Diagnostic value of lesion-to-spinal cord ratio and apparent diffusion coefficient in differentiating benign and...

Pulmonary nodules and masses (PNMs) are closely associated with lung cancer, for which early and accurate diagnosis is essential. However, current diagnostic methods have inherent limitations and may produce false-positive or false-negative results, thereby complicating clinical decision-making. Diffusion-weighted imaging (DWI) has emerged as a promising tool for distinguishing benign lesions from malignant lesions, although the combined diagnostic value of lesion-to-spinal cord ratio (LSR) and apparent diffusion coefficient (ADC) remains underexplored. This study aimed to evaluate the diagnostic value of LSR and ADC in distinguishing benign PNMs from malignant PNMs.

Multi-b-value DWI, commonly used for intravoxel incoherent motion (IVIM) acquisition, was retrospectively performed in patients with PNMs. Lesions were classified based on histopathological findings or clinical follow-up. ADC values were calculated from multi-b-value DWI data using a monoexponential fit, and LSR values were measured from DWI images at selected b-values. Independent-samples t-test was employed to assess statistical differences between the benign and malignant groups. Receiver operating characteristic curve analysis was conducted to determine the optimal cutoff values using the maximal Youden index and to evaluate the diagnostic performance of each parameter.

Among 56 lesions, 34 were benign and 22 were malignant. Malignant lesions exhibited significantly lower ADC values and higher LSR values compared with benign lesions (P<0.05). The area under the curve (AUC) for ADC was generally higher than that for LSR (P<0.05). Among ADC parameters, the maximum ADC value had the highest AUC [0.759, 95% confidence interval (CI): 0.631-0.887], with a sensitivity of 0.636 (95% CI: 0.407-0.828), a specificity of 0.824 (95% CI: 0.655-0.932), and a diagnostic threshold of 1.265×10^-3 mm²/s. Among the LSR parameters, LSR at b=500 s/mm² had the highest AUC (0.682, 95% CI: 0.530-0.834), with a sensitivity of 0.591 and a specificity of 0.794. The combination of ADC and LSR outperformed either parameter alone, achieving an AUC of 0.787 (95% CI: 0.658-0.885), a sensitivity of 0.909 (95% CI: 0.708-0.989), and a specificity of 0.559 (95% CI: 0.379-0.728).

Both ADC and LSR provide valuable diagnostic information for distinguishing benign PNMs from malignant PNMs. Integrating ADC and LSR further enhances diagnostic accuracy, supporting their combined application as a non-invasive, complementary imaging biomarker in clinical evaluation of PNMs.