Robustness of T1 and ECV mapping radiomics features: a between-session evaluation in young athletes

Funding Acknowledgements

Type of funding sources: None.

Introduction

Radiomics of cardiac MRI T1, T2 and extracellular volume (ECV) maps has the potential to add biomarkers that can aid in the detection and diagnosis of myocardial diseases. Recently, the feasibility of CMR mapping based radiomics to classify various myocardial diseases was demonstrated [1-6]. However, reproducibility studies have reported sensitivity of radiomics to acquisition parameters and processing steps involved concluding that only a limited number of features may be reproducible [7-8]. As CMR mapping guidelines recommend to use site-specific normal values [9], radiomics features derived likely also need careful site-specific evaluation to benchmark disease-related feature alterations.

Purpose

We aimed to assess the between-session reproducibility of radiomics features in a longitudinal dataset of MOLLI T1 and ECV maps obtained in young athletes at 1.5T.

Materials and methods

This study included data from 17 healthy subjects (15-20y; informed consent obtained) with data acquired two years apart [10] considered for this purpose as test-retest data since a prior standard analysis showed near identical average T1 (t1: 977±16 ms, t2: 982±20ms) and ECV (t1: 23.4±1.3%, t2: 23.4±1.5%). T1 mapping data was acquired on a 1.5T system (Ingenia, Philips) using MOLLI 5s(3s)3s. After motion correction and T1 and ECV map calculation [11], the left ventricular myocardium was manually delineated by two readers independently (3D Slicer [12]). In total 44 images (short and long axis) were included for each time point. The radiomics analysis resulted in 96 features per image (7 feature families, ‘shape’ excluded; no filters applied; Pyradiomics, [13]). The concordance correlation coefficient (CCC) was calculated to assess reproducibility, and features with CCCs ≥ 0.7 were considered reproducible. A coefficient of variation (CV) below 15% was considered low.

Results

Only a limited number of radiomics features had high CCC (T1: 6/96 ECV 0/96) or a low CV (T1: 32/96, ECV:30/96) in the between-session analysis. The inter-reader evaluation showed that the effect of the delineation on the results was limited. Features that were most robust in the between-session analysis were ‘first order (total)energy’ for T1 maps and ‘glcm_Autocorrelation’ for ECV maps (table 1). These results in young healthy subjects confirm previous test-retest reports [9-10]. Features with low CCC levels or high CV may however still be useful when discriminating between patient with myocardial diseases if the difference is larger than the confidence interval assessed via this reproducibility analysis.

Conclusion

In these healthy subjects, a strong variability in reproducibility of radiomics features of T1 and ECV mapping can be noted. Nonetheless, these variability measures are informative to determine features that are likely most robust when discriminating between health and disease and can be used as a benchmark towards radiomics AI-based diagnostic approaches. Top ranked features for either T1 or ECV

FRI0175 SEMI-DYNAMIC MRI OF THE EXTENSOR DIGITORUM TENDONS IN JACCOUD ARTHROPATHY

Background:

Jaccoud arthropathy (JA) is a rare joint deformity mostly associated with systemic lupus erythematosus (SLE) (1). JA may be defined as a reversible deformity of the hands during flexion with ulnar deviation and swan neck deformities of the fingers. The physiopathology and the mechanisms leading to JA still remain to be elucidated.

Objectives:

To test the hypothesis that Jaccoud arthropathy (JA) in systemic lupus erythematosus (SLE) patients is associated with abnormal extensor digitorum (ED) tendons displacement during flexion of the metacarpophalangeal (MCP) joints.

Methods:

16 SLE patients with JA (JA+), 12 SLE patients without JA (JA-) and 24 control subjects were included in the study. Transverse spin echo T1-weighted MR sequences of the MCP joints in flexion and in extension were obtained in each hand of SLE patients and in one randomly-selected hand of control subjects. Two radiologists separately measured the amplitude and the direction of the displacement of the ED tendons with respect to the midline at the level of the MCP joints. Statistical analysis included two-way ANOVA with random effects to assess differences in amplitude (p<0.0083), Fisher-Freeman-Halton exact test to assess differences in direction (p<0.0063) and Gwet’s AC1 score to determine interobserver agreement.

Results:

Amplitude of the displacement of the ED tendons was statistically significantly higher in JA+ patients than in JA- patients and controls in flexion for both readers (p<0.0001) and in extension for one reader (p<0.0048). Ulnar deviation of the ED tendons was statistically significantly more frequent in JA+ patients than in JA- patients and controls in flexion and in extension for both readers (p<0.0001). Interobserver agreement for the assessment of displacement was moderate to very good (0.51≤kappa≤0.93).

Conclusion:

JA is associated with abnormal displacement of the ED tendons in flexion and extension. Abnormal displacement of the ED tendons is absent in patients without JA.

References:

[1]Santiago MB. Miscellaneous non-inflammatory musculoskeletal conditions. Jaccoud’s arthropathy. Best Pract Res Clin Rheumatol. 2011;25(5):715-25.

Figure 1.

Axial spin echo T1-weighted images of a 49-year-old woman with SLE and JA (a) in flexion and (b) extension. In flexion, the extensor digitorum (ED) tendons of the 3rdand 4thfingers are completely displaced on the ulnar aspect of the metacarpal heads (white arrows) with no tendon visible on the dorsal aspect (arrowheads). In extension, the ED tendons of the 3rdand 4thfingers are slightly displaced on the ulnar side of the dorsal aspect of the metacarpal heads (black arrows).

Figure 2.

Axial spin echo T1-weighted images of the sex- and age-matched control subject of the patient from Figure 1 (a) in flexion and (b) extension. The extensor digitorum tendons of the 3rdand 4thfingers are normally positioned on the dorsal aspect of the metacarpal heads in flexion (white arrows) and extension (black arrows).

Disclosure of Interests:

Thomas Kirchgesner: None declared, Maria Stoenoiu: None declared, Nicolas Michoux: None declared, Xavier Libouton: None declared, Frederic Houssiau Grant/research support from: UCB, Consultant of: GSK, Bruno Vande Berg: None declared

THU0533 A SINGLE MRI DIXON SEQUENCE TO ASSESS DISEASE ACTIVITY AND CARTILAGE IN EARLY RHEUMATOID HANDS: ONE SEQUENCE TO ASSESS THEM ALL?

Background:

OMERACT recommends three “core set” MRI sequences with an optional cartilage-dedicated sequence to perform Rheumatoid Arthritis (RA) MRI scoring (RAMRIS) (1). Dixon method allows the production of four different images from a single MRI sequence.

Objectives:

To test a short MRI protocol based on a single Dixon sequence to assess disease activity and cartilage in hands of patients with early RA.

Methods:

Twenty-four patients (16 women, mean age 45.7 years old) with early DMARD-naive RA meeting the ACR/EULAR 2010 criteria were prospectively included. Both hands of each patient were imaged with MRI including contrast-enhanced T1-weighted Dixon and OMERACT “core set” MRI sequences and with conventional radiography.

Two musculoskeletal radiologists (R1 and R2) separately assessed disease activity according to RAMRIS scoring system based on the Dixon images (contrast-enhanced T1-weighted Dixon water-only images to score synovitis, tenosynovitis and osteitis and contrast-enhanced T1-weighted Dixon fat-only images to score erosions) and the three OMERACT “core set” MRI sequences (contrast-enhanced fat-saturated T1-weighted images to score synovitis and tenosynovitis, fat-saturated T2-weighted images to score osteitis and T1-weighted images without contrast-material injection to score erosions).

One radiologist (R1) separately measured the thickness of the cartilage in the joints corresponding to those assessed by the Sharp/van der Heijde modified scoring method on contrast-enhanced T1-weighted Dixon out-of-phase images and radiographs (2).

RAMRIS scoring and measurement of the cartilage thickness were repeated by R1 to assess intra-observer agreement. Statistical analysis was based on intra-class correlation coefficients (ICC) with 95% confidence interval to assess inter-technique, intra-observer and inter-observer agreement. The strength of agreement was interpreted as follows: ≤0, poor; 0.01-0.20, slight; 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, substantial and ≥0.81, excellent.

Results:

Agreement between total RAMRIS scores obtained with the Dixon water- and fat-only images and total RAMRIS scores obtained with the OMERACT sequences was excellent for R1 (0.94; 0.86-0.97) and R2 (0.91; 0.81-0.96). Intra-observer agreement was excellent with Dixon images (0.97; 0.92-0.98) and OMERACT sequences (0.96; 0.90-0.98). Inter-observer agreement was excellent with Dixon images (0.92; 0.82-0.96) and OMERACT sequences (0.93; 0.85-0.97).

Agreement between the measures of cartilage thickness on the Dixon out-of-phase images and the measures of cartilage thickness on radiographs was substantial (0.71; 0.66-0.75). Intra-observer agreement was excellent with Dixon out-of-phase images (0.94; 0.93-0.95) and radiographs (0.93; 0.92-0.94).

Conclusion:

An MRI protocol based on a single contrast-enhanced T1-weighted Dixon sequence allows reproducible RAMRIS scoring and measurement of the cartilage thickness. Further studies should be performed to evaluate the value of a short MRI protocol based on the Dixon method to monitor disease activity including cartilage loss in treated RA patients.

References:

[1]Ostergaard M, Peterfy CG, Bird P, Gandjbakhch F, Glinatsi D, Eshed I, et al. The OMERACT Rheumatoid Arthritis Magnetic Resonance Imaging (MRI) Scoring System: Updated Recommendations by the OMERACT MRI in Arthritis Working Group. J Rheumatol. 2017;44(11):1706-12.

[2]van der Heijde D. How to read radiographs according to the Sharp/van der Heijde method. J Rheumatol. 1999;26(3):743-5.

Disclosure of Interests:

Thomas Kirchgesner: None declared, Maria Stoenoiu: None declared, Nicolas Michoux: None declared, Patrick Durez Speakers bureau: AbbVie, Bristol-Myers Squibb, Celltrion, Eli Lilly, Pfizer, Sanofi, Bruno Vande Berg: None declared

Fat suppression at three-dimensional T1-weighted MR imaging of the hands: Dixon method versus CHESS technique

PURPOSE

To compare the effectiveness of fat suppression and the image quality of the Dixon method with those of the chemical shift-selective (CHESS) technique in hands of normal subjects at non-enhanced three-dimensional (3D) T1-weighted MR imaging.

MATERIALS AND METHODS

Both hands of 14 healthy volunteers were imaged with 3D fast spoiled gradient echo (FSPGR) T1-weighted Dixon, 3D FSPGR T1-weighted CHESS and 3D T1-weighted fast spin echo (FSE) CHESS sequences in a 1.5T MR scanner. Three radiologists scored the effectiveness of fat suppression in bone marrow (EFS^BM) and soft tissues (EFS^ST) in 20 joints per subject. One radiologist measured the signal-to-noise ratio (SNR) in 10 bones per subject. Statistical analysis used two-way ANOVA with random effects (P<0.0083), paired t-test (P<0.05) and observed agreement to assess differences in effectiveness of fat suppression, differences in SNR and interobserver agreement.

RESULTS

EFS^BM was statistically significantly higher for the 3D FSPGR T1-weighted Dixon than for the 3D FSPGR T1-weighted CHESS sequence and the 3D FSE T1-weighted CHESS sequence (P<0.0001). EFS^ST was statistically significantly higher for the 3D FSPGR T1-weighted Dixon than for the 3D FSPGR T1-weighted CHESS sequence (P<0.0011) and for the 3D FSE T1-weighted CHESS sequence in the axial plane (P=0.0028). Mean SNR was statistically significantly higher for 3D FSPGR T1-weighted Dixon sequence than for 3D FSPGR T1-weighted CHESS and 3D FSE T1-weighted CHESS sequences (P<0.0001).

CONCLUSION

The Dixon method yields more effective fat suppression and higher SNR than the CHESS technique at 3D T1-weighted MR imaging of the hands.

Application of a three-dimensional graft of autologous osteodifferentiated adipose stem cells in patients undergoing minimally invasive transforaminal lumbar interbody fusion: clinical proof of concept

BACKGROUND

The authors applied a scaffold-free osteogenic three-dimensional (3D) graft made of adipose-derived mesenchymal stem cells (AMSCs) in patients undergoing minimally invasive transforaminal lumbar interbody fusion (MI-TLIF).

METHODS

Three patients (two patients and one patient with 1 and 2 levels, respectively) with degenerative spondylolisthesis underwent MI-TLIF with 3D graft made of AMSCs. To obtain the AMSCs, fatty tissue was collected from the abdomen by lipoaspiration and differentiated afterwards in our Cell/Tissue bank. Clinical outcomes, including the Oswestry Disability Index (ODI) and visual analog scale (VAS) as well as fusion status were assessed preoperatively and up to 12 months postoperatively.

RESULTS

At 12 months, all four operated AMSC levels could be assessed (n = 4). Grade 3 fusion could be confirmed at two levels out of four. Mean VAS score improved from 8.3 to 2 and ODI also improved from 47 to 31%. No donor site complication was observed. The final AMSC osteogenic product was stable, did not rupture with forceps manipulation, and was easily implanted directly into the cage with no marked modification of operating time.

CONCLUSIONS

A scaffold-free 3D graft made of AMSCs can be manufactured and used as a promising alternative for spinal fusion procedures. Nevertheless, further studies of a larger series of patients are needed to confirm its effectiveness.

Imaging ACL reconstructions and their complications

Examination of ligament reconstructions, particularly of the anterior cruciate ligament (ACL), are common situations in everyday knee imaging practice. Knowledge of normal appearances, the expected changes over time and the potential complications of these plasties are essential. MRI is the imaging method of choice. This article illustrates the main complications specific to this procedure: suboptimal positioning of the femoral or tibial tunnels, impingement between the graft and bony contours, rupture (partial or complete) of the plasty due to friction or injury, arthrofibrosis and particularly the "Cyclops" syndrome, fragmentation or migration of the fixation materials and a granulomatous reaction to biomaterials.

[Transient bone marrow edema of the hip]

Transient bone marrow edema of the hip is characterized by moderate homogeneous low MR signal intensity with ill-defined margins that involves at least a portion of the femoral head. Spin echo T1-weighted images are helpful to exclude other underlying diseases (tumor, infection, necrosis from systemic origin…), for which marrow edema is secondary or no epiphyseal in location. High-resolution fat-suppressed T2-weighted or proton density images allow evaluation of the articular cartilage, subchondral bone and subchondal marrow: if the articular cartilage is abnormal, the lesion is irreversible (arthrosis or necrosis); if the subchondral bone is focally interrupted and/or if the femoral is no longer spherical, the lesion is irreversible (necrosis); if a focal linear fluid collection is present under the subchondral bone, the lesion is irreversible (necrosis). Finally, subchondral changes may provide useful prognostic information: the absence of any abnormality other than marrow edema typically indicates that complete resolution is likely; the presence of a focal T2-weighted hypointense lesion immediately next to the subchondral bone suggests an irreversible lesion, especially if it is equal to or thicker than 4mm or the joint space. In some instances, prognosis cannot be reliably determined requiring the need for follow-up imaging.

[Principles of analysis for sacroiliac joints imaging]

Interpretation of sacroiliac joints imaging is uneasy. Simultaneous erosions, hyperostosis, and ankylosis together make the diagnosis of inflammatory lesions. Articular abnormalities may also be induced by mechanical stress or ligament ossifications. Distribution of the lesions can help to establish a precise diagnosis. Inflammatory lesions may be located in any part of the articulation, including the posterior and inferior part. Mechanical lesions as in osteitis condensans ilii are commonly located in the anterior middle part of the joint. Ligament ossification in case of idiopathic skeletal hyperostosis is located at the margins of joint. The basic for the interpretation of sacroiliac joints is to look at high quality plain radiographs. When diagnosis is uncertain, complementary methods must be considered. CT is useful for analysis of subacute or chronic lesions and MRI is to be preferred for acute lesions, in young patients and when searching for signs of inflammatory activity in an already known chronic disease.

Bone marrow edema of the femoral head

This article addresses the MR features of the bone marrow edema syndrome (BMES) of the femoral head with emphasis on the prevalence and clinicopathology of the disorder and description of the current concepts on diagnosis and prognosis. BMES can be observed in self resolving conditions such as transient osteoporosis of the hip, spontaneous fracture of the femoral head, or post traumatic lesions. Rapidly destructive coxarthrosis, necrosis of the femoral head as well as certain forms of spontaneous fracture of the femoral head may present a similar MR pattern, though prognosis is definitely less favourable. The challenging role of the radiologist is to recognize BMES at an early stage and to provide adequate prognosis on the lesion outcome.

[Multiple diabetic muscular infarctions after liver and kidney transplantation: a case report]

INTRODUCTION

Several neuromuscular diseases may complicate diabetes mellitus and transplantation, including chronic sensorimotor length dependent polyneuropathy.

OBJECTIVE

Description of muscular infarction, a rare complication of diabetes mellitus, which occurred after liver and kidney transplantation.

CASE REPORT

A 57-year-old patient presented with long-term diabetes mellitus and multiple complications. End-stage renal and hepatic disease led to kidney and liver transplantation. Twenty-seven days after transplantation, swelling and induration appeared in the left shoulder and forearm. Forty-three days after transplantation, the same symptoms appeared in both lower limbs. Markedly reduced range of motion led to severe disability. Bone scans showed multiple spots following muscle anatomy. Computed tomography gave negative results. Magnetic resonance imaging (MRI) confirmed muscular infarction by a high T1 signal (muscular necrosis) and soft-tissue infiltration.

DISCUSSION AND CONCLUSION

Muscular infarction is a rare and unknown complication of diabetes mellitus. It is characterised by sudden painful muscular induration and swelling affecting one muscle at a time with recurrence. Our patient presented with simultaneous multiple muscular infarctions in 3 limbs. Diagnosis was based on clinical investigation and MRI. The treatment is conservative and the condition generally resolved by itself. However, the long-term prognosis of muscular infarction is not good because of the cardiovascular-associated complications of diabetes mellitus.

Papillomatose juvénile mammaire

The authors report a patient with juvenile papillomatosis of the breast presenting with a palpable mass and illustrate the correlation between mammographic, sonographic and pathologic features.

Post-renal transplant syndrome of transient lower limb joint pain: description under a tacrolimus-based immunosuppression

The occurrence of a post-renal transplant syndrome of lower limbs joint pain has been reported extensively over the last decade. Clinical examination of the symptomatic joints is often unremarkable and magnetic resonance imaging reveals abnormalities of the bone marrow suggestive of edema and/or hemorrhage. The main striking features of this syndrome are the spontaneous resolution of the symptoms within a few weeks as well as of the marrow abnormalities. This syndrome has been attributed to cyclosporine, given in the immunosuppression regimen or to epiphyseal impactions. We here document the occurrence of this syndrome in 5 kidney graft recipients given a tacrolimus-based immunosuppression.

Detection of flow limitation in mechanically ventilated patients

OBJECTIVE

In mechanically ventilated patients flow limitation often goes unrecognised. We compared three methods for detection of flow limitation in mechanically ventilated patients: the resistance method, the negative expiratory pressure (NEP) method, and the interrupter method.

DESIGN

Prospective study.

SETTING

A medical intensive care unit in a university hospital.

PATIENTS

Twenty-six patients (20 COPD, six other pathology), mechanically ventilated under sedation.

MEASUREMENTS AND RESULTS

Respiratory mechanics were obtained during application of the three methods. For the resistance method, flow limitation was expressed as percentage of expiratory tidal volume, in which flow did not decrease (FLP-R). For the NEP method, flow limitation was expressed as percentage of expiratory tidal volume in which flow did not increase (FLP-NEP). For the interrupter method, flow limitation was expressed as area of spike-overshoot in flow after interruption. In 18 COPD patients, flow limitation was detected by all methods; mean FLP-R 76% (SD 12%), mean FLP-NEP 90% (SD 11%), mean spike area 21 ml (SD 7 ml). In three patients with other pathology, these values were, respectively, 20% (SD 19%), 48% (SD 21%), and 5 ml (SD 4 ml). The three methods were in close agreement. In nine patients the resistance method increased flow and in six patients the NEP method decreased flow compared to the unimpeded breath.

CONCLUSIONS

In mechanically ventilated patients, flow limitation can well be detected by the resistance-, NEP-, and interrupter methods. However, the NEP method can overestimate the flow limited portion, while the resistance method can underestimate the flow limited portion. The interrupter method is found to be less practical.

Recovery of red blood cell mass following orthognathic surgery

The rates of recovery of red blood cell mass after surgery were compared in three groups of orthognathic surgery patients: 12 patients not receiving transfusions during Le Fort I osteotomy, 10 patients receiving transfusions during Le Fort I osteotomy, and 25 patients undergoing bilateral sagittal split osteotomy. Patients in these groups lost a mean of 421, 455, and 191 mL of blood, respectively, during surgery. Both groups of patients undergoing Le Fort I osteotomy donated a unit of blood prior to surgery, while the patients undergoing bilateral sagittal split osteotomy did not. A control group consisted of 77 presurgical orthognathic patients who donated a unit of blood (450 mL) a mean of 11 days prior to surgery. Among the control group, 38 had recovered red blood cell mass prior to surgery. The patients undergoing Le Fort I osteotomy recovered red blood cell mass after surgery more quickly than did the patients undergoing bilateral sagittal split osteotomy, despite having lost twice as much blood. This unexpected finding can be attributed to the need to lose sufficient red blood cell mass to stimulate the erythropoietic system. The patients undergoing Le Fort I osteotomy experienced two challenges to the system, predonation of blood and greater blood loss during surgery, resulting in a great enough decrease in red blood cell mass to stimulate the release of erythropoietin. Patients undergoing bilateral sagittal split osteotomy did not exceed this threshold. None of the differences among the groups were statistically significant, possibly because of the few number of subjects.

Costal cartilage fractures as revealed on CT and sonography

OBJECTIVE

We describe the CT and sonographic appearance of 15 costal cartilage fractures observed in eight patients.

CONCLUSION

On CT, fracture was seen as a low-density area through the costal cartilage, with surrounding calcifications present near old fractures, and gas density within the cleft in some cases. On sonography, cartilage fracture appeared as an interruption of the smooth anterior aspect of the cartilage.

[Radiologic aspects of the loosening of cemented hip prostheses: mechanical, septic or granulomatous etiology?]

Radiologic diagnosis of hip prosthesis loosening is based on the evaluation of each component (prosthesis, cement, bone) and of their interfaces. Both the prosthesis and the cement may deteriorate and the prosthesis/cement interface or cement/bone interface may become abnormal in prosthesis loosening of any etiology. In contrast, the aspect of the bone changes (erosion, periostitis) and their distribution vary according to the condition etiology. It appears from a retrospective study of 50 cases of chronic hip prosthesis loosening that the most specific signs for infection are unsharp bone resorption and acute-like or multifocal periostitis. In granulomatous loosening, bone resorption is sharp (as in mechanical loosening), but its distribution is not conform to the prosthesis shape (as in septic loosening), and periosteal changes are not observed.

Injury to the brachial plexus by a fragment of bone after fracture of the clavicle

Clavicular fractures are occasionally responsible for lesions of the brachial plexus. The symptoms are usually delayed and due to compression by hypertrophic callus, nonunion or a subclavian pseudoaneurysm. We describe a patient in whom a displaced bone fragment was pressing on the retroclavicular part of the brachial plexus, leading to early symptoms of a lesion of the posterior cord. Internal fixation of the clavicle and external neurolysis of the brachial plexus gave an almost full recovery.

INJURY TO THE BRACHIAL PLEXUS BY A FRAGMENT OF BONE AFTER FRACTURE OF THE CLAVICLE

Clavicular fractures are occasionally responsible for lesions of the brachial plexus. The symptoms are usually delayed and due to compression by hypertrophic callus, nonunion or a subclavian pseudoaneurysm.

We describe a patient in whom a displaced bone fragment was pressing on the retroclavicular part of the brachial plexus, leading to early symptoms of a lesion of the posterior cord. Internal fixation of the clavicle and external neurolysis of the brachial plexus gave an almost full recovery.

Hematological changes in anorexia nervosa are correlated with total body fat mass depletion

OBJECTIVE

To assess the relationships between total body fat mass (FM) and hematological abnormalities in anorexia nervosa (AN).

METHOD

Peripheral blood parameters and body composition were determined in 10 anorectic patients and 19 age- and sex-matched healthy subjects. In patients with AN, magnetic resonance imaging (MRI) studies of bone marrow were also performed.

RESULTS

Compared with controls, patients with AN had -41% body weight, -81% FM, -18.8% lean tissue mass (LTM), and -22.6% bone mineral content; they also had lower mean total leukocyte (4.52 +/- 0.47 vs. 6.28 +/- 0.33 x 10(3)/microliter, p < .005), neutrophil (2.45 +/- 0.34 vs. 3.46 +/- 0.20 x 10(3)/microliter, p < .005), monocyte (0.24 +/- 0.03 vs. 0.37 +/- 0.03 x 10(3)/microliter, p < .05), and platelet counts (184 +/- 17 vs. 238 +/- 9 x 10(3)/microliter, p < .005). Hemoglobin level was normal and comparable in both groups. In patients with AN, but not in controls, total leukocyte, neutrophil, eosinophil, and monocyte counts as well as hemoglobin level were highly correlated with FM expressed in absolute values or in percentage of body weight, but not with LTM. Moreover, AN patients with signal intensity patterns suggestive of serous atrophy of bone marrow at MRI had not only lower erythrocyte, leukocyte, neutrophil, and platelet counts, but they also had lower FM than AN patients with normal MRI patterns.

DISCUSSION

Hematological changes in AN, as assessed by peripheral blood parameters and MRI patterns of bone marrow, are correlated with total body FM depletion, suggesting that the reduction of adipose tissue adversely affects hematopoiesis.

Long-term effects of localized spinal radiation therapy on vertebral fractures and focal lesions appearance in patients with multiple myeloma

The occurrence of new vertebral fractures and focal marrow lesions was determined and compared in irradiated and nonirradiated vertebrae of 12 patients with multiple myeloma (MM), prospectively followed using magnetic resonance imaging (MRI) of the thoraco-lumbar spine after localized spinal radiation therapy. During follow-up (mean 35 months), fractures appeared in 5% of irradiated vertebrae and in 20% of nonirradiated vertebrae: new focal lesions appeared in 4% of irradiated vertebrae and in 27% of nonirradiated vertebrae. This study demonstrates a beneficial long-term effect of localized radiation therapy, consisting of a reduced incidence of vertebral fractures and focal marrow lesions in irradiated vertebrae.

Osteopenia, bone fragility and reflex sympathetic dystrophy syndrome in a man with ureterosigmoidostomy

A 68-year-old man is presented with a reflex sympathetic dystrophy syndrome (RSDS) of the right ankle diagnosed by radiography, magnetic resonance imaging and bone scintiscan. Investigations, including blood tests and bone biopsy, revealed a diagnosis of metabolic acidosis and osteomalacia. These appeared to result from a ureterosigmoidostomy performed 9 years previously for a transitional carcinoma of the bladder. Correction of the metabolic acidosis coincided with improvement in ankle pain. RSDS may be the initial presentation of osteomalacia, which in turn may be caused by the metabolic acidosis resulting from a ureterosigmoidostomy.

Apparent focal bone marrow ischemia in patients with marrow disorders: MR studies

OBJECTIVE

We sought to assess with MR the focal bone marrow abnormalities in patients with chronic marrow disorders and acute limb pain.

MATERIALS AND METHODS

We investigated and followed with MRI four patients with proliferative or dysplastic marrow disorders presenting with acute but spontaneously resolving hip or thigh pain.

RESULTS

Ten focal marrow lesions were demonstrated on T2-weighted images as high signal intensity (SI) areas. They remained undetected on T1-weighted images as they showed a low SI similar to the disease-related low SI of the entire marrow. Postcontrast images demonstrated lack of enhancement in the lesions. Follow-up enhanced MR images showed intense enhancement within the lesions, while unenhanced MR images remained unchanged. Later on, these focal marrow abnormalities completely resolved. These lesions most likely represent bone marrow ischemia, although histological proof is lacking.

CONCLUSION

Acute bone pain in patients with bone marrow disorders may be related to focal marrow lesions suggestive of bone marrow ischemia.

Intravertebral vacuum cleft: changes in content after supine positioning

Sequential radiographic and magnetic resonance (MR) imaging examinations were performed in nine patients with an intravertebral vacuum cleft indicative of avascular necrosis. Progressive changes in the content of the cleft occurred within an hour after the patients were placed in a supine position. Initially, the cleft showed a gaslike pattern during extension of the spine, with a radiolucent band on radiographs and a signal void on MR images. Later, the vacuum phenomenon disappeared on radiographs, and a fluidlike high-signal-intensity pattern appeared on T2- or T2*-weighted MR images, suggestive of a slow fluid inflow within the intravertebral cleft. Because the recognition of a vacuum cleft in a collapsed vertebral body helps avoid confusion with malignancy or infection, it is important to search for this in examinations performed immediately after supine positioning.

Avascular necrosis of the hip: comparison of contrast-enhanced and nonenhanced MR imaging with histologic correlation. Work in progress

In 15 hips with typical signs of avascular necrosis of the femoral head on plain radiographs and magnetic resonance (MR) images, gadolinium-enhanced spin-echo and fat-suppressed MR images were obtained and compared with nonenhanced T1- and T2-weighted images. Both enhanced and nonenhanced areas were consistently detected in the abnormal femoral heads. Enhanced areas showed a low signal intensity (SI) on T1-weighted MR images obtained before contrast material was administered and an intermediate to high SI on T2-weighted images. Nonenhanced areas showed an SI either identical (pattern 1) or hypointense (pattern 2) to that of fat on both sequences. Histologic correlation (six resected femoral heads) helped confirm that enhanced and nonenhanced areas corresponded respectively to viable and necrotic tissue. In most cases, SI analysis of nonenhanced T1- and T2-weighted images allows the differentiation of hypervascularized viable tissue from hypovascularized necrotic tissue of the sequestrum.

Benign osteochondromas and exostotic chondrosarcomas: evaluation of cartilage cap thickness by ultrasound

Ultrasonography (US) enables accurate assessment of the cartilage cap of exostoses. The cartilage cap appears as a hypoechoic layer covering the hyperechoic surface of the calcified part. Measurements of cap thickness with US were compared with measurements performed on pathological specimens in 22 resected exostoses and 2 exostotic chondrosarcomas. The US measurements proved to be very accurate, with a mean measurement error of less than 2 mm for cartilage caps less than 2 cm thick. The detection rate and measurement accuracy of US were higher than with computed tomography (CT) and comparable to magnetic resonance imaging (MRI), which were available in 14 and 10 cases, respectively. US appears to be a good procedure for evaluating the cartilage cap, which is usually thin for a benign exostosis and thick for a malignancy. In addition, other complications--such as bursa formation--are easily recognizable. The sole limitation is that US cannot visualize the cartilage cap when it is inwardly orientated or deeply located in soft tissues, which are both, however, relatively uncommon situations.