Differential diagnosis of benign and malignant vertebral fracture on CT using deep learning.

OBJECTIVES: To evaluate the performance of deep learning using ResNet50 in differentiation of benign and malignant vertebral fracture on CT. METHODS: A dataset of 433 patients confirmed with 296 malignant and 137 benign fractures was retrospectively selected from our spinal CT image database. A senior radiologist performed visual reading to evaluate six imaging features, and three junior radiologists gave diagnostic prediction. A ROI was placed on the most abnormal vertebrae, and the smallest square bounding box was generated. The input channel into ResNet50 network was 3, including the slice with its two neighboring slices. The diagnostic performance was evaluated using 10-fold cross-validation. After obtaining the malignancy probability from all slices in a patient, the highest probability was assigned to that patient to give the final diagnosis, using the threshold of 0.5. RESULTS: Visual features such as soft tissue mass and bone destruction were highly suggestive of malignancy; the presence of a transverse fracture line was highly suggestive of a benign fracture. The reading by three radiologists with 5, 3, and 1 year of experience achieved an accuracy of 99%, 95.2%, and 92.8%, respectively. In ResNet50 analysis, the per-slice diagnostic sensitivity, specificity, and accuracy were 0.90, 0.79, and 85%. When the slices were combined to ve per-patient diagnosis, the sensitivity, specificity, and accuracy were 0.95, 0.80, and 88%. CONCLUSION: Deep learning has become an important tool for the detection of fractures on CT. In this study, ResNet50 achieved good accuracy, which can be further improved with more cases and optimized methods for future clinical implementation. KEY POINTS: • Deep learning using ResNet50 can yield a high accuracy for differential diagnosis of benign and malignant vertebral fracture on CT. • The per-slice diagnostic sensitivity, specificity, and accuracy were 0.90, 0.79, and 85% in deep learning using ResNet50 analysis. • The slices combined with per-patient diagnostic sensitivity, specificity, and accuracy were 0.95, 0.80, and 88% in deep learning using ResNet50 analysis.

T1rho and T2 mapping of ankle cartilage of female and male ballet dancers.

BACKGROUND: Since ballet dancers begin their training before skeletal maturity, accurate and non-invasive identification of cartilage diseases is clinically important. Angle-dependent analysis of T1rho and T2 sequences can be useful for quantification of the composition of cartilage. PURPOSE: To investigate the angle-dependent T1rho and T2 profiles of ankle cartilage in non-dancers and dancers. MATERIAL AND METHODS: Ten female non-dancers, ten female dancers, and 9 male dancers were evaluated using T1rho and T2 mapping sequences. Manual segmentation of talar and tibial cartilage on these images was performed by two radiologists. Inter- and intra-rater reliabilities were calculated using intraclass correlation coefficients (ICCs) and Bland-Altman analysis. Mean thickness and volume of cartilage were estimated. Angle-dependent relaxation time profiles of talar and tibial cartilage were created. RESULTS: ICCs of the number of segmented pixels were poor to excellent. Bland-Altman plots indicated that differences were associated with segment sizes. Segmented cartilage on T1rho demonstrated larger thickness and volume than those on T2 in all populations. Male dancers showed larger cartilage thickness and volume than female dancers and non-dancers. Each cartilage demonstrated angular-dependent T1rho and T2 profiles. Minimal T1rho and T2 values were observed at approximately 180°-200°; higher values were seen at the angle closer to the magic angle. Minimal T2 value of talar cartilage of dancers was larger than that of non-dancers. CONCLUSION: In this small cohort study, regional and sex variations of ankle cartilage T1rho and T2 values in dancers and non-dancers were demonstrated using an angle-dependent approach.

Diagnosis of spinal lesions using perfusion parameters measured by DCE-MRI and metabolism parameters measured by PET/CT.

PURPOSE: To investigate the correlation of parameters measured by dynamic-contrast-enhanced MRI (DCE-MRI) and 18F-FDG PET/CT in spinal tumors, and their role in differential diagnosis. METHODS: A total of 49 patients with pathologically confirmed spinal tumors, including 38 malignant, six benign and five borderline tumors, were analyzed. The MRI and PET/CT were done within 3 days, before biopsy. On MRI, the ROI was manually placed on area showing the strongest enhancement to measure pharmacokinetic parameters Ktrans and kep. On PET, the maximum standardized uptake value SUVmax was measured. The parameters in different histological groups were compared. ROC was performed to differentiate between the two largest subtypes, metastases and plasmacytomas. Spearman rank correlation was performed to compare DCE-MRI and PET/CT parameters. RESULTS: The Ktrans, kep and SUVmax were not statistically different among malignant, benign and borderline groups (P = 0.95, 0.50, 0.11). There was no significant correlation between Ktrans and SUVmax (r = - 0.20, P = 0.18), or between kep and SUVmax (r = - 0.16, P = 0.28). The kep was significantly higher in plasmacytoma than in metastasis (0.78 ± 0.17 vs. 0.61 ± 0.18, P = 0.02); in contrast, the SUVmax was significantly lower in plasmacytoma than in metastasis (5.58 ± 2.16 vs. 9.37 ± 4.26, P = 0.03). In differential diagnosis, the AUC of kep and SUVmax was 0.79 and 0.78, respectively. CONCLUSIONS: The vascular parameters measured by DCE-MRI and glucose metabolism measured by PET/CT from the most aggressive tumor area did not show a significant correlation. The results suggest they provide complementary information reflecting different aspects of the tumor, which may aid in diagnosis of spinal lesions. These slides can be retrieved under Electronic Supplementary Material.

Morphological and dynamic contrast enhanced MR imaging features for the differentiation of chordoma and giant cell tumors in the Axial Skeleton.

PURPOSE: To characterize the morphological and dynamic-contrast-enhanced (DCE) MRI features of chordoma and giant cell tumor (GCT) of bone occurring in the axial skeleton. MATERIALS AND METHODS: A total of 13 patients with chordoma and 26 patients with GCT who received conventional T1, T2, and DCE-MRI on 3 Tesla MR scanners were retrospectively identified and analyzed. Two radiologists evaluated morphological features independently, including the lesion location, expansile bone changes, vertebral compression, presence of paraspinal soft tissue mass, fibrous septa, and the signal intensity on T1WI and T2WI. The inter-observer agreement was evaluated by kappa test. The DCE kinetics was measured to obtain the initial area under curve (IAUC) and the wash-out slope; also the two-compartmental pharmacokinetic model was applied to obtain Ktrans and kep . The diagnostic accuracy was evaluated by CHAID decision tree and ROC analysis. RESULTS: Chordomas were more likely to show soft tissue mass than GCTs (13/13 = 100% versus 15/26 = 58%; P = 0.007), as well as fibrous septa (9/13 = 69% versus 0; P < 0.001). In decision tree analysis, presence of fibrous septa and lesion location yield 31/39 = 79% accuracy. The DCE-MRI pharmacokinetic parameters Ktrans and kep of GCTs were significantly higher than those of chordomas, 0.13 ± 0.65 versus 0.06 ± 0.04 (1/min) for Ktrans , 0.62 ± 0.22 versus 0.17 ± 0.12 (1/min) for kep , P < 0.001 for both. If using kep = 0.43/min as the cut-off value, it achieved 100% sensitivity and 92% specificity to differentiate chordoma from GCT, with an overall accuracy of 37/39 = 95%. The IAUC was highly correlated with Ktrans (r = 0.94), and the slope was highly correlated with kep (r = 0.95). CONCLUSION: Several morphological features were significantly different between chordoma and GCT, but their diagnostic performance was inferior to that of DCE-MRI. LEVEL OF EVIDENCE: 4 J. Magn. Reson. Imaging 2017;45:1068-1075.

Advanced Imaging Techniques of the Wrist.

OBJECTIVE: This article covers the technical aspects and clinical applications of recent advancements in wrist MRI techniques, including T2 and T1rho mapping, compressed sensing, and isotropic 3D imaging using driven equilibrium sequences, variable-flip-angle refocusing pulse sequences, and parallel imaging. The clinical applications of these techniques include the quantitative analysis of cartilage and triangular fibrocartilaginous complex (TFCC) degeneration, faster scanning times, and improved resolution of complex wrist anatomy, allowing differentiation of degenerative from traumatic TFCC tears and improved morphologic evaluation of chondromalacia. CONCLUSION: MRI of the wrist and of the musculoskeletal system has had multiple novel and exciting advancements in recent years. Several of these advancements, such as parallel imaging, are already in clinical use, and others will be entering the clinical realm in the near future. An understanding of these techniques allows one to use their advantages to greatest effect.

T1rho and T2 relaxation times of the normal adult knee meniscus at 3T: analysis of zonal differences.

BACKGROUND: Prior studies describe histological and immunohistochemical differences in collagen and proteoglycan content in different meniscal zones. The aim of this study is to evaluate horizontal and vertical zonal differentiation of T1rho and T2 relaxation times of the entire meniscus from volunteers without symptom and imaging abnormality. METHODS: Twenty volunteers age between 19 and 38 who have no knee-related clinical symptoms, and no history of prior knee surgeries were enrolled in this study. Two T1rho mapping (b-FFE T1rho and SPGR T1rho) and T2 mapping images were acquired with a 3.0-T MR scanner. Each meniscus was divided manually into superficial and deep zones for horizontal zonal analysis. The anterior and posterior horns of each meniscus were divided manually into white, red-white and red zones for vertical zonal analysis. Zonal differences of average relaxation times among each zone, and both inter- and intra-observer reproducibility were statistically analyzed. RESULTS: In horizontal zonal analysis, T1rho relaxation times of the superficial zone tended to be higher than those of the deep zone, and this difference was statistically significant in the medial meniscal segments (84.3 ms vs 76.0 ms on b-FFE, p < 0.0001 and 96.5 ms vs 91.7 ms on SPGR, p = 0.004). In vertical zonal analysis, T1rho relaxation times of the white zone tended to be higher than those of the red zone, and this difference was statistically significant in the posterior horn of the medical meniscus (88.4 ms vs 77.1 ms on b-FFE, p < 0.001 and 104.9 ms vs 96.8 ms on SPGR, p =0.001). Likewise, T2 relaxation times of the superficial zone were significantly higher than those of the deep zone (80.4 ms vs 74.4 ms in the medial meniscus, p = 0.011). T2 relaxation times of the white zone were significantly higher than those of the red zone in the medial meniscus posterior horn (96.8 ms vs 84.3 ms, p < 0.001) and lateral meniscus anterior horn (104.6 ms vs 84.2 ms, p < 0.0001). Inter-class and intra-class correlation coefficients were excellent (>0.74) or good (0.60-0.74) in all meniscal segments on both horizontal and vertical zonal analysis, except for inter-class correlation coefficients of the lateral meniscus on SPGR. Compared with SPGR T1rho images, b-FFE T1rho images demonstrated more significant zonal differentiation with higher inter- and intra-observer reproducibility. CONCLUSIONS: There are zonal differences in T1rho and T2 relaxation times of the normal meniscus.

US-localized diffuse optical tomography in breast cancer: comparison with pharmacokinetic parameters of DCE-MRI and with pathologic biomarkers.

BACKGROUND: To correlate parameters of Ultrasonography-guided Diffuse optical tomography (US-DOT) with pharmacokinetic features of Dynamic contrast-enhanced (DCE)-MRI and pathologic markers of breast cancer. METHODS: Our institutional review board approved this retrospective study and waived the requirement for informed consent. Thirty seven breast cancer patients received US-DOT and DCE-MRI with less than two weeks in between imaging sessions. The maximal total hemoglobin concentration (THC) measured by US-DOT was correlated with DCE-MRI pharmacokinetic parameters, which included K(trans), k ep and signal enhancement ratio (SER). These imaging parameters were also correlated with the pathologic biomarkers of breast cancer. RESULTS: The parameters THC and SER showed marginal positive correlation (r = 0.303, p = 0.058). Tumors with high histological grade, negative ER, and higher Ki-67 expression ≥ 20% showed statistically higher THC values compared to their counterparts (p = 0.019, 0.041, and 0.023 respectively). Triple-negative (TN) breast cancers showed statistically higher K(trans) values than non-TN cancers (p = 0.048). CONCLUSION: THC obtained from US-DOT and K(trans) obtained from DCE-MRI were associated with biomarkers indicative of a higher aggressiveness in breast cancer. Although US-DOT and DCE-MRI both measured the vascular properties of breast cancer, parameters from the two imaging modalities showed a weak association presumably due to their different contrast mechanisms and depth sensitivities.

The effects of sequential treatments on hippocampal volumes in malignant glioma patients.

Malignant gliomas (MG) are very aggressive tumors. In an effort to improve the outcome, the patients receive multi-modal therapies such as surgery, radiation and chemotherapy (temozolomide followed in many cases by bevacizumab). The survivors are affected by multiple learning and memory deficits. Greater deterioration over time in hippocampal specific cognitive tasks was shown in patients receiving bevacizumab in addition to radiation and temozolomide for a longer period of time (RTOG 0825). The rate of hippocampal atrophy in patients treated with radiation and temozolomide followed by bevacizumab is not yet determined, and is the goal of the present study. We used the serial MRIs obtained as parts of standard clinical care in patients with MG. Measurements were done using the Medical Image Processing, Analysis and Visualization (MIPAV) software. The hippocampus in the contralateral hemisphere was manually traced and measured, to avoid morphological structure changes induced by the tumor, radiation fields or surgical markers. We determined a longitudinal progression of hippocampal atrophy-with the maximum volume loss (33.26 %) for the patients that were on treatment for 5 years. There was no detectable hippocampal atrophy during the chemo-radiation followed by adjuvant temozolomide. A significant decrease in the absolute hippocampus volume was noted after 6 months of continuous bevacizumab treatment (p < 0.05). The hippocampal volume loss progressed over the next 3 years, and was higher than the one previously reported in Alzheimer disease patients. The hippocampal volume loss is minimal during the 1 month after diagnosis, when the patients receive chemo-radiation and adjuvant temozolomide. However, prolonged treatment including bevacizumab is associated with a significant rate of hippocampal volume loss.

T1rho mapping of entire femoral cartilage using depth- and angle-dependent analysis.

OBJECTIVES: To create and evaluate normalized T1rho profiles of the entire femoral cartilage in healthy subjects with three-dimensional (3D) angle- and depth-dependent analysis. METHODS: T1rho images of the knee from 20 healthy volunteers were acquired on a 3.0-T unit. Cartilage segmentation of the entire femur was performed slice-by-slice by a board-certified radiologist. The T1rho depth/angle-dependent profile was investigated by partitioning cartilage into superficial and deep layers, and angular segmentation in increments of 4° over the length of segmented cartilage. Average T1rho values were calculated with normalized T1rho profiles. Surface maps and 3D graphs were created. RESULTS: T1rho profiles have regional and depth variations, with no significant magic angle effect. Average T1rho values in the superficial layer of the femoral cartilage were higher than those in the deep layer in most locations (p < 0.05). T1rho values in the deep layer of the weight-bearing portions of the medial and lateral condyles were lower than those of the corresponding non-weight-bearing portions (p < 0.05). Surface maps and 3D graphs demonstrated that cartilage T1rho values were not homogeneous over the entire femur. CONCLUSIONS: Normalized T1rho profiles from the entire femoral cartilage will be useful for diagnosing local or early T1rho abnormalities and osteoarthritis in clinical applications. KEY POINTS: • T1rho profiles are not homogeneous over the entire femur. • There is angle- and depth-dependent variation in T1rho profiles. • There is no influence of magic angle effect on T1rho profiles. • Maps/graphs might be useful if several difficulties are solved.

In vivo quantification of lumbar disc degeneration: assessment of ADC value using a degenerative scoring system based on Pfirrmann framework.

PURPOSE: To evaluate the role of apparent diffusion coefficient (ADC) as a quantitative means to assess the degree of intervertebral disc (IVD) degeneration contextually within the framework of a widely used Pfirrmann classification rather than in a direct correlation with Pfirrmann grades. METHODS: DWI and T2-weighted (T2w) of lumbar spine were acquired from nine healthy volunteers (age range 27-62 years, mean age 45 years) with a 3T MR scanner. ADC values were obtained from each of the five lumbar discs via a pixel-by-pixel ADC calculation as well as via region of interest-averaged image intensities. Disc degeneration was assessed by a scoring system via sequential application of Pfirrmann scale and use of intensity ratio of IVD/cerebrospinal fluid in T2w for discs in each Pfirrmann grade to be further separated. RESULTS: A significant correlation was observed between degenerative scores and ADC independent of how ADC was obtained (Spearman's ρ < -0.85, P < 2 × 10(-14)). CONCLUSIONS: This study demonstrates that previously perceived as an overlap in ADC value existing between different degenerative categories based on a visual inspection can be viewed as a quantitative role of ADC in assessment of disc degeneration. This reinforces the Pfirrmann classification system but also proceeds beyond mere qualitatively determining morphologic states.

Differentiation of tuberculosis and metastatic cancer in the spine using dynamic contrast-enhanced MRI.

PURPOSE: To investigate the differences between imaging features of spinal tuberculosis (TB) and metastatic cancer measured by dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). The presentation of TB on convention MRI may not show the typical TB signs, and they may be mis-diagnosed as malignant diseases. DCE-MRI may provide additional information to help making differential diagnosis. MATERIALS AND METHODS: DCE-MRI was performed in 24 TB and 22 metastatic cancer patients. The DCE kinetic pattern was determined as "wash-out", "plateau" or "persistent enhancement". The characteristic DCE parameters were calculated from the signal intensity time course. The two-compartmental pharmacokinetic model was used to obtain K (trans), which is the parameter associated with the delivery of MR contrast agents into the lesion, and k ep, which is the parameter associated with the distribution and clearance of contrast agents from the lesion. RESULTS: Of the 24 TB, one case showed the wash-out kinetic pattern, 12 cases showed the plateau pattern, and 11 cases showed the persistent enhancement pattern. Of the 22 metastatic cancers, 12 cases showed wash-out, 7 cases showed plateau, and 3 cases showed persistent enhancement patterns. Compared to the metastatic cancer group, the TB group had a lower k ep (0.27 ± 0.15 vs. 0.49 ± 0.23 min(-1), P < 0.001). The ROC analysis showed that the area under the curve was 0.780 for k ep. CONCLUSIONS: DCE-MRI may provide additional information for differentiation between spinal TB and metastasis, when their manifestations on conventional imaging were similar.

Association between intervertebral disc degeneration and endplate perfusion studied by DCE-MRI.

PURPOSE: The goal of this study was to study the association between solute transport mechanisms in cartilaginous disc endplates and the degeneration of intervertebral discs. Intervertebral disc degeneration is a multi-factorial process. It is suspected that poor nutrient delivery to discs might be a factor leading to degeneration. Several studies suggest that defects in disc endplates could lead to poor transport of nutrients. An imaging technique assessing endplate perfusion could be a valuable tool in investigating disc degeneration. There is currently no universally accepted technique assessing endplate perfusion in vivo. METHODS: Nine adult patients exhibiting varying levels of intervertebral disc degeneration were included. MRI was used to study the association between blood perfusion in 90 lumbar disc endplates and disc degeneration in 45 lumbar discs. Solute transport mechanism through endplates was assessed indirectly by dynamic contrast enhanced (DCE) MRI. T2-weighted MRI was used for conventional Pfirrmann classification. RESULTS: A positive association was observed between Pfirrmann grades and endplate DCE-MRI enhancement. A differential enhancement between cranial and caudal endplates was also observed, which increased with Pfirrmann grades. This differential enhancement was also dependent on the lumbar level. CONCLUSIONS: Increased MRI signal enhancement in the cartilaginous endplates of degenerated discs might indicate damage to the subchondral bone of the vertebral bodies. The endplate enhancement characteristic could aid in understanding the pathophysiology of disc degeneration and planning treatment more effectively.

Magnetic Resonance Imaging Findings in High School Football Players: Brain and Cervical Spine.

Football exposes its players to traumatic brain, neck, and spinal injury. It is unknown whether the adolescent football player develops imaging abnormalities of the brain and spine that are detectable on magnetic resonance imaging (MRI). The objective of this observational study was to identify potential MRI signatures of early brain and cervical spine (c-spine) injury in high school football players. Eighteen football players (mean age, 17.0 ± 1.5 years; mean career length, 6.3 ± 4.0 years) had a baseline brain MRI, and 7 had a follow-up scan 9-42 months later. C-spine MRIs were performed on 11 of the 18 subjects, and 5 had a follow-up scan. C-spine MRIs from 12 age-matched hospital controls were also retrospectively retrieved. Brain MRIs were reviewed by a neuroradiologist, and no cerebral microbleeds were detected. Three readers (a neuroradiologist, a neurosurgeon, and an orthopedic spine surgeon) studied the cervical intervertebral discs at six different cervical levels and graded degeneration using an established five-grade scoring system. We observed no statistically significant difference in disc degeneration or any trend toward increased disc degeneration in the c-spine of football players as compared with age-matched controls. Further research is needed to validate our findings and better understand the true impact of contact sports on young athletes.

Breast cancer: evaluation of response to neoadjuvant chemotherapy with 3.0-T MR imaging.

PURPOSE: To assess how the molecular biomarker status of a breast cancer, including human epidermal growth factor receptor 2 (HER2), hormone receptors, and the proliferation marker Ki-67 status, affects the diagnosis at 3.0-T magnetic resonance (MR) imaging. MATERIALS AND METHODS: This study was approved by the institutional review board and was HIPAA compliant. Fifty patients (age range, 28-82 years; mean age, 49 years) receiving neoadjuvant chemotherapy were monitored with 3.0-T MR imaging. The longest dimension of the residual cancer was measured at MR imaging and correlated with pathologic findings. Patients were further divided into subgroups on the basis of HER2, hormone receptor, and Ki-67 status. Pathologic complete response (pCR) was defined as when there were no residual invasive cancer cells. The Pearson correlation was used to correlate MR imaging-determined and pathologic tumor size, and the unpaired t test was used to compare MR imaging-pathologic size discrepancies. RESULTS: Of the 50 women, 14 achieved pCR. There were seven false-negative diagnoses at MR imaging. The overall sensitivity, specificity, and accuracy for diagnosing invasive residual disease at MR imaging were 81%, 93%, and 84%, respectively. The mean MR imaging-pathologic size discrepancy was 0.5 cm ± 0.9 (standard deviation) for HER2-positive cancer and 2.3 cm ± 3.5 for HER2-negative cancer (P = .009). In the HER2-negative group, the size discrepancy was smaller for hormone receptor-negative than for hormone receptor-positive cancers (1.0 cm ± 1.1 vs 3.0 cm ± 4.0, P = .04). The size discrepancy was smaller in patients with 40% or greater Ki-67 expression (0.8 cm ± 1.1) than in patients with 10% or less Ki-67 expression (3.9 cm ± 5.1, P = .06). CONCLUSION: The diagnostic accuracy of breast MR imaging is better in more aggressive than in less aggressive cancers. When MR imaging is used for surgical planning, caution should be taken with HER2-negative hormone receptor-positive cancers.

Comparison of breast density measured on MR images acquired using fat-suppressed versus nonfat-suppressed sequences.

PURPOSE: To investigate the difference of MR percent breast density measured from fat-suppressed versus nonfat-suppressed imaging sequences. METHODS: Breast magnetic resonance imaging (MRI) with and without fat suppression was acquired from 38 subjects. Breasts were divided into subgroups of different morphological patterns ("central" and "intermingled" types). Breast volume, fibroglandular tissue volume, and percent density were measured. The results were compared using nonparametric statistical tests and regarded as significant at p < 0.05. RESULTS: Breast volume, fibroglandular volume, and percent density between fat-suppressed and nonfat-suppressed sequences were highly correlated. Breast volumes measured on these two sequences were almost identical. Fibroglandular tissue volume and percent density, however, had small (<5%) yet significant differences between the two sequences-they were both higher on the fat-suppressed sequence. Intraobserver variability was within 4% for both sequences and different morphological types. The fibroglandular tissue volume measured on downsampled images showed a small (<5%) yet significant difference. CONCLUSIONS: The measurement of breast density made on MRI acquired using fat-suppressed and nonfat-suppressed T1W images was about 5% difference, only slightly higher than the intraobserver variability of 3%-4%. When the density data from multiple centers were to be combined, evaluating the degree of difference is needed to take this difference into account.

Quantitative MRI Analysis of the Talocrural and Talonavicular Joints in Ballet Dancers.

The ankles of ballet dancers are routinely under heavy loading that may lead to osteoarthritic changes. It would be clinically useful to identify such pathology as early as possible in a dancer's career. Therefore, the purpose of this study was to compare quantitative measurements in magnetic resonance (MR) images of the talocrural and talonavicular joints in ballet dancers and healthy non-dancers for use in formulating prediction of chronic injury and degenerative joint disease in these locations. Quantitative measurements in MR images of the talocrural and talonavicular joints were compared in 10 female ballet dancers, 10 healthy female non-dancers, and nine male ballet dancers. Fat-suppressed density-weighted proton, T1rho, and T2 mapping images were acquired with a 3.0 T MR scanner. Medial and lateral subchondral bone distance between the tibia and talus (MSBD and LSBD), axial navicular-talus axis angle (ANT angle), sagittal talar neck angle against the posterior talocalcaneal joint (TN angle), and curvature of navicular surface at the talonavicular joint were measured on sagittal images. The medial subchondral bone distance was found to be significantly larger in female dancers than female non-dancers (4.05 mm vs. 2.75 mm, p < 0.05), whereas there were no significant differences in LSBD (2.63 mm vs. 2.63 mm, p = 0.87). Axial navicular talus angles in female dancers were significantly larger than those in female non-dancers (38.9° vs. 24.3°, p < 0.05). There was a tendency for the TN angle to be smaller and navicular curvature (NC) to be larger in female dancers compared to female non-dancers, though the differences were not significant (TN angle: 16.6° vs. 22.3°, p = 0.09, and NC: 0.186 vs. 0.165, p = 0.28). There were no significant differences in T1rho or T2 values of talonavicular joint cartilage. These results show that the bony anatomy of dancers' ankles may adapt to the stresses placed on them by ballet.

Selection of diagnostic features on breast MRI to differentiate between malignant and benign lesions using computer-aided diagnosis: differences in lesions presenting as mass and non-mass-like enhancement.

PURPOSE: To investigate methods developed for the characterisation of the morphology and enhancement kinetic features of both mass and non-mass lesions, and to determine their diagnostic performance to differentiate between malignant and benign lesions that present as mass versus non-mass types. METHODS: Quantitative analysis of morphological features and enhancement kinetic parameters of breast lesions were used to differentiate among four groups of lesions: 88 malignant (43 mass, 45 non-mass) and 28 benign (19 mass, 9 non-mass). The enhancement kinetics was measured and analysed to obtain transfer constant (K(trans)) and rate constant (k(ep)). For each mass eight shape/margin parameters and 10 enhancement texture features were obtained. For the lesions presenting as nonmass-like enhancement, only the texture parameters were obtained. An artificial neural network (ANN) was used to build the diagnostic model. RESULTS: For lesions presenting as mass, the four selected morphological features could reach an area under the ROC curve (AUC) of 0.87 in differentiating between malignant and benign lesions. The kinetic parameter (k(ep)) analysed from the hot spot of the tumour reached a comparable AUC of 0.88. The combined morphological and kinetic features improved the AUC to 0.93, with a sensitivity of 0.97 and a specificity of 0.80. For lesions presenting as non-mass-like enhancement, four texture features were selected by the ANN and achieved an AUC of 0.76. The kinetic parameter k(ep) from the hot spot only achieved an AUC of 0.59, with a low added diagnostic value. CONCLUSION: The results suggest that the quantitative diagnostic features can be used for developing automated breast CAD (computer-aided diagnosis) for mass lesions to achieve a high diagnostic performance, but more advanced algorithms are needed for diagnosis of lesions presenting as non-mass-like enhancement.

Predicting pathologic response to neoadjuvant chemotherapy in breast cancer by using MR imaging and quantitative 1H MR spectroscopy.

PURPOSE: To compare changes in the concentration of choline-containing compounds (tCho) and in tumor size at follow-up after neoadjuvant chemotherapy (NAC) between patients who achieved pathologic complete response (pCR) and those who did not (non-pCR). MATERIALS AND METHODS: This study was approved by the institutional review board and was compliant with HIPAA; each patient gave informed consent. Thirty-five patients (mean age, 48 years +/- 11 [standard deviation]; range, 29-75 years) with breast cancer were included. Treatment included doxorubicin and cyclophosphamide followed by a taxane-based regimen. Changes in tCho and tumor size in pCR versus non-pCR groups were compared by using the two-way Mann-Whitney nonparametric test. Receiver operating characteristic (ROC) analysis was performed to differentiate between them and the area under the ROC curve (AUC) was compared. RESULTS: In the pCR group, the tCho level change was greater compared with change in tumor size (P = .003 at first follow-up, P = .01 at second follow-up), but they were not significantly different in the non-pCR group. Changes in tumor size and tCho level at the first follow-up study were not significantly different between the pCR and non-pCR groups but reached significance at the second follow-up. In ROC analysis, the magnetic resonance (MR) imaging and MR spectroscopic parameters had AUCs of 0.65-0.68 at first follow-up; at second follow-up, AUC for change in tumor size was 0.9, AUC for change in tCho was 0.73. CONCLUSION: Patients who show greater reduction in tCho compared with changes in tumor size are more likely to achieve pCR. The change in tumor size halfway through therapy was the most accurate predictor of pCR.

Residual breast cancer diagnosed by MRI in patients receiving neoadjuvant chemotherapy with and without bevacizumab.

PURPOSE: To investigate the impact of antiangiogenic therapy with bevacizumab on pathological response and the diagnostic performance of magnetic resonance imaging (MRI) in breast cancer patients. METHODS: Thirty-six patients (aged 31-69 years) with breast cancer were included. Sixteen patients received neoadjuvant chemotherapy (NAC) containing bevacizumab, and 20 patients received the same NAC protocol without bevacizumab. Serial MRI studies were performed to evaluate response. All patients received surgery after completing NAC. The extent of residual disease was examined by histopathology, and classified into three types (pCR-pathologic complete response, confined nodules, and scattered cells). Fisher's exact test and general logistic regression models were applied to analyze differences between two groups. RESULTS: pCR rates and residual disease (nodular and scattered cell) patterns were comparable between the two groups. The diagnostic accuracy rate of MRI (true positive and true negative) was 13/17 (76%) for patients with bevacizumab, and 14/20 (70%) for patients without bevacizumab. The size measured on MRI was accurate for mass lesions that shrank down to nodules, showing <0.7 cm discrepancy from pathological size. For residual disease presenting as scattered cells within a large fibrotic region, MRI could not predict them correctly, resulting in a high false-negative rate and a large size discrepancy. CONCLUSION: The pathological response and the diagnostic performance of MRI are comparable between patients receiving NAC with and without bevacizumab. In both groups MRI has a limitation in detecting residual disease broken down to small foci and scattered cells/clusters. When MRI is used to evaluate the extent of residual disease for surgical treatment, the limitations, particularly for nonmass lesions, should be considered.

Differentiation of spinal metastases originated from lung and other cancers using radiomics and deep learning based on DCE-MRI.

PURPOSE: To differentiate metastatic lesions in the spine originated from primary lung cancer and other cancers using radiomics and deep learning, compared to traditional hot-spot ROI analysis. METHODS: In a retrospective review of clinical spinal MRI database with a dynamic contrast enhanced (DCE) sequence, a total of 61 patients without prior cancer diagnosis and later confirmed to have metastases (30 lung; 31 non-lung cancers) were identified. For hot-spot analysis, a manual ROI was placed to calculate three heuristic parameters from the wash-in, maximum, and wash-out phases in the DCE kinetics. For each case, the 3D tumor mask was generated by using the normalized-cut algorithm. Radiomics analysis was performed to extract histogram and texture features from three DCE parametric maps. Deep learning was performed using these maps as inputs into a conventional convolutional neural network (CNN), as well as using all 12 sets of DCE images into a convolutional long short term memory (CLSTM) network. RESULTS: For hot-spot ROI analysis, mean wash-out slope was 0.25 ±â€¯10% for lung metastases and -9.8 ±â€¯12.9% for other tumors. CHAID classification using a wash-out slope of -6.6% followed by wash-in enhancement ratio of 98% achieved a diagnostic accuracy of 0.79. Radiomics analysis using features representing tumor heterogeneity only reached the highest accuracy of 0.71. Classification using CNN achieved a mean accuracy of 0.71 ±â€¯0.043, whereas a CLSTM improved accuracy to 0.81 ±â€¯0.034. CONCLUSIONS: DCE-MRI machine-learning analysis methods have potential to predict lung cancer metastases in the spine, which may be used to guide subsequent workup for confirmed diagnosis.