Union is strength: the combination of radiomics features and 3D-deep learning in a sole model increases diagnostic accuracy in demented patients: a whole brain 18FDG PET-CT analysis.

OBJECTIVE: FDG PET imaging plays a crucial role in the evaluation of demented patients by assessing regional cerebral glucose metabolism. In recent years, both radiomics and deep learning techniques have emerged as powerful tools for extracting valuable information from medical images. This article aims to provide a comparative analysis of radiomics features, 3D-deep learning convolutional neural network (CNN) and the fusion of them, in the evaluation of 18F-FDG PET whole brain images in patients with dementia and normal controls. METHODS: 18F-FDG brain PET and clinical score were collected in 85 patients with dementia and 125 healthy controls (HC). Patients were assigned to various form of dementia on the basis of clinical evaluation, follow-up and voxels comparison with HC using a two-sample Student's t -test, to determine the regions of brain involved. Radiomics analysis was performed on the whole brain after normalization to an optimized template. After selection using the minimum redundancy maximum relevance method and Pearson's correlation coefficients, the features obtained were added to a neural network model to find the accuracy in classifying HC and demented patients. Forty subjects not included in the training were used to test the models. The results of the three models (radiomics, 3D-CNN, combined model) were compared with each other. RESULTS: Four radiomics features were selected. The sensitivity was 100% for the three models, but the specificity was higher with radiomics and combined one (100% vs. 85%). Moreover, the classification scores were significantly higher using the combined model in both normal and demented subjects. CONCLUSION: The combination of radiomics features and 3D-CNN in a single model, applied to the whole brain 18FDG PET study, increases the accuracy in demented patients.

Integrated FDG PET/CT in patients with persistent ovarian cancer: correlation with histologic findings.

PURPOSE: To prospectively evaluate the accuracy of integrated positron emission tomography (PET) and computed tomography (CT) for depiction of persistent ovarian carcinoma after first-line treatment, with use of histologic findings as the reference standard. MATERIALS AND METHODS: Thirty-one women (mean age, 55.9 years) with ovarian carcinoma treated with primary cytoreductive surgery and followed up with platinum regimen chemotherapy were included. All 31 patients were scheduled for surgical second-look. Before surgical second-look, all patients underwent fluorodeoxyglucose (FDG) PET/CT. At PET/CT, three main categories of persistent disease were considered for data analysis: lymph nodal lesion, peritoneal lesion, and pelvic lesion. In all patients, imaging findings were compared with results of histologic examination after surgical second-look to determine the diagnostic accuracy of PET/CT in the evaluation of disease status. The kappa statistic (Cohen kappa) was used for statistical analysis. RESULTS: Seventeen (55%) of 31 patients had persistent tumor at histologic analysis after surgical second-look, and fourteen (45%) had no histologically proved tumor. The total number of lesions that was positive for tumor cells at histologic analysis was 41 (lymph nodes, n = 16; peritoneal lesions, n = 21; pelvic lesions, n = 4); maximum diameter of these lesions was 0.3-3.2 cm (mean, 1.7 cm). A correlation was found between PET/CT and histologic analysis (kappa = 0.48). The overall lesion-based sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of PET/CT were 78%, 75%, 77%, 89% and 57%, respectively. In the detection of a tumor, a size threshold could be set at 0.5 cm, as this was the largest diameter of a lesion missed at PET/CT. CONCLUSION: Integrated PET/CT depicts persistent ovarian carcinoma with a high positive predictive value.

PET/CT and breast cancer.

During the past decade, the application of positron emission tomography with [(18)F]fluoro-2-deoxy-D-glucose (FDG-PET) has remarkably improved the management of cancer patients. Nevertheless, the clinical interpretation of FDG-PET scan can be difficult for two main reasons: (1) anatomical localisation of FDG uptake is not easy, (2) normal physiological accumulation of FDG can be misinterpreted as a pathologic area. It has been demonstrated that the visual correlation of PET with morphological procedures, such as computed tomography or magnetic resonance imaging, can improve the accuracy of PET alone. However, the time interval between the two scans, the time employed by the operator and difficulties in co-registering imaging of the abdomen and pelvis make the co-registration of separately obtained images clinically difficult. A novel combined PET/CT system has been built that improves the capacity to correctly localise and interpret FDG uptake. To date only a few studies have been conducted on the potential role of PET/CT in the management of breast cancer patients, but the better performance of this technique compared with PET alone should also be relevant for breast cancer application. In this review, we evaluate the possible impact on breast cancer diagnosis of PET/CT compared with PET alone, with respect to disease re-staging, treatment monitoring, preoperative staging and primary diagnosis. In addition, the possible role of PET/CT for radiotherapy planning is evaluated.