Computed tomography features for differentiating malignant and benign focal liver lesions in dogs: A meta-analysis.

Computed tomography (CT) is often performed to complement ultrasound following detection of focal liver lesions (FLL). There is no consensus in the literature regarding the CT features that might be helpful in the distinction between benign and malignant FLL. The aim of this meta-analysis was to identify, based on the available literature, the qualitative and quantitative CT features able to distinguish between benign and malignant FLL. Studies on the diagnostic accuracy of CT in characterising FLL were searched in MEDLINE, Web of Science, and Scopus databases. Pooled sensitivity, pooled specificity, diagnostic odds ratio (DOR), receiver operator curve (ROC) area, were calculated for qualitative features. DOR were used to determine which qualitative features were most informative to detect malignancy; quantitative features were selected/identified based on standardised mean difference (SMD). Well-defined margins, presence of a capsule, abnormal lymph nodes, and heterogeneity in the arterial, portal and delayed phase were classified as informative qualitative CT features. The pooled sensitivity ranged from 0.630 (abnormal lymph nodes) to 0.786 (well-defined margins), while pooled specificity ranged from 0.643 (well-defined margins) to 0.816 (heterogeneous in delayed phase). Maximum dimensions, ellipsoid volume, attenuation of the liver in the pre-contrast phase, and attenuation of the liver in the arterial, portal, and delayed phase were found to be informative quantitative CT features. Larger maximum dimensions and volume (positive SMD), and lower attenuation values (negative SMD) were more associated with malignancy. This meta-analysis provides the evidence base for the interpreting CT imaging in the characterization of FLL.

Use of deep learning to detect cardiomegaly on thoracic radiographs in dogs.

The purpose of this study was to develop a computer-aided detection (CAD) device based on convolutional neural networks (CNNs) to detect cardiomegaly from plain radiographs in dogs. Right lateral chest radiographs (n = 1465) were retrospectively selected from archives. The radiographs were classified as having a normal cardiac silhouette (No-vertebral heart scale [VHS]-Cardiomegaly) or an enlarged cardiac silhouette (VHS-Cardiomegaly) based on the breed-specific VHS. The database was divided into a training set (1153 images) and a test set (315 images). The diagnostic accuracy of four different CNN models in the detection of cardiomegaly was calculated using the test set. All tested models had an area under the curve >0.9, demonstrating high diagnostic accuracy. There was a statistically significant difference between Model C and the remainder models (Model A vs. Model C, P = 0.0298; Model B vs. Model C, P = 0.003; Model C vs. Model D, P = 0.0018), but there were no significant differences between other combinations of models (Model A vs. Model B, P = 0.395; Model A vs. Model D, P = 0.128; Model B vs. Model D, P = 0.373). Convolutional neural networks could therefore assist veterinarians in detecting cardiomegaly in dogs from plain radiographs.

Use of transfer learning to detect diffuse degenerative hepatic diseases from ultrasound images in dogs: A methodological study.

The aim of this methodological study was to develop a deep convolutional neural network (DNN) to detect degenerative hepatic disease from ultrasound images of the liver in dogs and to compare the diagnostic accuracy of the newly developed DNN with that of serum biochemistry and cytology on the same samples, using histopathology as a standard. Dogs with suspected hepatic disease that had no prior history of neoplastic disease, no hepatic nodular pathology, no ascites and ultrasonography performed 24h prior to death were included in the study (n=52). Ultrasonography and serum biochemistry were performed as part of the routine clinical evaluation. On the basis of histopathology, dogs were categorised as 'normal' (n=8), or having 'vascular abnormalities'(n=8), or 'inflammatory'(n=0), 'neoplastic' (n=4) or 'degenerative'(n=32) disease; dogs with 'neoplastic' disease were excluded from further analysis. On cytological evaluation, dogs were categorised as 'normal' (n=11), or having 'inflammatory' (n=0), 'neoplastic' (n=4) or 'degenerative' (n=37) disease. Dogs were categorised as having 'degenerative' (n=32) or 'non-degenerative' (n=16) liver disease for analysis due to the limited sample size. The DNN was developed using a transfer learning methodology on a pre-trained neural network that was retrained and fine-tuned to our data set. The resultant DNN had a high diagnostic accuracy for degenerative liver disease (area under the curve 0.91; sensitivity 100%; specificity 82.8%). Cytology and serum biochemical markers (alanine transaminase and aspartate transaminase) had poor diagnostic accuracy in the detection of degenerative liver disease. The DNN outperformed all the other non-invasive diagnostic tests in the detection of degenerative liver disease.