Radiologists and thoracic surgeons interested in differentiation of malignant and benign pulmonary lesions.The differential diagnosis of benign and malignant focal lesions of the lung is a hot and difficult problem in daily chest imaging. Computed tomography (CT) is a common choice of modality for diagnosis of solitary pulmonary lesions because of its high spatial resolution and excellent contrast between lesions and normal lung parenchyma. However, there remains an overlap on CT signs between benign and malignant lesions. The purpose of our study was to evaluate the potential of intravoxel incoherent motion (IVIM)–derived parameters as well as apparent diffusion coefficient (ADC) in differentiating solitary pulmonary lesions. Subjects Sixty-four consecutive patients ( 23 women, 41 men; mean age, 56.1 years; age range, 30-81 years) with solitary pulmonary nodules or masses(including 52 cases of lung cancer; 12 cases of benign lesions) who underwent MR imaging were enrolled in this study. All lesions were pathologically confirmed within 10 days after MRI by surgery or biopsy. The entry criteria for patients were: (a) a pulmonary nodule or mass that showed solid opacity without an air-containing area; (b) a minimum nodule diameter of 1.0 cm; (c) no previous treatment was given; (d) absence of any contraindications to MRI. Data acquisition All patients were examined with a 3.0-T MRI (Achieva, Philips Healthcare, Best, The Netherlands), using a 16 channel phased array body coil. The MRI sequences included: axial gradient echo T1-weighted (T1W) imaging, and axial and coronal turbo spin-echo T2-weighted (T2W) imaging. MR images were obtained during end-inspiration breath-holding. DWI scan with b value = 0, 600s/mm2 was used to produce apparent diffusion coefficient (ADC) values. DWI scan with 13 b value (b = 0, 5, 10, 15, 20, 25, 50, 80, 150, 300, 500, 800, 1000 s/mm2) was used to calculate IVIM derived parameters(true diffusion coefficient[D], pseudo-diffusion coefficient [D*], and perfusion fraction [f] ). Diagnostic efficiency of each parameter was analysed using receiver operating characteristic ( ROC ) curve.Diffusivity (D) of lung cancer was significantly lower than that of benign lesions (0.92±0.20×10-3 mm²/s v.s. 1.39±0.34×10-3 mm²/s; P < 0.001). The ADCb=600 s/mm² was also significantly lower for lung cancer than for benign lesions (1.24±0.29×10-3 mm²/s v.s. 1.85±0.60×10-3 mm²/s; P < 0.001). The difference of pseudo-diffusion coefficient (D*) and perfusion fraction (f) in the two groups is not statistically significant (P > 0.05). In ROC analysis, D showed the highest area under curve(AUC) for distinguishing malignant from benign lesions, 0.867(95% Confidence Interval, 0.759 to 0.939), followed by ADCb=600 s/mm², with AUC=0.817(95% Confidence Interval, 0.701 to 0.903). the optimal threshold for differentiating malignant from benign lesions was 1.25×10-3mm²/s for D, and 1.54×10-3 mm²/s for ADCb=600 s/mm². In this study, b=600 s/mm² was used to carry out a mono-exponential model calculations, because our pre-experimental results show that the diagnosis efficacy was highest when b values equeals 600 s/mm² in discriminating benign and malignant lesions in lung. Our study demonstrated that both D and ADC b=600 s/mm² showed significant difference in benign lesions and lung cancer groups, which was believed to result from high cellularity and consequent diffusion restriction. Furthermore, D outperformed ADC in the differentiation since D reduces the bias of perfusion fraction and helps to identify malignant lesions. In our study, D* and f didn’t showed statistical difference between two group. This was different from a previous study in which f was found useful in distinguishing malignant and inflammatory lesions in the lung1. Such difference may own to the different case selection and a relatively small sample size in that study.IVIM-DWI could be more helpful for distinguishing malignant from benign lesions in lung. D has the best diagnostic efficiency.