V Pedoia1, J Lee2, B Norman3, T M Link4, S Majumdar5. 1. Department of Radiology and Biomedical Imaging, University of California, San Francisco, USA; Center of Digital Health Innovation (CDHI), USA. Electronic address: valentina.pedoia@ucsf.edu. 2. Department of Radiology and Biomedical Imaging, University of California, San Francisco, USA. Electronic address: Jinhee.Lee@ucsf.edu. 3. Department of Radiology and Biomedical Imaging, University of California, San Francisco, USA. Electronic address: berknorman@me.com. 4. Department of Radiology and Biomedical Imaging, University of California, San Francisco, USA. Electronic address: Thomas.Link@ucsf.edu. 5. Department of Radiology and Biomedical Imaging, University of California, San Francisco, USA; Center of Digital Health Innovation (CDHI), USA. Electronic address: Sharmila.Majumdar@ucsf.edu.
Abstract
OBJECTIVE: We aim to study to what extent conventional and deep-learning-based T2 relaxometry patterns are able to distinguish between knees with and without radiographic osteoarthritis (OA). METHODS: T2 relaxation time maps were analyzed for 4,384 subjects from the baseline Osteoarthritis Initiative (OAI) Dataset. Voxel Based Relaxometry (VBR) was used for automatic quantification and voxel-based analysis of the differences in T2 between subjects with and without radiographic OA. A Densely Connected Convolutional Neural Network (DenseNet) was trained to diagnose OA from T2 data. For comparison, more classical feature extraction techniques and shallow classifiers were used to benchmark the performance of our algorithm's results. Deep and shallow models were evaluated with and without the inclusion of risk factors. Sensitivity and Specificity values and McNemar test were used to compare the performance of the different classifiers. RESULTS: The best shallow model was obtained when the first ten Principal Components, demographics and pain score were included as features (AUC = 77.77%, Sensitivity = 67.01%, Specificity = 71.79%). In comparison, DenseNet trained on raw T2 data obtained AUC = 83.44%, Sensitivity = 76.99%, Specificity = 77.94%. McNemar test on two misclassified proportions form the shallow and deep model showed that the boost in performance was statistically significant (McNemar's chi-squared = 10.33, degree of freedom (DF) = 1, P-value = 0.0013). CONCLUSION: In this study, we presented a Magnetic Resonance Imaging (MRI)-based data-driven platform using T2 measurements to characterize radiographic OA. Our results showed that feature learning from T2 maps has potential in uncovering information that can potentially better diagnose OA than simple averages or linear patterns decomposition.
OBJECTIVE: We aim to study to what extent conventional and deep-learning-based T2 relaxometry patterns are able to distinguish between knees with and without radiographic osteoarthritis (OA). METHODS: T2 relaxation time maps were analyzed for 4,384 subjects from the baseline Osteoarthritis Initiative (OAI) Dataset. Voxel Based Relaxometry (VBR) was used for automatic quantification and voxel-based analysis of the differences in T2 between subjects with and without radiographic OA. A Densely Connected Convolutional Neural Network (DenseNet) was trained to diagnose OA from T2 data. For comparison, more classical feature extraction techniques and shallow classifiers were used to benchmark the performance of our algorithm's results. Deep and shallow models were evaluated with and without the inclusion of risk factors. Sensitivity and Specificity values and McNemar test were used to compare the performance of the different classifiers. RESULTS: The best shallow model was obtained when the first ten Principal Components, demographics and pain score were included as features (AUC = 77.77%, Sensitivity = 67.01%, Specificity = 71.79%). In comparison, DenseNet trained on raw T2 data obtained AUC = 83.44%, Sensitivity = 76.99%, Specificity = 77.94%. McNemar test on two misclassified proportions form the shallow and deep model showed that the boost in performance was statistically significant (McNemar's chi-squared = 10.33, degree of freedom (DF) = 1, P-value = 0.0013). CONCLUSION: In this study, we presented a Magnetic Resonance Imaging (MRI)-based data-driven platform using T2 measurements to characterize radiographic OA. Our results showed that feature learning from T2 maps has potential in uncovering information that can potentially better diagnose OA than simple averages or linear patterns decomposition.
Authors: Bhushan R Deshpande; Jeffrey N Katz; Daniel H Solomon; Edward H Yelin; David J Hunter; Stephen P Messier; Lisa G Suter; Elena Losina Journal: Arthritis Care Res (Hoboken) Date: 2016-11-03 Impact factor: 4.794
Authors: Julio Carballido-Gamio; Gabby B Joseph; John A Lynch; Thomas M Link; Sharmila Majumdar Journal: Magn Reson Med Date: 2010-12-02 Impact factor: 4.668
Authors: R B Souza; D Kumar; N Calixto; J Singh; J Schooler; K Subburaj; X Li; T M Link; S Majumdar Journal: Osteoarthritis Cartilage Date: 2014-04-30 Impact factor: 6.576
Authors: Gabby B Joseph; Charles E McCulloch; Jae Ho Sohn; Valentina Pedoia; Sharmila Majumdar; Thomas M Link Journal: Skeletal Radiol Date: 2021-11-04 Impact factor: 2.199
Authors: Bragi Sveinsson; Akshay S Chaudhari; Bo Zhu; Neha Koonjoo; Martin Torriani; Garry E Gold; Matthew S Rosen Journal: Radiol Artif Intell Date: 2021-05-26
Authors: Kevin A Thomas; Dominik Krzemiński; Łukasz Kidziński; Rohan Paul; Elka B Rubin; Eni Halilaj; Marianne S Black; Akshay Chaudhari; Garry E Gold; Scott L Delp Journal: Cartilage Date: 2021-09-08 Impact factor: 3.117