What’s New in MSK Imaging? – March 2022

2 years ago

Cotyloid Fossa Coverage Percentages May Be Associated With Alpha Angle, Labral Tear, and Clinical Outcomes in Patients With Femoroacetabular Impingement

Yoichi Murata, Lauren Pierpoint,  Madeleine DeClercq,  Carly Lockard, Maitland Martin, Naomasa Fukase, Rui Soares, Patrick Quinn, Charles P Ho, Soshi Uchida, Marc J Philippon.


American Journal of Sports Medicine




Background: Within the hip joint, the anatomy of the acetabulum and cotyloid fossa is well established. There is little literature describing the association between the size of the cotyloid fossa relative to the acetabulum and characteristics of patients with femoroacetabular impingement (FAI).  The purpose was to calculate the cotyloid fossa coverage percentage in the acetabulum and determine its association with patient characteristics, radiographic parameters, intra-articular findings, and preoperative patient-reported outcomes in patients with FAI.


Question: Is there an association between the cotyloid fossa coverage percentage of the acetabulum and characteristics of patients with FAI?


Design: Cross-sectional study, Level of evidence, 3.


Participants: Patients were included who underwent standard clinical 3-T magnetic resonance imaging of the hip and primary arthroscopic FAI correction surgery during 2015 and 2016. An overall 146 patients were included.


Exclusion criteria: Age <18 or >40 years, osteoarthritis, labral reconstruction, previous ipsilateral hip surgery, and hip dysplasia.


Methods:  Measurements of the cotyloid fossa and surrounding lunate cartilage were performed to calculate cotyloid fossa width (CFW) and cotyloid fossa height (CFH) coverage percentages. The relationships between coverage percentages and patient characteristics and intraoperative findings were assessed using independent t tests or Pearson correlations.


Posterior lunate cartilage width (PLCW), cotyloid fossa width (CFW), and anterior lunate cartilage width (ALCW) measurement locations: (A) 3-dimensional representation of the pelvis and (B) axial PDw TSE non–fat suppressed slice, as used for measurement in this study. Superior lunate cartilage width (SLCW) and cotyloid fossa height (CFH) measurement locations: (C) 3-dimensional representation of the pelvis and (D) coronal PDw TSE fat-suppressed slice, as used for measurement in this study.


Main Results:  Alpha angle negatively correlated with CFH coverage percentage and positively correlated with labral tear size. CFH coverage percentage was negatively correlated with labral tear size.  Cotyloid fossa coverage was not associated with the condition of the cotyloid fossa synovium (synovitis vs no synovitis). CFW coverage percentage was negatively correlated with the 12-Item Short Form Health Survey (SF-12) physical component summary score.


Conclusion: The CFW and CFH coverage percentages may be associated with alpha angle, labral tear size, and SF-12 physical component summary score in patients with FAI. We may be able to predict the labral condition based on preoperative measurements of CFH and CFW coverage percentages.


Senior editorial comment: Interesting research. Not sure of the clinical value though. It is difficult to establish cartilage thickness or labral tear size without 3D imaging and intervening fluid in hip joint will alter thickness. Questionable results and their value in my opinion. Alpha angle itself has variability in its measurements.


Take home message:

  • Think of cartilage coverage of the acetabulum. The loss of cartilage can affect treatment options.
  • Cotyloid fossa size can impact the area of cartilage that articulates with the femoral head, large cotyloid fossa can result in a smaller zone of articular cartilage.
  • Limited clinical impact of the presented study, as no clinical correlation was made.


Fully Automated Deep Learning Tool for Sarcopenia Assessment on CT: L1 Versus L3 Vertebral Level Muscle Measurements for Opportunistic Prediction of Adverse Clinical Outcomes

Perry J. Pickhardt, MD, Alberto A. Perez, MD, John W. Garrett, PhD, Peter M. Graffy, BS, MPH, Ryan Zea, MS and Ronald M. Summers, MD, PhD


American Journal of Roentgenology




Background: Sarcopenia is associated with adverse clinical outcomes. CT-based skeletal muscle measurements for sarcopenia assessment are most commonly performed at the L3 vertebral level. The purpose of this article is to compare the utility of fully automated deep learning CT-based muscle quantitation at the L1 versus L3 level for predicting future hip fractures and death.


Question: Can CT-based measurements of muscle attenuation for myosteatosis at the L1 level be used to predict hip fracture and death instead of the previously established L3-level measurements?


Design: Retrospective study


Participants: 9223 asymptomatic adults (mean age, 57 ± 8 [SD] years; 4071 men, 5152 women) who underwent unenhanced low-dose abdominal CT.


Exclusion criteria: Excluded were 63 patients with symptoms that prompted the CT colonography evaluation, 82 who had less than 1 year of clinical follow-up after the CT evaluation in the absence of an earlier defined adverse event, three with unavailable or corrupted CT DICOM data, and 28 with muscle tool failure.


Methods:  A previously validated fully automated deep learning tool was used to assess muscle for myosteatosis (by mean attenuation) and myopenia (by cross-sectional area) at the L1 and L3 levels. Performance for predicting hip fractures and death was compared between L1 and L3 measures. Performance for predicting hip fractures and death was also evaluated using the established clinical risk scores from the fracture risk assessment tool (FRAX) and Framingham risk score (FRS), respectively.


Main Results: Median clinical follow-up interval after CT was 8.8 years, yielding hip fractures and death in 219 (2.4%) and 549 (6.0%) patients, respectively. L1-level and L3-level muscle attenuation measurements were not different in 2-, 5-, or 10-year AUC for hip fracture or death. Lowest quartile hazard ratios (HRs) for hip fracture were 2.20 (L1 attenuation), 2.45 (L3 attenuation), and 2.53 (FRAX score), and for death were 3.25 (L1 attenuation), 3.58 (L3 attenuation), and 2.82 (FRS). CT-based muscle cross-sectional area measurements at L1 and L3 were less predictive for hip fracture and death.


Sarcopenia (myosteatosis) at screening CT colonoscopy in 79-year-old woman with subsequent hip fracture.

A, CT images show L1 vertebral level without (A) and with (B) overlay of automated skeletal muscle segmentation (B, red). Intramuscular fat is present within paraspinal muscles (circle, A). Mean muscle attenuation is similar for manually placed ROI (1.8 HU) and automated tool (3.9 HU) and is markedly decreased for both approaches. Muscle cross-sectional area is relatively preserved when intramuscular fat is included.



Conclusion: Automated CT-based measurements of muscle attenuation for myosteatosis at the L1 level compare favorably with previously established L3-level measurements and clinical risk scores for predicting hip fracture and death. Assessment for myopenia was less predictive of outcomes at both levels. Alternative use of the L1 rather than L3 level for CT-based muscle measurements allows sarcopenia assessment using both chest and abdominal CT scans, greatly increasing the potential yield of opportunistic CT screening.


Senior editorial comment: Interesting research and insights. Thanks! Hope one day, such results will translate into clinical practice. Whole abdomen muscle segmentation and automation might produce better results and mitigate variability on single slice segmentation among patients.


Take home message:

  • Understanding muscle bulk can provide an assessment of a patient’s general condition
  • A vast amount of “free” data is available on multiple cross-sectional imaging studies. If there is an automatic way of assessing the general condition of a patient using muscle mass measurements, it will certainly be useful for predictive analysis and provide guidance for general patient assessment.
  • The lumbar spine region is a potential area to measure Sarcopenia (myosteatosis) and this study suggests that the levels where this can be measured are at L1 and L3 levels.


CT in Patients With External Fixation for Complex Lower Extremity Fractures: Impact of Iterative Metal Artifact Reduction Techniques on Metal Artifact Burden and Subjective Quality

Andreas Stefan Brendlin, MD, Christian Philipp Reinert, MD, Heiko Baumgartner, MD, Malte Niklas Bongers, MD, Christoph Thomas, MD, Saif Afat, MD, Fabian Springer, MD, MHBA and Haidara Almansour, MD, MEng.


American Journal of Roentgenology




Background:  Lower extremity external fixators have complex geometries that induce pronounced metal artifact on CT. Iterative metal artifact reduction (iMAR) algorithms help reduce such artifact, although no dedicated iMAR preset exists for external fixators.


Question:  Which iMAR presets for CT examinations achieves the greatest metal artifact burden reduction and highest subjective image quality in patients with external fixators for complex lower extremity fractures?


Design: Retrospective study


Participants: 72 CT examinations in 56 patients (20 women, 36 men; mean age, 56 ± 18 [SD] years) with lower extremity external fixators (regular, hybrid, or monotube).


Exclusion criteria: No patients or examinations during the study were excluded.


Method: Examinations were reconstructed without iMAR (hereafter referred to as “noMAR”) and with three iMAR presets (iMARspine, iMARhip, iMARextremity). A radiology resident quantified metal artifact burden using software. Two radiology residents independently assessed overall image quality and diagnostic confidence using 4-point scales (4 = excellent [highest quality or highest confidence]). Techniques were compared using Bonferroni-corrected post hoc tests. Interreader agreement was assessed by intraclass correlation coefficients (ICCs). A post hoc multinomial regression model was used for predicting overall image quality.


Main Results:  Metal artifact burden was lower and overall image quality was higher for iMARhip and iMARextremity than noMAR and iMARspine for all fixators (aside from image quality of iMARhip and iMARextremity vs iMARspine for regular fixators) but were not different between iMARhip and iMARextremity. Median diagnostic confidence was 4 for all fixators and reconstructions. Independent predictors of overall quality relative to noMAR were iMARspine (odds ratio [OR] = 1.92–5.51), iMARhip (OR = 5.56–31.10), and iMARextremity (OR = 7.07–38.21). All iMAR presets introduced new reconstruction artifacts for all examinations for both readers.


Conclusion:  For the three fixator types, iMARhip and iMARextremity achieved greatest metal artifact burden reduction and highest subjective image quality, although both introduced new reconstruction artifacts. CT using the two identified iMAR presets may facilitate perioperative management of external fixators.



Senior editorial comment: Metal artifact reduction is a hot topic and has a lot of clinical utility. Hopefully, such improvements will improve image quality and help patient care.


Take home message:

  • Multiple newer techniques of metal suppression are available with software and hardware techniques.
  • The suppression techniques can be optimized for various body parts with varying types of software algorithms.


Prediction of Early Treatment Response in Multiple Myeloma Using MY-RADS Total Burden Score, ADC, and Fat Fraction From Whole-Body MRI: Impact of Anemia on Predictive Performance


Huazheng Dong, MS, Wenyang Huang, MS, Xiaodong Ji, MD, Lixiang Huang, MS, Dehui Zou, MS, Mu Hao, MD, Shuhui Deng, MD, Zhiwei Shen, PhD, Xiudi Lu, MD, Jian Wang, BS, Zhiyi Song, BS, Xuening Zhang, PhD, MD, Huadan Xue, PhD, MD and Shuang Xia, PhD, MD

American Journal of Roentgenology



Background: The recently released Myeloma Response Assessment and Diagnosis System (MY-RADS) for multiple myeloma (MM) evaluation using whole-body MRI (WB-MRI) describes the total burden score. However, assessment is confounded by red bone marrow hyperplasia in anemia. The purpose of this study is to assess the utility of the MY-RADS total burden score, ADC, and fat fraction (FF) from WB-MRI in predicting early treatment response in patients with newly diagnosed MM and to compare the utility of these measures between patients with and without anemia.


Question: Can WB-MRI predict early treatment response in patients with MM in patients with and without anemia?


Design: Retrospective study


Participants: 56 patients (40 men, 16 women; mean age, 57.4 ± 9.6 [SD] years) with newly diagnosed MM.


Exclusion criteria: 79 were excluded because they were not newly diagnosed. An additional 72 patients were excluded for the following reasons: whole-body DWI did not show bone marrow involvement, lack of treatment or post-treatment evaluation after baseline WB-MRI, unavailable DWI or modified Dixon sequence on baseline WB-MRI, severe motion artifact on baseline WB-MRI, largest lesion diameter smaller than 5 mm on baseline WB-MRI, and interval of more than 1 week between baseline WB-MRI and laboratory data acquisition.


Methods:  The study included 56 patients with newly diagnosed MM who underwent baseline WB-MRI including DWI and modified Dixon sequences. Two radiologists recorded total burden score using MY-RADS and measured the ADC and FF of diffuse and focal disease sites. Mean values across sites were derived. Interobserver agreement was evaluated, and the mean assessments of the readers were used for further analyses. Presence of deep response after four cycles of induction chemotherapy was recorded. Patients were classified as having anemia if their hemoglobin level was less than 100 g/L. The utility of WBMRI parameters in predicting deep response was assessed.


Main Results:  A total of 24 of 56 patients showed deep response, and 25 of 56 patients had anemia. Among patients without anemia, those with deep response compared with those without deep response had a lower total burden score, a lower ADC, and a higher FF. The combination of these three parameters (optimal cutoffs: ≤ 15 for total burden score, ≤ 0.84 × 10−3 mm2/s for ADC, and > 0.16 for FF) achieved sensitivity of 93.8%, specificity of 93.3%, and accuracy of 93.5% for predicting deep response. In patients with anemia, none of the three parameters were significantly different between patients with and without deep response (all p > .05), and the combination of parameters achieved sensitivity of 56.3%, specificity of 100.0%, and accuracy of 72.0%.


Conclusion:  Low total burden score, low ADC, and high FF from WB-MRI may predict deep response in patients with MM, although only among those without anemia. WB-MRI findings may help guide determination of prognosis and initial treatment selection in MM.


Senior editorial comment: Results are intriguing as the ADC should increase with necrosis of a cellular lesion in the interval and later-on decrease due to fatty metamorphosis. Timing of f/up should dictate how changes evolve.


Take home message:

  • Myeloma Response Assessment and Diagnosis System (MY-RADS) for multiple myeloma can be evaluated by using whole-body MRI
  • Diffusion Weighting Imaging and modified Dixon sequence with ADC measurement and fat fraction measurements can help with assessment of multiple myeloma.

Osteoid Osteoma: Percutaneous CT-guided Cryoablation Is a Safe, Effective, and Durable Treatment Option in Adults

Thomas Le Corroller Thomas VivesJean-Camille MatteiVanessa PaulyDaphné GuenounAlexandre RochwergerPierre Champsaur






Background:  Cryoablation is playing an increasing role in the percutaneous treatment of bone tumors. However, despite its potential advantages over heat-based ablation techniques, the clinical safety and efficacy of cryoablation have not been established for osteoid osteoma treatment.


Question:  How effective is percutaneous CT-guided cryoablation for the treatment of osteoid osteoma in young patients and adults?


Design: Retrospective study


Participants: 50 consecutive patients (median age, 24 years; interquartile range [IQR], 19–38 years; 31 men) who underwent percutaneous CT-guided cryoablation for the treatment of osteoid osteoma between January 2013 and June 2019 in a single institution.


Exclusion criteria: Three patients were excluded because no follow-up data were available after their discharge from the hospital.


Methods: In 30 of 50 patients (60%), the procedure was carried out with the patient under local anesthesia and conscious sedation, with the cryoprobe covering the lesion from an extraosseous position, avoiding direct penetration of the nidus. Clinical and radiologic features, procedure-related data, visual analog scale (VAS) pain scores, complications, and overall success rate were evaluated. Statistical analyses were performed by using the nonparametric Friedman test and Wilcoxon signed rank test for repeated measures.


Main Results: Fifty patients underwent CT-guided cryoablation for the treatment of osteoid osteoma, with a 96% (48 of 50 patients) overall clinical success rate. Of the two patients without clinical success, one patient had incomplete pain relief and the other experienced a recurrence of osteoid osteoma at 11 months, which was successfully treated with a second cryoablation procedure. Three of the 50 patients had minor complications (6%); no major complications were reported.



Conclusion: Osteoid osteoma was safely, effectively, and durably treated with CT-guided percutaneous cryoablation. In the majority of patients, treatment could be performed without general anesthesia, with the cryosphere covering the nidus from an extraosseous position.


Senior editorial comment: Thanks for your work. Cryoablation has been found to be useful for tumors of several organs. Successful management of osteoid osteoma management is no surprise either and this work documents the utility in this domain as well.


Take home message:

  • CT-guided percutaneous cryoablation for osteoid osteomas are a viable and safe method of treatment


Artificial Intelligence Algorithm Improves Radiologist Performance in Skeletal Age Assessment: A Prospective Multicenter Randomized Controlled Trial

David K. Eng Nishith B. KhandwalaJin LongNancy R. FeffermanShailee V. LalaNaomi A. StrubelSarah S. MillaRoss W. FiliceSusan E. SharpAlexander J. TowbinMichael L. FrancavillaSummer L. KaplanKirsten EcklundSanjay P. PrabhuBrian J. DillonBrian M. EveristChristopher G. AntonMark E. BittmanRebecca DennisDavid B. LarsonJayne M. SeekinsCicero T. SilvaArash R. ZandiehCurtis P. LanglotzMatthew P. LungrenSafwan S. Halabi






Background: Previous studies suggest that use of artificial intelligence (AI) algorithms as diagnostic aids may improve the quality of skeletal age assessment, though these studies lack evidence from clinical practice. The purpose of this study was to compare the accuracy and interpretation time of skeletal age assessment on hand radiograph examinations with and without the use of an AI algorithm as a diagnostic aid.


Question: What is the effect of AI algorithm on the quality of skeletal age assessment?


Design: Prospective randomized controlled trial


Methods: The accuracy of skeletal age assessment on hand radiograph examinations was performed with (n = 792) and without (n = 739) the AI algorithm as a diagnostic aid. For examinations with the AI algorithm, the radiologist was shown the AI interpretation as part of their routine clinical work and was permitted to accept or modify it. Hand radiographs were interpreted by 93 radiologists from six centers. The primary efficacy outcome was the mean absolute difference between the skeletal age dictated into the radiologists’ signed report and the average interpretation of a panel of four radiologists not using a diagnostic aid. The secondary outcome was the interpretation time. A linear mixed-effects regression model with random center- and radiologist-level effects was used to compare the two experimental groups.


Main Results:

■ Skeletal age was assessed on hand radiographs by 93 radiologists at six centers without (n = 739 radiographs) and with (n = 792) an artificial intelligence (AI) algorithm. Comparison was made to a reference panel of four experts.

■ Use of the AI algorithm showed a smaller difference in skeletal age compared with the reference panel (5.4 months with AI vs 6.0 months without AI, P = .04). This improvement was seen at five of the six radiology centers.



Conclusion: Use of an artificial intelligence algorithm improved skeletal age assessment accuracy and reduced interpretation times for radiologists, although differences were observed between centers.


Senior editorial comment: Thanks for your work. Excellent use of AI- repeatable and mundane work. Such AI softwares can be placed in clinics where expert radiologists are not available.


Take home message:

  • Improvement in the training methodology and improvement in large database analysis provides a higher level of AI interpretation.
  • Changes in AI implementation will affect clinical care delivery and future workflow for Radiologists.


  • Share