Search

Abstract Knee effusion is a common comorbidity in osteoarthritis. To quantify the amount of effusion, semi quantitative assessment scales have been developed that classify fluid levels on an integer scale from 0 to 3. In this work, we investigated the use of a neural network (NN) that used MRI Osteoarthritis Knee Scores effusion-synovitis (MOAKS-ES) values to distinguish physiologic fluid levels from higher fluid levels in MR images of the knee. We evaluate its effectiveness on low-resolution images to examine its potential in low-field, low-cost MRI. We created a dense NN (dNN) for detecting effusion, defined as a nonzero MOAKS-ES score, from MRI scans. Both the training and performance evaluation of the network were conducted using public radiological data from the Osteoarthritis Initiative (OAI). The model was trained using sagittal turbo-spin-echo (TSE) MR images from 1628 knees. The accuracy was compared to VGG16, a commonly used convolutional classification network. Robustness of the dNN was assessed by adding zero-mean Gaussian noise to the test images with a standard deviation of 5–30% of the maximum test data intensity. Also, inference was performed on a test data set of 163 knees, which includes a smaller test set of 36 knees that was also assessed by a musculoskeletal radiologist and the performance of the dNN and the radiologist compared. For the larger test data set, the dNN performed with an average accuracy of 62%. In addition, the network proved robust to noise, classifying the noisy images with minimal degradation to accuracy. When given MRI scans with 5% Gaussian noise, the network performed similarly, with an average accuracy of 61%. For the smaller 36-knee test data set, assessed both by the dNN and by a radiologist, the network performed better than the radiologist on average. Classifying knee effusion from low-resolution images with a similar accuracy as a human radiologist using neural networks is feasible, suggesting automatic assessment of images from low-cost, low-field scanners as a potentially useful assessment tool.

Abstract Root phenotyping provides critical information to plant breeders for developing varieties with improved drought tolerance, greater root biomass, and greater nutrient use efficiency. Phenotyping roots in the natural environment is important for understanding the effect of the soil environment on root genotypic expressions. The goal of this work was to design and test a field‐scale mobile low‐field magnetic resonance imaging (LF‐MRI) Rhizotron that produces actionable root phenotyping data. We demonstrated this novel technology for root visualization and quantification using a LF‐MRI Rhizotron operating at 47 mT with two soil types. The LF‐MRI Rhizotron weights 453 kg, with a height of 90 cm, a diameter of 28 cm and an imaging field of view of 28 cm × 28 cm. The unit was operated in a Belk clay (Entic Hapluderts) and Weswood silt loam (Udifluventic Halustepts) generating 2‐D and 3‐D image data sets. The 2‐D image data had a collection time of 16.5 min per image at an image resolution of 2.2 mm per pixel. The 3‐D data had a collection time of 13 h per image with a 2.2 × 2.2 × 2.2 mm voxel resolution. Low‐field magnetic resonance imaging worked well for visualizing roots in moderate to high clay soils, demonstrating the potential for this technology; however, the broad application of this platform is hampered due to the prohibitively long scanning time to obtain 3‐D images. By increasing the field strength, and therefore the signal‐to‐noise ratio, faster scan times can enable a more useful system for root phenotyping.