develop a simple prediction model for the underlying diagnosis of solitary pulmonary nodules (SPN) based on clinical characteristics and thin-section CT findings. METHODS: Retrospective analysis was carried out on 452 patients with SPN (113 benign and 339 malignant) smaller than 30 mm, who underwent thin-section CT followed by surgical resection and histological diagnosis. The clinical characteristics were collected from medical records, and radiographic characteristics from thin-section CT findings. The prediction model was determined using multivariate logistic analysis. The prediction model was validated in 148 consecutive patients with undiagnosed SPN, and the diagnostic accuracy of the model was compared with that of an experienced chest radiologist. RESULTS: The prediction model comprised the level of serum CRP, the level of carcinoembryonic antigen, the presence or absence of calcification, spiculation and CT bronchus sign. The areas under the receiver-operating characteristic curve in training and validation sets were 0.966 and 0.840, respectively. The diagnostic accuracies of the prediction model and the experienced chest radiologist for the validation set were 0.858 and 0.905, respectively. CONCLUSION: The simple prediction model consisted of two biochemical and three radiographic characteristics. The diagnostic accuracy of an experienced chest radiologist was higher compared with the prediction model.
Authors: Chad V Pecot; Ming Li; Xueqiong J Zhang; Rama Rajanbabu; Ciara Calitri; Aaron Bungum; James R Jett; Joe B Putnam; Carol Callaway-Lane; Steve Deppen; Eric L Grogan; David P Carbone; John A Worrell; Karel G M Moons; Yu Shyr; Pierre P Massion Journal: Cancer Epidemiol Biomarkers Prev Date: 2012-02-28 Impact factor: 4.254
Authors: Anil Vachani; Nichole T Tanner; Jyoti Aggarwal; Charles Mathews; Paul Kearney; Kenneth C Fang; Gerard Silvestri; Gregory B Diette Journal: Ann Am Thorac Soc Date: 2014-12
Authors: Alex A Balekian; Gerard A Silvestri; Suzanne M Simkovich; Peter J Mestaz; Gillian D Sanders; Jamie Daniel; Jackie Porcel; Michael K Gould Journal: Ann Am Thorac Soc Date: 2013-12
Authors: Dan Bai; Cheng-Shiun Leu; Joanne E Mantell; Theresa M Exner; Diane Cooper; Susie Hoffman; Elizabeth A Kelvin; Landon Myer; Debbie Constant; Jennifer Moodley Journal: AIDS Behav Date: 2017-02
Authors: Michael K Gould; Jessica Donington; William R Lynch; Peter J Mazzone; David E Midthun; David P Naidich; Renda Soylemez Wiener Journal: Chest Date: 2013-05 Impact factor: 9.410