Markus B. Huber

Meniscal Measurements of T1ρ and T2 at MR Imaging in Healthy Subjects and Patients with Osteoarthritis

PURPOSE To prospectively evaluate differences in T1(rho) (T1 relaxation time in the rotating frame) and T2 values in the meniscus at magnetic resonance (MR) imaging in both patients with varying degrees of osteoarthritis (OA) and healthy control subjects. MATERIALS AND METHODS The study was institutional review board approved and HIPAA compliant. Written informed consent was obtained from all subjects. T1(rho) and T2 measurements were performed at 3.0-T MR imaging in 60 subjects deemed to be healthy (n = 23; mean age, 34.1 years +/- 10.0 [standard deviation]; age range, 23-59 years), having mild OA (n = 27; mean age, 52.5 years +/- 10.9; age range, 32-69 years), or having severe OA (n = 10; mean age, 61.6 years +/- 11.6; age range, 50-86 years). Semiautomatic segmentation was performed to generate T1(rho) and T2 maps of the menisci. Clinical findings were assessed by using Western Ontario and McMaster Osteoarthritis (WOMAC) questionnaires. Differences in T1(rho) and T2 values between the three subject groups were calculated by using two-tailed t tests (with P < .05 indicating significance), and receiver operating characteristic analyses were performed. Correlations of meniscal T1(rho) and T2 values with age, cartilage-derived T1(rho) and T2 parameters, and WOMAC scores were calculated. RESULTS Significant differences between the three subject groups were found: Mean T1(rho) values were 14.7 msec +/- 5.5, 16.1 msec +/- 6.6, and 19.3 msec +/- 7.6 for the healthy, mild OA, and severe OA groups, respectively. Mean T2 values were 11.4 msec +/- 3.9, 13.5 msec +/- 4.7, and 16.6 msec +/- 8.2 for the healthy, mild OA, and severe OA groups, respectively. Correlations of meniscal T1(rho) and T2 values with subject age (R(2) = 0.18, for correlation with T2 only), cartilage-derived parameters (R(2) = 0.14-0.29), and WOMAC scores (R(2) = 0.11-0.45) were significant. CONCLUSION Meniscal T1(rho) and T2 values correlate with clinical findings of OA and can be used to differentiate healthy subjects from patients with mild or severe OA.

Proximal Femur Specimens: Automated 3D Trabecular Bone Mineral Density Analysis at Multidetector CT—Correlation with Biomechanical Strength Measurement

PURPOSE To prospectively evaluate an automated volume of interest (VOI)-fitting algorithm for quantitative computed tomography (CT) of proximal femur specimens, correlate bone mineral density (BMD) with biomechanically determined bone strength in vitro, and compare that correlation with those observed at dual-energy x-ray absorptiometry (DXA) measurement of BMD. MATERIALS AND METHODS The study was compliant with institutional and legislative requirements; donors had dedicated their body for education and research before death. Multidetector CT and DXA scans were acquired in 178 proximal femur specimens harvested from human cadavers (91 women, 87 men; mean age at death, 79 years +/- 10.2; range, 52-100 years). An automated VOI-fitting algorithm was used to calculate BMD and bone mineral content (BMC) in the head, neck, and trochanter from CT findings and pixel distribution parameters. The femur failure load (FL) was determined by using a mechanical test. Quantitative CT BMD, quantitative CT pixel distribution parameters, DXA BMD, and FL were correlated at multiple regression analysis. RESULTS Mean precision errors in quantitative CT BMD measurements at segmentation with repositioning were 0.56%, 2.26%, and 0.61% for the head, neck, and trochanter, respectively. For the head, neck, and trochanter, respectively, r values were 0.77, 0.53, and 0.59 for the correlation between quantitative CT BMD and FL and 0.74, 0.55, and 0.65 for the correlation between quantitative CT BMC and FL (P < .001). Values ranged from 0.77 to 0.80 for correlations between DXA BMD and FL and from 0.73 to 0.82 for correlations between DXA BMC and FL (P < .001). In a multiple regression model that included quantitative CT pixel distributions, adjusted multivariate correlation coefficient values for correlations with FL increased to up to 0.88. CONCLUSION Regional BMD of the proximal femur can be determined in vitro from quantitative CT data with high precision by using an automated VOI-fitting algorithm. The best multiple regression model for predicting FL included DXA BMD and regional quantitative CT BMD measurements.

Classification of Small Lesions in Breast MRI: Evaluating The Role of Dynamically Extracted Texture Features Through Feature Selection

Dynamic texture quantification, i.e., extracting texture features from the lesion enhancement pattern in all available post-contrast images, has not been evaluated in terms of its ability to classify small lesions. This study investigates the classification performance achieved with texture features extracted from all five post-contrast images of lesions (mean lesion diameter of 1.1 cm) annotated in dynamic breast magnetic resonance imaging exams. Sixty lesions are characterized dynamically using Haralick texture features. The texture features are then used in a classification task with support vector regression and a fuzzy k-nearest neighbor classifier; free parameters of these classifiers are optimized using random sub-sampling cross-validation. Classifier performance is determined through receiver-operator characteristic (ROC) analysis, specifically through computation of the area under the ROC curve (AUC). Mutual information is used to evaluate the contribution of texture features extracted from different post-contrast stages to classifier performance. Significant improvements (p < 0.05) are observed for six of the thirteen texture features when the lesion enhancement pattern is quantified using the proposed approach of dynamic texture quantification. The highest AUC value observed (0.82) is achieved with texture features responsible for capturing aspects of lesion heterogeneity. Mutual information analysis reveals that texture features extracted from the third and fourth post-contrast images contributed most to the observed improvement in classifier performance. These results show that the performance of automated character classification with small lesions can be significantly improved through dynamic texture quantification of the lesion enhancement pattern.

Prediction of Biomechanical Properties of Trabecular Bone in MR Images With Geometric Features and Support Vector Regression

Whole knee joint MR image datasets were used to compare the performance of geometric trabecular bone features and advanced machine learning techniques in predicting biomechanical strength properties measured on the corresponding ex vivo specimens. Changes of trabecular bone structure throughout the proximal tibia are indicative of several musculoskeletal disorders involving changes in the bone quality and the surrounding soft tissue. Recent studies have shown that MR imaging also allows non-invasive 3-D characterization of bone microstructure. Sophisticated features like the scaling index method (SIM) can estimate local structural and geometric properties of the trabecular bone and may improve the ability of MR imaging to determine local bone quality in vivo. A set of 67 bone cubes was extracted from knee specimens and their biomechanical strength estimated by the yield stress (YS) [in MPa] was determined through mechanical testing. The regional apparent bone volume fraction (BVF) and SIM derived features were calculated for each bone cube. A linear multiregression analysis (MultiReg) and a optimized support vector regression (SVR) algorithm were used to predict the YS from the image features. The prediction accuracy was measured by the root mean square error (RMSE) for each image feature on independent test sets. The best prediction result with the lowest prediction error of RMSE = 1.021 MPa was obtained with a combination of BVF and SIM features and by using SVR. The prediction accuracy with only SIM features and SVR (RMSE = 1.023 MPa) was still significantly better than BVF alone and MultiReg (RMSE = 1.073 MPa). The current study demonstrates that the combination of sophisticated bone structure features and supervised learning techniques can improve MR-based determination of trabecular bone quality.

Computer-Aided Diagnosis in Phase Contrast Imaging X-Ray Computed Tomography for Quantitative Characterization of ex vivo Human Patellar Cartilage

Visualization of ex vivo human patellar cartilage matrix through the phase contrast imaging X-ray computed tomography (PCI-CT) has been previously demonstrated. Such studies revealed osteoarthritis-induced changes to chondrocyte organization in the radial zone. This study investigates the application of texture analysis to characterizing such chondrocyte patterns in the presence and absence of osteoarthritic damage. Texture features derived from Minkowski functionals (MF) and gray-level co-occurrence matrices (GLCM) were extracted from 842 regions of interest (ROI) annotated on PCI-CT images of ex vivo human patellar cartilage specimens. These texture features were subsequently used in a machine learning task with support vector regression to classify ROIs as healthy or osteoarthritic; classification performance was evaluated using the area under the receiver operating characteristic curve (AUC). The best classification performance was observed with the MF features perimeter (AUC: 0.94 ±0.08) and “Euler characteristic” (AUC: 0.94 ± 0.07), and GLCM-derived feature “Correlation” (AUC: 0.93 ± 0.07). These results suggest that such texture features can provide a detailed characterization of the chondrocyte organization in the cartilage matrix, enabling classification of cartilage as healthy or osteoarthritic with high accuracy.

Isotropic 3-dimensional fast spin echo imaging versus standard 2-dimensional imaging at 3.0 T of the knee: artificial cartilage and meniscal lesions in a porcine model.

Purpose: To compare different fat-saturated (FS) 3-dimensional (3D) intermediate-weighted (IM-w) fast spin echo (FSE) sequences with a standard FS 2-dimensional (2D) IM-w FSE sequence using a porcine in vitro model with artificially created cartilage and meniscus lesions. Methods: Using a ceramic scalpel, cartilage lesions with different depths and sizes were created in porcine knee specimens at the patella as well as the medial and lateral femoral and tibial cartilage. In addition, lateral and medial meniscal lesions were produced. Magnetic resonance imaging was performed at 3.0 T in sagittal plane using an 8-channel knee coil. A standard FS 2D IM-w FSE sequence and 3 newly developed isotropic 3D FSE sequences: (i) non-FS echo train length (ETL): 78, (ii) FS ETL: 44, and (iii) FS ETL: 44, were used. The images were independently analyzed by 4 radiologists concerning image quality (1 = optimal image quality, 4 = substantially limited quality) and absence or presence of lesions using a 5-level confidence score (1 = definite no presence of abnormality, 5 = definite presence of abnormality). Radiologists were also asked to measure diameter and categorize the depth of cartilage lesions using a modified Noyes classification. Average scores for image quality, confidence of diagnosis, and sensitivity, specificity, and accuracy were calculated. In addition, contrast-to-noise ratios were calculated. Results: Image quality was significantly (P < 0.05) lower on the 3D FSE images than on the 2D FSE images [3D (i): 1.6 (SD, 0.43); 3D (ii): 2.35 (SD, 0.7); 3D (iii): 2.35 (SD, 0.5); 2D: 1.3 (SD, 0.35)]. No significant differences in diagnostic performance were found between 3D (i) and 2D FSE sequences. However, 16% fewer lesions were correctly detected with the 3D (ii) and (iii) sequences. Sensitivity was highest for the 2D sequence, and specificity was highest for the 3D (i) sequence. Confidence scores were higher for the 3D (i) sequence than for the 2D sequence. A significant increase (P < 0.05) in correctly measured cartilage lesions size and depth was found for the 3D (i) sequence over the standard 2D FSE sequence. Conclusions: Although the 3D FSE sequence performed better in depiction and characterization of cartilage abnormalities than the standard 2D FSE sequence, we currently do not recommend to use it as substitute. For the diagnosis of meniscal defects, however, no significant improvement was found.

Computer-Aided Diagnosis for Phase-Contrast X-ray Computed Tomography: Quantitative Characterization of Human Patellar Cartilage with High-Dimensional Geometric Features

Phase-contrast computed tomography (PCI-CT) has shown tremendous potential as an imaging modality for visualizing human cartilage with high spatial resolution. Previous studies have demonstrated the ability of PCI-CT to visualize (1) structural details of the human patellar cartilage matrix and (2) changes to chondrocyte organization induced by osteoarthritis. This study investigates the use of high-dimensional geometric features in characterizing such chondrocyte patterns in the presence or absence of osteoarthritic damage. Geometrical features derived from the scaling index method (SIM) and statistical features derived from gray-level co-occurrence matrices were extracted from 842 regions of interest (ROI) annotated on PCI-CT images of ex vivo human patellar cartilage specimens. These features were subsequently used in a machine learning task with support vector regression to classify ROIs as healthy or osteoarthritic; classification performance was evaluated using the area under the receiver-operating characteristic curve (AUC). SIM-derived geometrical features exhibited the best classification performance (AUC, 0.95 ± 0.06) and were most robust to changes in ROI size. These results suggest that such geometrical features can provide a detailed characterization of the chondrocyte organization in the cartilage matrix in an automated and non-subjective manner, while also enabling classification of cartilage as healthy or osteoarthritic with high accuracy. Such features could potentially serve as imaging markers for evaluating osteoarthritis progression and its response to different therapeutic intervention strategies.

Improving bone strength prediction in human proximal femur specimens through geometrical characterization of trabecular bone microarchitecture and support vector regression

Abstract. We investigate the use of different trabecular bone descriptors and advanced machine learning techniques to complement standard bone mineral density (BMD) measures derived from dual-energy x-ray absorptiometry (DXA) for improving clinical assessment of osteoporotic fracture risk. For this purpose, volumes of interest were extracted from the head, neck, and trochanter of 146 ex vivo proximal femur specimens on multidetector computer tomography. The trabecular bone captured was characterized with (1) statistical moments of the BMD distribution, (2) geometrical features derived from the scaling index method (SIM), and (3) morphometric parameters, such as bone fraction, trabecular thickness, etc. Feature sets comprising DXA BMD and such supplemental features were used to predict the failure load (FL) of the specimens, previously determined through biomechanical testing, with multiregression and support vector regression. Prediction performance was measured by the root mean square error (RMSE); correlation with measured FL was evaluated using the coefficient of determination R2. The best prediction performance was achieved by a combination of DXA BMD and SIM-derived geometric features derived from the femoral head (RMSE: 0.869±0.121, R2: 0.68±0.079), which was significantly better than DXA BMD alone (RMSE: 0.948±0.119, R2: 0.61±0.101) (p<10−4). For multivariate feature sets, SVR outperformed multiregression (p<0.05). These results suggest that supplementing standard DXA BMD measurements with sophisticated femoral trabecular bone characterization and supervised learning techniques can significantly improve biomechanical strength prediction in proximal femur specimens.

Volumetric characterization of human patellar cartilage matrix on phase contrast x-ray computed tomography

Phase contrast X-ray computed tomography (PCI-CT) has recently emerged as a novel imaging technique that allows visualization of cartilage soft tissue, subsequent examination of chondrocyte patterns, and their correlation to osteoarthritis. Previous studies have shown that 2D texture features are effective at distinguishing between healthy and osteoarthritic regions of interest annotated in the radial zone of cartilage matrix on PCI-CT images. In this study, we further extend the texture analysis to 3D and investigate the ability of volumetric texture features at characterizing chondrocyte patterns in the cartilage matrix for purposes of classification. Here, we extracted volumetric texture features derived from Minkowski Functionals and gray-level co-occurrence matrices (GLCM) from 496 volumes of interest (VOI) annotated on PCI-CT images of human patellar cartilage specimens. The extracted features were then used in a machine-learning task involving support vector regression to classify ROIs as healthy or osteoarthritic. Classification performance was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC). The best classification performance was observed with GLCM features correlation (AUC = 0.83 ± 0.06) and homogeneity (AUC = 0.82 ± 0.07), which significantly outperformed all Minkowski Functionals (p < 0.05). These results suggest that such quantitative analysis of chondrocyte patterns in human patellar cartilage matrix involving GLCM-derived statistical features can distinguish between healthy and osteoarthritic tissue with high accuracy.

Classifying small lesions on breast MRI through dynamic enhancement pattern characterization

Dynamic characterization of the lesion enhancement pattern can improve the classification performance of small diagnostically challenging lesions on dynamic-contrast enhanced MRI. This involves extraction of texture features from all post-contrast images of the lesion rather than using the first post-contrast image alone. In this study, statistical texture features derived from gray-level co-occurrence matrices are extracted from all five post-contrast images of 60 lesions and then used in a supervised learning task with a support vector regressor. Our results show that this approach significantly improves the performance of classifying small lesions (p < 0.05). This suggests that such dynamic characterization of lesion enhancement has significant potential in assisting breast cancer diagnosis for small lesions.