BACKGROUND AND PURPOSE: The surgical approach to parotid tumors is different for benign and malignant neoplasms, but the clinical symptoms do not correlate well with histology. Difficulties in tumor classification also arise in imaging modalities, in which sonography has the lowest and MR imaging, the highest accuracy. The purpose of this study was to review our experience using conventional MR imaging of the neck in the evaluation of parotid tumors and to evaluate which MR imaging findings are best able to predict malignant histology. MATERIALS AND METHODS: Eighty-four consecutive patients (43 males, 41 females; median age, 56 years; range, 9–85 years) with parotid gland tumors who underwent MR imaging before surgery were prospectively included in the present study and retrospectively analyzed. Histology was available for all tumors. We analyzed the following MR imaging parameters: signal intensity, contrast enhancement, lesion margins (well-defined versus ill-defined), lesion location (deep/superficial lobe), growth pattern (focal, multifocal, or diffuse), and extension into neighboring structures, perineural spread, and lymphadenopathy. RESULTS: The 57 (68%) benign and 27 (32%) malignant tumors consisted of 29 pleomorphic adenomas, 17 Warthin tumors, 11 various benign tumors, 5 mucoepidermoid carcinomas, 3 adenoid cystic carcinomas, 1 acinic cell carcinoma, 1 carcinoma ex pleomorphic adenoma, 9 metastases, and 8 various malignant neoplasms. Specific signs predictive of malignancy were the following: T2 hypointensity of the parotid tumor (P = .048), ill-defined margins (P = .001), diffuse growth (P = .012), infiltration of subcutaneous tissue (P = .0034), and lymphadenopathy (P = .012). CONCLUSIONS: Low signal intensity on T2-weighted images and postcontrast ill-defined margins of a parotid tumor are highly suggestive of malignancy.

Objectives: The objective of this study is to assess the performance of a computer-aided diagnosis (CAD) system (INTACT system) for the automatic classification of high-resolution computed tomography images into 4 radiological diagnostic categories and to compare this with the performance of radiologists on the same task. Materials and Methods: For the comparison, a total of 105 cases of pulmonary fibrosis were studied (54 cases of nonspecific interstitial pneumonia and 51 cases of usual interstitial pneumonia). All diagnoses were interstitial lung disease board consensus diagnoses (radiologically or histologically proven cases) and were retrospectively selected from our database. Two subspecialized chest radiologists made a consensual ground truth radiological diagnosis, according to the Fleischner Society recommendations. A comparison analysis was performed between the INTACT system and 2 other radiologists with different years of experience (readers 1 and 2). The INTACT system consists of a sequential pipeline in which first the anatomical structures of the lung are segmented, then the various types of pathological lung tissue are identified and characterized, and this information is then fed to a random forest classifier able to recommend a radiological diagnosis. Results: Reader 1, reader 2, and INTACT achieved similar accuracy for classifying pulmonary fibrosis into the original 4 categories: 0.6, 0.54, and 0.56, respectively, with P > 0.45. The INTACT system achieved an F-score (harmonic mean for precision and recall) of 0.56, whereas the 2 readers, on average, achieved 0.57 (P = 0.991). For the pooled classification (2 groups, with and without the need for biopsy), reader 1, reader 2, and CAD had similar accuracies of 0.81, 0.70, and 0.81, respectively. The F-score was again similar for the CAD system and the radiologists. The CAD system and the average reader reached F-scores of 0.80 and 0.79 (P = 0.898). Conclusions: We found that a computer-aided detection algorithm based on machine learning was able to classify idiopathic pulmonary fibrosis with similar accuracy to a human reader.

Early and accurate diagnosis of interstitial lung diseases (ILDs) is crucial for making treatment decisions, but can be challenging even for experienced radiologists. The diagnostic procedure is based on the detection and recognition of the different ILD pathologies in thoracic CT scans, yet their manifestation often appears similar. In this study, we propose the use of a deep purely convolutional neural network for the semantic segmentation of ILD patterns, as the basic component of a computer aided diagnosis (CAD) system for ILDs. The proposed CNN, which consists of convolutional layers with dilated filters, takes as input a lung CT image of arbitrary size and outputs the corresponding label map. We trained and tested the network on a dataset of 172 sparsely annotated CT scans, within a cross-validation scheme. The training was performed in an end-to-end and semi-supervised fashion, utilizing both labeled and non-labeled image regions. The experimental results show significant performance improvement with respect to the state of the art.

The purpose of this article is to provide a comprehensive overview of imaging findings in patients with hepatic fibrosis and cirrhosis; and to describe which radiological/clinical modality is best for staging hepatic fibrosis. MR elastography (MRE) appears to be the most reliable method for grading liver fibrosis, although the CT fibrosis score derived from the combination of caudate-to-right-lobe ratio and the diameters of the liver veins significantly correlates with the stage of fibrosis.