M E Schweitzer

Stony Brook University Hospital, Stony Brook, New York, United States

Are you M E Schweitzer?

Claim your profile

Publications (393)849.2 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Rationale and Objectives To determine if differentiation of lipoma from liposarcoma on magnetic resonance imaging can be improved using computer-assisted diagnosis (CAD). Materials and Methods Forty-four histologically proven lipomatous tumors (24 lipomas and 20 liposarcomas) were studied retrospectively. Studies were performed at 1.5T and included T1-weighted, T2-weighted, T2-fat-suppressed, short inversion time inversion recovery, and contrast-enhanced sequences. Two experienced musculoskeletal radiologists blindly and independently noted their degree of confidence in malignancy using all available images/sequences for each patient. For CAD, tumors were segmented in three dimensions using T1-weighted images. Gray-level co-occurrence and run-length matrix textural features, as well as morphological features, were extracted from each tumor volume. Combinations of shape and textural features were used to train multiple, linear discriminant analysis classifiers. We assessed sensitivity, specificity, and accuracy of each classifier for delineating lipoma from liposarcoma using 10-fold cross-validation. Diagnostic accuracy of the two radiologists was determined using contingency tables. Interreader agreement was evaluated by Cohen kappa. Results Using optimum-threshold criteria, CAD produced superior values (sensitivity, specificity, and accuracy are 85%, 96%, and 91%, respectively) compared to radiologist A (75%, 83%, and 80%) and radiologist B (80%, 75%, and 77%). Interreader agreement between radiologists was substantial (kappa [95% confidence interval] = 0.69 [0.48–0.90]). Conclusions CAD may help radiologists distinguish lipoma from liposarcoma.
    Academic Radiology. 09/2014; 21(9):1185–1194.
  • [Show abstract] [Hide abstract]
    ABSTRACT: T1ρ MRI is an imaging technique sensitive to proteoglycan (PG) content of hyaline cartilage. However, normative T1ρ values have not been established for the weightbearing cartilage of the hip, and it is not known whether it is uniform or whether there is topographic variation. Knowledge of the T1ρ profile of hyaline cartilage in the normal hip is important for establishing a baseline against which comparisons can be made to experimental and clinical arthritic subjects.
    Clinical Orthopaedics and Related Research 08/2014; · 2.79 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Purpose To characterize the incidence, location, grade, and patterns of magnetic resonance (MR) imaging findings in the tibia in asymptomatic recruits before and after 4-month basic training and to investigate whether MR imaging parameters correlated with pretraining activity levels or with future symptomatic injury. Materials and Methods This study was approved by three institutional review boards and was conducted in compliance with HIPAA requirements. Volunteers were included in the study after they signed informed consent forms. MR imaging of the tibia of 55 men entering the Israeli Special Forces was performed on recruitment day and after basic training. Ten recruits who did not perform vigorous self-training prior to and during service served as control subjects. MR imaging studies in all recruits were evaluated for presence, type, length, and location of bone stress changes in the tibia. Anthropometric measurements and activity history data were collected. Relationships between bone stress changes, physical activity, and clinical findings and between lesion size and progression were analyzed. Results Bone stress changes were seen in 35 of 55 recruits (in 26 recruits at time 0 and in nine recruits after basic training). Most bone stress changes consisted of endosteal marrow edema. Approximately 50% of bone stress changes occurred between the middle and distal thirds of the tibia. Lesion size at time 0 had significant correlation with progression. All endosteal findings smaller than 100 mm resolved or did not change, while most findings larger than 100 mm progressed. Of 10 control subjects, one had bone stress changes at time 0, and one had bone stress changes at 4 months. Conclusion Most tibial bone stress changes occurred before basic training, were usually endosteal, occurred between the middle and distal thirds of the tibia, were smaller than 100 mm, and did not progress. These findings are presumed to represent normal bone remodeling. © RSNA, 2014.
    Radiology 07/2014; · 6.34 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The purpose of this study was to determine if bone metastasis characteristics on axial skeleton MRI are associated with either skeletal-related events (SREs) or survival in breast cancer patients. A retrospective review was performed on 247 breast cancer patients with bone metastases identified on axial skeleton MRI. MRI studies were reviewed for metastases T1 signal, signal uniformity, complete vertebral metastatic marrow replacement, metastases quantity, and distribution. Odds ratio (OR) and hazard ratios (HR) were calculated, with 95 % confidence intervals (95 % CI), to determine association with either future SREs or survival. At the time of analysis, 174 (70 %) patients had developed SREs and 176 (71 %) patients were dead. Features of skeletal metastases associated with SREs included the presence of complete metastatic marrow replacement within any vertebra; OR 2.363 (95 % CI 1.240-4.504, P = 0.0090), and more widely distributed metastases; OR 1.239 (95 % CI 1.070-1.435, P = 0.0040). Features associated with shorter survival included the presence of complete metastatic marrow replacement within any vertebra; HR 1.500 (95 % CI 1.105-2.036, P = 0.0093), and more widely distributed metastases; HR 1.141 (95 % CI 1.047-1.243, P = 0.0027). Metastases T1 signal, signal uniformity, and surprisingly quantity were not associated with SREs or survival. Axial skeleton MRI was able to identify characteristics predictive of future SREs and survival. These characteristics could be used for risk stratification for future trials if prospectively validated.
    Breast Cancer Research and Treatment 07/2014; · 4.47 Impact Factor
  • Neera Malik, Shilpa Lad, Jean Seely, Mark Schweitzer
    [Show abstract] [Hide abstract]
    ABSTRACT: Objective: To determine the rate of underestimation of malignancy in patients with biopsy-proven stromal fibrosis. Methods: Following IRB approval, We retrospectively reviewed the charts of patients with biopsy-proven stromal fibrosis who underwent percutaneous biopsy over a 5 year period. The medical records and histopathology in patients who underwent repeat biopsy and/or surgical excision at the site of stromal fibrosis within two years were reviewed. Interval stability for up to two years was documented in patients who did not undergo additional biopsy/surgical excision. An upgrade was defined as any patient with biopsy-proven stromal fibrosis with evidence of malignancy at the site of biopsy within two years. Results: 365 cases of stromal fibrosis were identified, of which 25 (7%) were upgraded to in-situ or invasive malignancy on repeat biopsy or surgical excision. 7 cases were upgraded to DCIS and 18 to invasive cancer. Of upgraded cases, 8 of 24 (32%) were considered concordant with a benign diagnosis. The false negative rate, i.e. cases of stromal fibrosis concordant with benignity, but with subsequent upgrade, comprised 2% of all cases. Conclusion: In biopsy-proven cases of stromal fibrosis, there is a 7% upgrade to malignancy,and all instances of stromal fibrosis with radiology-pathology discordance should undergo repeat biopsy or surgical excision. Cases that demonstrate radiology-pathology concordance can be safely followed due to a false-negative rate of 2%.
    The British journal of radiology. 05/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: We compared preoperative and postoperative computed tomography (CT) versus radiographic imaging in the evaluation of acetabular fractures (AFs). Fifty-four patients who underwent imaging for AFs were retrospectively evaluated. Postoperative reduction quality was assessed on radiographs and CT scan by 2 observers. Rate of reintervention was noted. Radiation exposure from CT was calculated. After reduction, 24 patients had significant findings on postoperative CT. Five patients required reintervention, all of whom had significant postoperative CT findings and complex fractures. Notably, only 1 of the 5 patients had an indication for reintervention based on radiographs alone.The average dose for preoperative/postoperative CT study was 11.5/12.3 mSv, respectively, with a cumulative average dose of 23.8 mSv. Although reoperation rate is low after fixation of AFs, CT is required to identify those requiring reintervention. However, postoperative CT should be used judicially, only in patients presenting with complex acetabular fractures.
    Journal of computer assisted tomography 03/2014; · 1.38 Impact Factor
  • Conference Paper: Appendicular Trauma
    Mark E. Schweitzer
    [Show abstract] [Hide abstract]
    ABSTRACT: LEARNING OBJECTIVES 1) Review some subtle imaging findings indicative of significant derangements. 2) Review the locations where joint effusions can be reliably described radiographically. 3) Emphasize some normal variant of appendicular trauma. ABSTRACT Appendicular trauma is common and most patients are imaged initally by conventional radiography. In these patients it is important to use an a systematic approach looking at alignment and soft tissues changes, especialy the presence of effusions. It is also important to be cognizant that there are specific fractures and soft tissue injuries that are indicative of major soft tissue or visercal trauma. It is no less important to be aware that some glaring radiographic findings may be associated with less severe trauma. We will review the concepts behind and locations of osteoporotic fractures, and what defines the risk of and the locations for pathologic fractures. Specific mechanisms for specific injuries will be discussed as well.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE To determine the hip capsule thickness in patients with cam-FAI and non-FAI hip pathology using preoperative magnetic resonance imaging (MRI). The hypothesis was that in cam-FAI the capsule would be thicker related to chronic impingement of the cam deformity against the capsule. METHOD AND MATERIALS Research ethics board approval was obtained. Forty-one hips (40 patients) were included, 16 with surgically proven cam-FAI (9M,7F; age 22-58 yrs) and 25 with non-FAI chondrolabral pathology (4M,21F; age 18-63 yrs). All subjects had undergone preoperative 3T MRI including oblique axial and oblique coronal, FSE proton density weighted sequences, with parameters: FOV 180mm, Matrix 320 x 256, Slice thickness 3.5mm, TE 30ms, TR 2310ms, ETL 7, NEx=2. The hip capsule thickness was measured at two locations, anteriorly (3 o’clock) and superiorly (12 o’clock) on single oblique axial and oblique coronal images, respectively, at the thickest portion of the visualized capsule on the given slices. Comparison of the maximal hip capsule thickness between the two groups and gender analysis (two-tailed paired t-test for both), and correlation of capsule thickness with the alpha angle (Pearson correlation coefficient) were performed. A p-value <0.05 was considered significant. RESULTS The mean maximal hip capsule thicknesses(mm) at the anterior and superior locations were: cam-FAI 4.99, 6.97; Non-FAI 4.94, 6.68, respectively. There was no significant difference between the cam-FAI and non-FAI groups at either location. The mean maximal capsule thicknesses (mm) anteriorly for males, females were: cam-FAI 5.02, 4.96, respectively, and non-FAI 5.03, 4.93, respectively. There was no significant gender difference in capsule thickness within either subject group. The mean alpha angle measurements were 57.3 degrees in the cam-FAI group and 46.5 degrees in the non-FAI group (p<0.05). There was no significant correlation between the alpha angle values and the anterior capsule thickness. CONCLUSION The hip capsule thickness is no different in cam-FAI patients compared to those with non-FAI hip pathology. There is no gender difference in hip capsule thickness with either cam-FAI or non-FAI hip pathology. In cam-FAI subjects, there is no correlation between the alpha angle and hip capsule thickness. CLINICAL RELEVANCE/APPLICATION Hip capsule thickness cannot be used as a specific marker or sign of cam-FAI.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE Bone metastases from breast cancer are associated with significant morbidity and mortality. Skeletal related events (SREs) are defined as radiation/surgery to bone, pathologic fractures, malignant spinal cord compression and hypercalcemia. We hypothesized that certain skeleton MRI characteristics were associated with developing an SRE. METHOD AND MATERIALS Over a 3-year period, 250 unselected patients were identified with skeletal metastases on axial skeleton MRI, histologically proven breast cancer and no other malignancy. Two radiologists reviewed each study independently. Patients were divided into 3 groups: 1) metastatic marrow replacement, 2) > 20 skeletal metastases, and 3) ≤ 20 metastases. Marrows were divided into 7 subregions (cervical, thoracic, lumbo-sacral spine, pelvis, proximal femurs, sternum and shoulder girdles). The number of regions with metastases was summed. Presence of a pathologic fracture was documented. Charts were reviewed for the subsequent development of SREs. Logistic regression was applied to determine the odds ratio (OR) for developing an SRE, with 95% CI and p-values. RESULTS 176/250 (70%) patients developed an SRE. Presence of a prior pathologic fracture demonstrated the strongest association with developing an SRE; OR 2.506 (1.112-5.648, p=0.03). Marrow replacement and > 20 metastases were associated with developing an SRE compared with ≤ 20 metastases; OR 2.048 (1.019-4.118) and 2.176 (1.086-4.361), respectively (p=0.03). A higher sum of involved skeletal regions was associated with developing an SRE; OR 1.251 (1.081-1.448, p=0.003). Metastases involving predominantly the axial skeleton, increasing metastases quantity, metastases T1 signal and signal uniformity were not statistically associated with SREs. Increased metastatic involvement of the proximal femur showed a trend for shorter time to hip fracture (Figure). CONCLUSION Several characteristics of axial skeleton metastatic disease on MRI are associated with developing SREs. These include presence of a pathologic fracture, marrow replacing process, > 20 skeletal metastases, and more skeletal regions of involvement. These findings require prospective confirmation. CLINICAL RELEVANCE/APPLICATION Features of bone metastases on axial skeleton MRI may be of utility for predicting development of skeletal related events in patients with breast cancer.
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • Shaoyin Duan, Gina Diprimio, Mark E. Schweitzer
    [Show abstract] [Hide abstract]
    ABSTRACT: CONCLUSION Conclusion: Patients with the culture negative disc space infection have obvious marrow edema and endplate erosion, more frequency of disc fluid and long complaint days and less frequency of facet fliuid. Follow up MR at more than 1 month can find these lesions with significant absorption or shrinkage, while within 1 month follow-up maybe not. BACKGROUND Objective: As many disc space infections are culture negative, as there is increasing evidence that a subgroup of end plate reactive changes actually represent indolent infections, we sought to compare the MR findings of disc space infections with and without positive cultures EVALUATION Methods: the clinical and MR imaging of 26 patients with disc space infections, of which 9 cases (15 discs) who were blood and local tissue culture negative, and 17 cases (21 disc) who were culture positive, were retrospectively evaluated by two radiologists. All patients had the histological confirmation, the negative group have the follow up MR, performed within 1 month in 5 cases, 1-3 months in 7 and more than 3 months in 5 and responded appropriately to antibiotic treatment; the positive group were histological diagnosis as the acute inflammation. MR findings , epidemiology were compared. Statistical analysis consisted of T and chi squarred tests DISCUSSION Results: Nine patients of the negative group had the complaint times of 2 days to 4 months, lesions in single disc in 6 cases and 2-4 discs in 3. MR showed the marrow edema, endplate erosions, disc fluid, paraspinal mass and lesions CE with 100.0% sensitivity, The negative group were compared with the positive,The former is more obvious in the extent of the marrow edema and endplate erosion, and more frequency of disc fluid and longer days of complaint, while the latter is more frequency of facet fluid (P<0.05). follow-up MR of negative group showed the obvious absorption or shrinkage of marrow edema, disc fluid, paraspinal mass, epidural collection, facet fluid and the decreased lesions CE among within 1 month, 1-3 months and more than 3 months (P<0.05), but without the significant difference between the first examination and the within 1 month follow up ( P>0.05).
    Radiological Society of North America 2013 Scientific Assembly and Annual Meeting; 12/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Objective Prospective evaluation of optimal MR arthrographic concentration of two gadolinium-based agents of different relaxivities, either in saline or iodinated carriers at 1.5, 3 and 7 T field strengths was evaluated in vitro. Materials and methods At 1.5, 3.0 and 7.0 T, gadobenate and gadoteridol were studied at different concentrations in both normal saline and iodinated contrast. IRT1W and multi-echo T2FSE sequences were obtained. Signal–concentration relationship of both used agents in different carriers and field strengths was plotted from their collected T1 and T2 values, as well as their derived longitudinal (r1) and transverse (r2) relaxivities. Results Significant increase in T1 signal intensity (p < 0.001) of both gadobenate and gadoteridol was observed in higher fields of 3 T and 7 T, stronger for gadoteridol. Contrarily, gadoteridol showed a significantly lesser decrease (p < 0.0001) of T2 signal intensity with increasing field from 3 T to 7 T. Iodinated carriers significantly diminished T2 signal intensity (p < 0.0001) except for highest concentration (10 mmol/L) (p = 0.8899) while this was only significant (p = 0.0279) between extreme concentrations for T1 signal. Conclusion On transitioning to high-field (3 T and 7 T) MR arthrography, keeping the least amount of iodine for intra-articular contrast delivery, and when choice is available, gadoteridol is preferable to gadobenate. However, gadobenate was still optimal at 3 T.
    The Egyptian Journal of Radiology and Nuclear Medicine. 06/2013; 44(2):283-289.
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: Cam-type femoroacetabular impingement (FAI) deformities have been associated with early osteoarthritic degeneration of the hip. Degeneration depends on many factors such as joint morphology and dynamics of motion. Bone mineral density (BMD) appears to be a manifestation of the above, and may be a potentiater. Thus the goal of this study was to assess subchondral BMD of cam deformities in symptomatic and asymptomatic FAI subjects, and to compare to normal controls. METHODS: Subjects undergoing surgical correction of a symptomatic cam-type deformity were recruited ("Surgical"). Asymptomatic volunteers were also recruited and classified as normal ("Control") or having a deformity ("Bump") based on their alpha angle measurement. All subjects (n=12 per group) underwent CT with a calibration phantom. BMD was determined in volumes of interest around the femoral head and neck to a depth of 5 mm. BMD was compared between groups in each section using spine BMD as a covariate. RESULTS: No differences were seen between groups in the peripheral bearing surface. The Bump group exhibited higher BMD than Controls within the head/neck junction (p<0.05). When compared to normal subchondral bone in the peripheral level of Controls, BMD in the deformity was up to 78% higher in Bump subjects and up to 47% higher Surgical subjects (p<0.05). CONCLUSION: Subchondral BMD of cam deformities is higher than that of normal subchondral bone in the peripheral region of the femoral head, regardless of symptom status. The expected increased subchondral stiffness may increase contact stresses in the joint tissues leading to accelerated degeneration.
    Osteoarthritis and Cartilage 05/2013; · 4.26 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Peripheral neurolymphomatosis is a rare manifestation of advanced lymphoproliferative disorders. It is often associated with B cell lymphomas and rarely with cutaneous T cell lymphomas, such as mycosis fungoides and Sézary syndrome. In this case report, we present a 78-year-old male with a long-standing history of mycosis fungoides. The patient initially presented with chronic peripheral neuropathy in an ulnar nerve distribution. After an unsuccessful ulnar nerve transposition, the ulnar nerve was re-explored and a mass consistent with diffuse lymphomatous infiltration was diagnosed. Magnetic resonance (MR) imaging of the left brachial plexus and later of the sacral plexus demonstrated diffuse thickening and peripheral nodularity in keeping with neurolymphomatosis. The patient's clinical course rapidly deteriorated thereafter and the patient succumbed to his disease. Although uncommon, neurolymphomatosis may be considered in patients with chronic peripheral neuropathy and an underlying history of a lymphoproliferative disorder. US and MR may serve as helpful non-invasive adjuncts in making the diagnosis and identifying sites for biopsy.
    Skeletal Radiology 03/2013; · 1.74 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Sonographically guided biopsy is performed by one of two techniques: the freehand and needle-guided techniques. To our knowledge, the relationship between the location of the local anesthetic tract and the biopsy needle tract as well as direct comparison of the two biopsy techniques has not been previously validated. The aim of this study was to validate the different parameters related to the two biopsy techniques using computed tomography as the reference standard for assessing final tract positions. There were statistically significant differences between the freehand and guided techniques in the following parameters: number of passes required for contrast agent injection (P = .003), number of passes required to insert the needle (P = .005), time required to inject the anesthetic/contrast agent (P = .005), time required to insert the biopsy needle (P = .02), and distance between contrast tract and final needle position (P = .03). No statistical difference was identified for the angle between the contrast tract and needle position. This difference likely reflects the confidence of the radiologist in identifying the needle location during the procedure. Using a commercially available guide that has a fixed angle can result in a faster, more efficient, and reproducible biopsy technique compared to the freehand technique, especially for those who have less experience in performing sonographically guided biopsies.
    Journal of ultrasound in medicine: official journal of the American Institute of Ultrasound in Medicine 03/2013; 32(3):535-40. · 1.40 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: MR imaging, because of its multiplanar capability and superior soft tissue contrast resolution, is the preferred modality to assess osseous and soft tissue structures around the hip joint. This article reviews the clinical presentation, disease process, and imaging findings of important congenital and acquired osseous disorders of the pediatric and adult hip.
    Magnetic resonance imaging clinics of North America 02/2013; 21(1):111-25.
  • [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: Femoro-acetabular impingement (FAI) has been associated with significant acetabular cartilage damage and subsequent degenerative arthritis. Subchondral bone, often neglected in osteoarthritis studies, may play an important role in the degenerative cascade. Hence the goal of this study was to assess acetabular subchondral bone mineral density (BMD) in subjects with asymptomatic or symptomatic cam deformities compared to normal control subjects. The relationship between BMD and the alpha angle, a quantitative measure of the deformity, was also analyzed. METHODS: Patients diagnosed with symptomatic cam FAI were recruited ('Surgical') as well as subjects from the general asymptomatic population, classified from CT imaging as normal ('Control') or having a cam deformity ('Bump') based on their alpha angle measurement. There were 12 subjects in each group. All subjects underwent a CT scan with a calibration phantom. BMD was calculated in regions of interest around the acetabulum from CT image intensity and the phantom calibration. BMD was compared between groups using spine BMD as a covariate. The relationship between BMD and alpha angle was assessed by linear regression. RESULTS: In the antero-superior regions bone density was 15-34% higher in the Bump group (p<0.05) and 14-38% higher in the Surgical group (p<0.05) compared to Controls. BMD correlated positively with the alpha angle measurements (R(2)=0.44, p<0.001). CONCLUSION: BMD was elevated in subjects with cam-type deformities, with the severity of the deformity more correlative than symptom status. Similarities to the symptomatic group suggest that hips with an asymptomatic deformity may already be in early stages of joint degeneration.
    Osteoarthritis and Cartilage 01/2013; · 4.26 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE PET CT is the current gold standard for lymphoma response assessment , however the anatomic detail is somewhat limited . Moreover, PET is somewhat limited in distribution, of high cost, and especially important as many of these patients are young has a relatively high radiation doses. MRI has provided morphologic information in these patients, but most current protocols are unable to provide information about biologic activity. We hypothesized that diffusion weighted imaging might provide information similar to that provided by PET/CT thus we assessed the feasibility of diffusion whole body imaging, for the assessment of treatment response in patients with B cell lymphoma. METHOD AND MATERIALS Eight patients with pathology proven B cell lymphoma was prospectively recruited to participate in this study. Each scan consisted of T2 Weighted whole body, in coronal and axial as well Diffusion weighted whole body scan with 8 different b values (0, 50,100, 200, 300, 400, 600, and 800). Each patient had a pretreatment baseline scan and after 4 weeks of the first cycle of baseline chemotherapy (CHOP-R). ADC values were calculated as well the lesion homogeneity. 6 patients had a follow up at the end of their chemotherapy courses. RESULTS There was restricted diffusivity of all the lesions prior to treatment. The overall mean ADC value is 0.7± 0.4x10-3 mm2/sec. A ADC value was 0.5±0.1 x10-3 mm2/sec in 75% . After treatment the The overall mean ADC value is 1.7±0.9x10-3 mm2/sec with signal change of 54±15%. A mean ADC value increased to 1.9 ±0.8 x10-3 mm2/sec in 75% of the cases with a signal change of 57±13%.All 6 patients demonstrated complete disease remission at the end of their treatment course. These 6 patients had a second follow up at the end of chemotherapy, with ADC value 1.8±1.5 x10-3 mm2/sec . The lesion homogeneity is 0.8 ±0.1 (varying between 0 heterogeneous and 1 completely homogenous). No statistical significant difference in homogeneity between the pre and post treatment scan p= 0.05. CONCLUSION Diffusion weighed MRI is feasible in following lymphoma patients and pilot data seems to show that it may be able to provide overran assessment of treatment response. Further longitudinal study is needed to assess the predictability of these findings in assessing treatment response. CLINICAL RELEVANCE/APPLICATION Diffusion MR is feasible and shows promise in evaluating treatment response in lymphoma patients.
    Radiological Society of North America 2012 Scientific Assembly and Annual Meeting; 11/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE We sought to use post-processing analysis of margins and tumor texture to determine the ability of computer-aided diagnosis (CAD) to differentiate lipoma and liposarcoma on MRI images. METHOD AND MATERIALS Forty-four histologically proven cases, including 24 lipoma and 20 liposarcoma were retrospectively studied. Patients were age 20 - 86 years (mean = 56) and included 16 males and 28 females. All studies were performed at 1.5T scanner and included T1-weighted, T2-weighted and contrast enhanced sequences. Two experienced musculoskeletal radiologists blindly and independently assessed degree of confidence in malignancy using all available images for each patient. Seventy textural features and 73 shape features were derived based on the defined region-of-interest morphologies. For feature selection, we reduced the dimensionality of our feature space from 143 to 17 on the basis of the Fisher coefficient of each feature. Linear discriminant analysis (LDA) was used to reduce the feature space to its ‘most discriminative features’ (MDF). All textural and shape features were computed using MaZda version 4.6. Sensitivity, specificity and accuracy values of the two radiologists were calculated using contingency table and inter-reader correlation by Kappa statistics. ROC, Fisher’s exact test, and Kappa test were used for data analysis. RESULTS Using optimum-threshold criteria, the CAD method produced superior values (sensitivity: 85%, specificity: 100%, accuracy: 93%) compared to radiologist A (sensitivity: 75%, specificity: 83.3%, accuracy: 79.5%) and radiologist B(sensitivity: 84.2%, specificity: 73.9%, accuracy: 78.5%). Also, our proposed CAD system achieved a higher AUC (0.96) than either of the radiological evaluations (AUC =0.89 for each radiologist). A Fisher’s exact test determined a statistically significant difference (P=0.001) between CAD and the radiologists’evaluations. Additionally, there was moderate agreement between CAD and radiologists (kappa = 0.5). Inter-reader agreement was good (kappa =0.72). CONCLUSION In this preliminary study, texture and border features not discernable to the naked eye allowed determination of malignancy significantly better than experienced, specialized musculoskeletal radiologists. CLINICAL RELEVANCE/APPLICATION Our proposed image-processing algorithm can provide radiologists in practice with objective diagnostic information or simply a second opinion.
    Radiological Society of North America 2012 Scientific Assembly and Annual Meeting; 11/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE An increasing number of early adverse tissue reactions, including pseudotumors, that can lead to substantial soft tissue destruction and early implant failure have recently been reported in patients with metal-on-metal (MM) total hip replacements (THR). Although metal artifact reduction MRI is now effective in visualizing these reactions, it is unclear how benign and potentially aggressive ones can be differentiated. Since histological differences have been observed between MM and metal-on-polyethylene (MPE) periprosthetic tissues, a difference in the overall tissue perfusion is expected. The purpose of this study was to develop and apply a quantitative dynamic contrast-enhanced (DCE) MRI technique to measure and compare the perfusion parameter Ktrans (index of flow and permeability) near MM and MPE implants. METHOD AND MATERIALS Eight patients (55-74 y; equal #s men & women) with a THR (4 MM and 4 MPE) were imaged with DCE-MRI at 1.5T (post-op time 40 +/- 7 months). Turbo spin-echo sequences were used to minimize metal artifacts (echo train length=7, BW=349 Hz, TE=6.8 ms, 16 slices, thick=6 mm). DCE was performed with TR=865 ms, temporal resolution=7.8 s, 0.1 mmol/kg Gd contrast agent. Inversion-recovery T1 measurements (TR=8000 ms, TI=500, 1000, 3000 ms) were performed before and after DCE to allow calculation of Gd concentration-vs-time in tissues. Tracer kinetic modeling was used to obtain Ktrans maps. Average steady-state Gd concentration and Ktrans values (excluding zero values) in all periprosthetic tissues were calculated for each patient. T-test was used for statistical analysis. RESULTS A pseudotumor was detected in one MM patient. Excluding this case, which had the highest Ktrans average (0.072 units of inverse minutes) and highest Gd concentration average (0.22 mM), Ktrans values were significantly lower in MM patients vs. MPE patients (0.033 +/- 0.005 for MM vs. 0.052 +/- 0.006 for MPE patients (units of inverse minutes), p=0.008). On the other hand, steady-state Gd concentration was not significantly different between MM and MPE patients (0.15 +/- 0.01 mM for MM vs. 0.17 +/- 0.03 mM for MPE patients). CONCLUSION This preliminary study shows that DCE MRI can be successfully used to measure Ktrans near hip implants and detect differences between patients, depending on implant type and tissue reaction. CLINICAL RELEVANCE/APPLICATION Ktrans measurement through DCE MRI shows potential for early detection of THR failure.
    Radiological Society of North America 2012 Scientific Assembly and Annual Meeting; 11/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: PURPOSE Edema in Hoffa’s and quadriceps fat pads have been previously described as sources of pain. The prefemoral fat pad, the third intracapsular fat pad of the knee, has not been systematically evaluated. The purpose of the study is to determine the prevalence, patterns and clinical correlation of prefemoral fat pad edema. METHOD AND MATERIALS Initially, 11 clinical cases were accumulated and used as a model for a population study. Next, 478 consecutive knee MRI exams were retrospectively reviewed by 2 observers for the presence of prefemoral fat pad edema. The presence, location, pattern and extent of edema, as well as concurrent MRI findings were recorded. Subsequently, patients’ medical records were reviewed for site of pain, history of trauma/surgery, and physical exam findings. The results of the positive cases were compared to the imaging and clinical findings of an age and sex matched control group. RESULTS 33/478 (6.9%) retrospectively reviewed cases had prefemoral fat pad edema. Of all 44 positive cases, the average lesion volume was 1.10 cm3. Two distinct edema patterns were noted: In group A, 8 cases, there was a superior patellar osteophyte impinging the fat pad, with superior-central pad edema. In group B, 36 cases, edema involved the inferolateral aspect. 64% of group B patients had a history of trauma, compared to only 38% in group A. Conversly, 38% of group A patients had imaging findings of patellar tendon-lateral femoral condyle friction syndrome, versus 19% in group B. Taken as a whole, there was a significantly increased number of ACL tears (p=0.002), patellofemoral compartment osteoarthritis (p=0.043), and patellar tendon-lateral femoral condyle friction syndrome (p=0.039) in the 44 patient study group when compared to the control group. CONCLUSION There are 2 patterns of focal prefemoral fat pad edema: (A) Superior central edema, directly related to a prominent superior patellar osteophyte impinging on the pad; (B) inferolateral edema, with suggested mechanism being a prior sheer injury or impingement. ACL tears, patellar tendon-lateral femoral condyle friction syndrome and patellofemoral osteoarthritis are more frequently observed in patients with prefemoral fat pad edema. CLINICAL RELEVANCE/APPLICATION Inferolateral prefemoral pad edema has a high association with lateral patellofemoral friction syndrome and can result from trauma. Superior prefemoral fat edema occurs when impinged by an osteophyte.
    MRI Findings in 44 Patients Along with Clinical Correlation and Assessment of Prevalence. Radiological Society of North America 2012 Scientific Assembly and Annual Meeting; 11/2012

Publication Stats

5k Citations
849.20 Total Impact Points


  • 2014
    • Stony Brook University Hospital
      Stony Brook, New York, United States
  • 2009–2013
    • University of Ottawa
      • Department of Radiology
      Ottawa, Ontario, Canada
    • Medical University of Vienna
      • Universitätsklinik für Radiodiagnostik
      Vienna, Vienna, Austria
  • 2012
    • Johns Hopkins University
      Baltimore, Maryland, United States
    • University of San Francisco
      San Francisco, California, United States
  • 2008–2012
    • The Ottawa Hospital
      • • Department of Medical Imaging
      • • Department of Radiology
      Ottawa, Ontario, Canada
  • 2005–2012
    • NYU Langone Medical Center
      • • Department of Radiology
      • • Division of Rheumatology
      New York City, NY, United States
    • Boston Medical Center
      Boston, Massachusetts, United States
  • 2004–2012
    • Sheba Medical Center
      Gan, Tel Aviv, Israel
  • 2011
    • Brooks Rehabilitation Hospital
      Jacksonville, Florida, United States
    • Sun Yat-Sen University
      • Department of Oral Radiology
      Guangzhou, Guangdong Sheng, China
    • Cooper University Hospital
      Worcester, Massachusetts, United States
  • 2008–2010
    • Sunnybrook Health Sciences Centre
      • Department of Medical Imaging
      Toronto, Ontario, Canada
  • 2006–2009
    • CUNY Graduate Center
      New York City, New York, United States
    • The University of Calgary
      • Department of Radiology
      Calgary, Alberta, Canada
  • 2007–2008
    • Simon Fraser University
      • School of Computing Science
      Burnaby, British Columbia, Canada
    • Southern California Orthopedic Institute
      Los Angeles, California, United States
  • 2005–2008
    • New York University
      • • Department of Radiology
      • • Department of Chemistry
      New York City, NY, United States
  • 2002–2008
    • Rothman Institute
      Philadelphia, Pennsylvania, United States
    • Universitätsspital Basel
      Bâle, Basel-City, Switzerland
  • 1992–2008
    • Thomas Jefferson University
      • • Department of Radiology
      • • Department of Orthopaedic Surgery
      Philadelphia, PA, United States
  • 2005–2006
    • State University of New York Downstate Medical Center
      • Department of Radiology
      Brooklyn, NY, United States
  • 2004–2006
    • National Yang Ming University
      T’ai-pei, Taipei, Taiwan
  • 1998–2006
    • Boston Children's Hospital
      • Department of Radiology
      Boston, MA, United States
  • 1992–2005
    • Thomas Jefferson University Hospitals
      • Department of Radiology
      Philadelphia, Pennsylvania, United States
  • 2001–2004
    • University of Zurich
      Zürich, Zurich, Switzerland
  • 2003
    • Johns Hopkins Medicine
      • Department of Radiology and Radiological Science
      Baltimore, MD, United States
  • 2002–2003
    • Yale-New Haven Hospital
      New Haven, Connecticut, United States
  • 2000
    • Humboldt-Universität zu Berlin
      Berlín, Berlin, Germany
  • 1992–1999
    • University of California, San Diego
      • Department of Radiology
      San Diego, CA, United States
  • 1997
    • Leiden University Medical Centre
      • Department of Radiology
      Leiden, South Holland, Netherlands
    • Alexandria University
      Al Iskandarīyah, Alexandria, Egypt
  • 1996
    • University of San Diego
      San Diego, California, United States
  • 1995
    • The Ohio State University
      • Department of Radiology
      Columbus, OH, United States