Purpose: To prospectively compare the capability for bone metastasis assessment of whole‐body diffusion‐weighted imaging (DWI), magnetic resonance imaging (MRI) without and with DWI, [¹⁸F] fluoro‐2‐D‐glucose positron emission tomography with computed tomography (FDG‐PET/CT) and bone scintigraphy in non‐small cell carcinoma (NSCLC) patients.Materials and Methods: In all, 115 consecutive NSCLC patients (66 men, 49 women; mean age 72 years) prospectively underwent whole‐body MRI, PET/CT, and bone scintigraphy before treatment. For each method, probability of metastasis was independently assessed by using a 5‐point visual scoring system on a per‐site basis. Receiver operating characteristic (ROC)‐based positive tests were used to determine the practical threshold value for each method on a per‐site basis. Sensitivities, specificities, and accuracies were then compared on a per‐site and per‐patient basis by means of McNemar's test.Results: When the practical threshold values were adapted, specificity and accuracy of whole‐body MRI with DWI were significantly higher than those of bone scintigraphy and PET/CT (P < 0.05). On a per‐patient basis, specificity and accuracy of whole‐body MRI with DWI were significantly higher than those of bone scintigraphy (P < 0.05).Conclusion: Whole‐body MRI with DWI can be used for bone metastasis assessment of NSCLC patients as accurate as bone scintigraphy and/or PET/CT. J. Magn. Reson. Imaging 2009;30:298–308.

Objectives: This study aimed to evaluate the diagnostic performance of the lung zero-echo time (ZTE) sequence in FDG PET/MRI for detection and differentiation of lung lesions in oncologic patients in comparison with conventional two-point Dixon-based MR imaging. Methods: In this single-institution retrospective study approved by the institutional review board, 209 patients with malignancies (97 men and 112 women; age range, 17–89 years; mean age, 66.5 ± 12.9 years) underwent ¹⁸F-FDG PET/MRI between August 2017 and August 2018, with diagnostic Dixon and ZTE under respiratory gating acquired simultaneously with PET. Image analysis was performed for PET/Dixon and PET/ZTE fused images by two readers to assess the detectability and differentiation of lung lesions. The reference standard was pathological findings and/or the data from a chest CT. The detection and differentiation abilities were evaluated for all lesions and subgroups divided by lesion size and maximum standardized uptake value (SUVmax). Results: Based on the reference standard, 227 lung lesions were identified in 113 patients. The detectability of PET/ZTE was significantly better than that of PET/Dixon for overall lesions, lesions with a SUVmax less than 3.0 and lesions smaller than 4 mm (p < 0.01). The diagnostic performance of PET/ZTE was significantly better than that of PET/Dixon for overall lesions and lesions smaller than 4 mm (p < 0.01). Conclusions: ZTE can improve diagnostic performance in the detection and differentiation of both FDG-avid and non-FDG-avid lung lesions smaller than 4 mm in size, yielding a promising tool to enhance the utility of FDG PET/MRI in oncology patients with lung lesions. Key Points: • The detection rate of PET/ZTE for lesions with a SUVmax of less than 1.0 was significantly better than that of PET/Dixon. • The performance for differentiation of PET/ZTE for lesions that were even smaller than 4 mm in size were significantly better than that of PET/Dixon. • Inter-rater agreement of PET/ZTE for the differentiation of lesions less than 4 mm in size was substantial and better than that of PET/Dixon.

The usefulness of sparse-sampling CT with deep learning-based reconstruction for detection of metastasis of malignant ovarian tumors was evaluated. We obtained contrast-enhanced CT images (n = 141) of ovarian cancers from a public database, whose images were randomly divided into 71 training, 20 validation, and 50 test cases. Sparse-sampling CT images were calculated slice-by-slice by software simulation. Two deep-learning models for deep learning-based reconstruction were evaluated: Residual Encoder-Decoder Convolutional Neural Network (RED-CNN) and deeper U-net. For 50 test cases, we evaluated the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) as quantitative measures. Two radiologists independently performed a qualitative evaluation for the following points: entire CT image quality; visibility of the iliac artery; and visibility of peritoneal dissemination, liver metastasis, and lymph node metastasis. Wilcoxon signed-rank test and McNemar test were used to compare image quality and metastasis detectability between the two models, respectively. The mean PSNR and SSIM performed better with deeper U-net over RED-CNN. For all items of the visual evaluation, deeper U-net scored significantly better than RED-CNN. The metastasis detectability with deeper U-net was more than 95%. Sparse-sampling CT with deep learning-based reconstruction proved useful in detecting metastasis of malignant ovarian tumors and might contribute to reducing overall CT-radiation exposure.