In this work we describe a Convolutional Neural Network (CNN) to accurately predict image quality without a reference image. Taking image patches as input, the CNN works in the spatial domain without using hand-crafted features that are employed by most previous methods. The network consists of one convolutional layer with max and min pooling, two fully connected layers and an output node. Within the network structure, feature learning and regression are integrated into one optimization process, which leads to a more effective model for estimating image quality. This approach achieves state of the art performance on the LIVE dataset and shows excellent generalization ability in cross dataset experiments. Further experiments on images with local distortions demonstrate the local quality estimation ability of our CNN, which is rarely reported in previous literature.

No abstract

In this paper, we present an efficient general-purpose objective no-reference (NR) image quality assessment (IQA) framework based on unsupervised feature learning. The goal is to build a computational model to automatically predict human perceived image quality without a reference image and without knowing the distortion present in the image. Previous approaches for this problem typically rely on hand-crafted features which are carefully designed based on prior knowledge.In contrast, we use raw-image-patches extracted from a set of unlabeled images to learn a dictionary in an unsupervised manner. We use soft-assignment coding with max pooling to obtain effective image representations for quality estimation. The proposed algorithm is very computationally appealing, using raw image patches as local descriptors and using soft-assignment for encoding. Furthermore, unlike previous methods, our unsupervised feature learning strategy enables our method to adapt to different domains. CORNIA (Codebook Representation for No-Reference Image Assessment) is tested on LIVE database and shown to perform statistically better than the full-reference quality measure, structural similarity index (SSIM) and is shown to be comparable to state-of-the-art general purpose NR-IQA algorithms.

ObjectiveTo investigate the stiffness values obtained by acoustic radiation force impulse (ARFI) quantification in assessing renal histological fibrosis of chronic kidney disease (CKD).Methods163 patients with CKD and 32 healthy volunteers were enrolled between June 2013 and April 2014. ARFI quantification, given as shear wave velocity (SWV), was performed to measure renal parenchyma stiffness. Diagnostic performance of ARFI imaging and conventional ultrasound (US) were compared with histologic scores at renal biopsy. Intra- and inter-observer reliability of SWV measurement was analyzed.ResultsIn CKD patients, SWV measurements correlated significantly with pathological parameters (r = −0.422–−0.511, P<0.001), serum creatinine (r = −0.503, P<0.001), and glomerular filtration rate (r = 0.587, P<0.001). The mean SWV in kidneys with severely impaired (histologic score: ≥19 points) was significant lower than that mildly impaired (histologic score: ≤9 points), moderately impaired (histologic score: 10–18 points), and control groups (all P<0.001). Receiver operating characteristic (ROC) curves analyses indicated that the area under the ROC curve for the diagnosis of renal histological fibrosis using ARFI imaging was superior to these conventional US parameters. Using the optimal cut-off value of 2.65 m/s for the diagnosis of mildly impaired kidneys, 2.50 m/s for moderately impaired kidneys, and 2.33 m/s for severely impaired kidneys, the corresponding area under the ROC curves were 0.735, 0.744, and 0.895, respectively. Intra- and intre-observer agreement of SWV measurements were 0.709 (95% CI: 0.390–0.859, P<0.001) and 0.627 (95% CI: 0.233–0.818, P = 0.004), respectively.ConclusionsARFI may be an effective tool for evaluating renal histological fibrosis in CKD patients.

No abstract