RPS 1010a - Artificial intelligence (AI) and new techniques in MRI

RPS 1010a-1
Computer-assisted diagnosis of hip dysplasia and femoroacetabular impingement FAI using automatic reconstruction of MRI-based 3D models of the hip joint: a deep learning-based study
Purpose: Hip dysplasia and FAI are complex three-dimensional hip pathologies that can cause hip pain in young and active patients of child-bearing age. Imaging is static and based on 2D pelvic radiographs. Computer-assisted CT-based diagnosis of hip dysplasia and FAI was introduced for patient-specific planning of surgical treatment but MRI-based 3D models would offer a radiation-free alternative.
This study explores the difference between automatic and manual MRI-based 3D-models and if diagnostic parameters correlate between automatic and manual MR-based 3D-models.
Methods: We performed an IRB-approved comparative, retrospective study of 31 hips of 26 symptomatic patients with FAI and hip dysplasia. We compared CT-and MR-based osseous 3D models of the hip joint of the same patients. A 3D-CT scan (isovoxel: 1 mm3) of the entire pelvis and the distal femoral condyles were obtained. Preoperative MR arthrograms of the hip were obtained including axial-oblique T1 3D-VIBE sequence (0.8 mm3 isovoxel) and two axial anisotropic (1.2 x1.2 x 1 mm) 3D-T1 VIBE DIXON sequences of the entire pelvis and the distal femoral condyles. Automatic reconstruction of MRI-based 3D models was performed using machine-learning (deep learning). Threshold-based manual reconstruction of CT- and MR-based 3D-models was performed using commercial software (AMIRA).
Results: The difference between MRI-based 3D-models was below 1 mm. The median difference was 0.2±0.1 mm for the proximal femur and 0.2±0.5 mm for the acetabulum. The Dice coefficient was 97% for the acetabulum and 98% for the femur.
The correlation for 6 diagnostic parameters was excellent and significant (r=0.99, p<0.001) between automatic and manual MR-based 3D-models. The absolute difference was below 2°.
Conclusion: Automatic MRI-based 3D models can replace CT-based 3D-models for patients of childbearing age with hip dysplasia and FAI.
Limitations: A retrospective study.
Ethics: IRB approval was obtained.
Funding: No funding was received for this work.
RPS 1010a-2
To test the ability of artificial intelligence to differentiate between benign and malignant soft tissue masses in ultrasonography
Purpose: Ultrasound evaluation of superficial soft tissue masses is increasingly utilised in routine clinical practice. However, differentiating benign from malignant aetiologies may be challenging. Deep convolutional neural networks (CNNs) have shown the potential to classify images with good accuracy. We propose to test the ability of this methodology to differentiate between benign and malignant superficial soft tissue masses.
Methods: A total of 760 ultrasound images from focused ultrasounds of superficial soft tissue masses were selected to train a CNN classifier. The dataset consisted of 523 images of benign soft tissue masses and 237 images of malignant masses, all confirmed by pathologic diagnosis. We selected the pre-trained VGG-16 architecture implemented on Keras. We performed a binary classification (benign and malignant). Then, we tried to differentiate between 3 subgroups of benign masses: lipomas, benign nerve sheath tumours, and vascular malformations.
Results: The test accuracy of the model to determine if the mass was benign or malignant mass was 0.75. The CNN outperformed two experienced musculoskeletal radiologists on a test sample dataset. The accuracy of the model to differentiate between the 3 benign subgroups test dataset was 0.66.
Conclusion: As an initial step, artificial intelligence (AI) shows promise in the differentiation of benign from malignant soft masses, outperforming two musculoskeletal radiologists. The addition of a larger variety and number of soft tissue masses will be necessary to establish the utility of AI as a useful tool going forward.
Limitations: The limited size of our dataset restricted the number of subgroups that could be investigated. We tested the accuracy of only two radiologists, which does not necessarily reflect the accuracy of radiologists with varying experience.
Ethics: This study has institutional board review approval.
Funding: No funding was received for this work.
RPS 1010a-4
Paediatric radiographic detection of the acute distal tibial fracture using trained AI-networks
Purpose: Tibial fracture detection is a routine task in paediatric radiology. Its precision depends on knowledge of anatomy and experience-based training. Until recently, there were no computer-aided detection tools due to anatomical complexity, a variety of 2d projections, additional objects in the image, the presentation of pathology, the age of patients, and gender.
Methods: We presented a CNN-based system that can automatise radiographic distal tibia fracture detection. The system used two connected CNNs. The first CNN was trained to localise the distal tibia using 500 normal and 500 pathological studies, including displaced and nondisplaced fractures. The images were manually annotated with a box containing the distal tibia. This CNN had a 98% accuracy with zero false-positives. The second CNN was chosen from previously developed CNNs trained on a different, broader set of manually segmented distal tibia images and had undergone the 10 fold cross-validation. The best CNN, which formerly achieved an accuracy of 95.9%, was selected for validation in this study.
Results: The whole system combined two models into the pipeline. An x-ray image was read and the distal tibia localised as an ROI. The ROI was transferred to the fracture detection algorithm. If the distal tibia was not localised, then it was not processed by the second model. The whole system achieved an area under the receiver-operating curve (AUC-ROC) of 0.89.
Conclusion: This method could become a computer-aided diagnosis (CAD) tool, accelerating the workflow in paediatric radiology departments.
Limitations: This study was limited to images acquired and read at a single institution.
Ethics: This work was conducted under a protocol approved by the Institutional Review Board.
Funding: No funding was received for this work.
RPS 1010a-5
Quantitative ultrashort echo time MRI (UTE-MRI) for the diagnosis of early cartilage degeneration: a comparison with T2 mapping (recorded)
Purpose: To investigate the feasibility of using a quantitative ultrashort echo time MRI (UTE-MRI) technique to diagnose early cartilage degeneration and to compare the technique?s diagnostic efficacy with T2 mapping.
Methods: 20 patients with a total knee arthroplasty were obtained and underwent an MR scan on a clinical 3.0T scanner. 72 regions of interest were manually drawn on specimens for UTE-adiabaticT1?, UTE-T2*, and T2 measurement, and the corresponding cartilage-bone regions were further divided into normal (n=11, Mankin scores 0-1), mild (n=28, Mankin scores 2-5), moderate(n=21, Mankin scores 6-9), and severe (n=12, Mankin scores 10-14) degeneration groups based on histological measures of degeneration (Mankin scores) as a reference standard. Statistical analysis was performed to compare the differences among the groups.
Results: The UTE-adiabaticT1? in the normal group was significantly different from the mild group moderate group and severe group. The adiabaticT1?, UTE-T2*, and T2 values correlated significantly with Mankin scores, with UTE-adiabaticT1? and UTE-T2* markedly correlated with Mankin scores, and with T2 values weakly correlated with Mankin scores. The diagnostic efficacy of UTE-adiabatic T1? was better than UTE T2* mapping and T2 mapping for the diagnosis of early cartilage degeneration.
Conclusion: UTE-adiabaticT1? may provide a more promising imaging biomarker with potential application in more comprehensive diagnoses and the monitoring of cartilage degeneration compared to UTE-T2* and T2.
Limitations: The sample size was small. The study was performed ex vivo and the measurement condition was different from the real physiological environment. Additionally, this study could not utilise quantitative biochemical measurements for PG and the collagen of cartilage.
Ethics: The study was approved by Ethics Committee at our Hospital (Ethics Committee approval number: shsy-iec-ky-3964).
Funding: National Natural Science Foundation of China (81871325,81801656).
RPS 1010a-7
Dixon or DWI: quantitative discrimination between malignant and acute osteoporotic vertebral fractures
Purpose: To compare Dixon MRI derived fat-fraction (FF) and ADC to discriminate between malignant and acute osteoporotic vertebral fractures.
Methods: 1.5T MRI of 42 malignant and 27 acute osteoporotic vertebral fractures were retrospectively reviewed. Fractured vertebrae showed extensive hyperintensity on fluid-sensitive fat-suppressed sequences. FF was derived from axial T1 VIBE two-point Dixon images with dedicated in and opposed echo timing by dividing the fat-only sequence through the sum of the fat-plus water-only sequence. ADC maps were calculated from axial b50 and b900 images. ROI-based measurements of mean FF and ADC were performed for each fractured vertebra in the axial plane. A blinded MSK-radiologist classified fractures as malignant or osteoporotic based only on morphologic features, also noting diagnostic confidence. Diagnostic performance was compared by ROC curve analyses and discriminating cut-off values were derived. P<0.05 was deemed significant.
Results: FF and ADC were significantly smaller in malignant fractures (p<0.01). AUC was significantly larger for FF than for ADC (p=0.03). The FF AUC (0.92), sensitivity (86%), and specificity (85%) were not significantly inferior compared to the radiologist's interpretations (0.93, 98%, and 89%; each p>0.05). ADC AUC (0.78) and sensitivity (62%) were significantly smaller compared to the radiologist (each p<0.01). Low-reader confidence was reported in 11% of fractures. In this subgroup, the specificity of FF and ADC (83% each) exceeded the radiologist?s specificity (50%).
Conclusion: Quantitative FF and ADC can distinguish malignant from acute osteoporotic fractures. FF showed superior diagnostic accuracy over ADC and was non-inferior to an MSK-radiologist?s evaluation.
Limitations: The study size, retrospective design, lack of histopathology, and recall bias.
Ethics: This study follows the regulations of the local ethics committee. Patient consent was waived.
Funding: No funding was received for this work.
RPS 1010a-8
Shape-based machine learning for three-dimensional phenotyping of the lumbosacral spine and dural sac: a prediction of Fibrillin-1 gene mutations pathogenic for Marfan syndrome
Purpose: To test our hypothesis that three-dimensional phenotyping of the lumbosacral spine and dural sac using shape-based machine learning enables the prediction of Fibrillin-1 gene mutations pathogenic for Marfan syndrome.
Methods: 184 patients being evaluated for Marfan syndrome, 01/2012-12/2016, underwent 1.5T-MRI with 3D T2-weighted TSE-sequence of the lumbosacral spine with 1x1x1 mm³ spatial resolution. 110 patients (32.4±11.6 years, 50 female) agreed to genetic testing; 38 had a Fibrillin-1 gene mutation pathogenic for Marfan syndrome (FBN1+) and 72 were tested negative (FBN1-). Shape-based machine learning allowed three-dimensional segmentation and quantification of volumes and volume ratios of vertebral bodies L3-L5 and dural sac segments L3-S1.
Results: Dural sac volumes were significantly enlarged in FBN1+ vs FBN1- patients (in mL): segment L3 12.5±3.0 vs 10.3±2.3, L4 11.8±3.5 vs 9.0±2.5, L5 11.7±4.4 vs 7.9±3.1, and S1 13.5±8.7 vs 5.3±3.1 (all p<0.001). Vertebral body volumes did not significantly differ between FBN1+ vs FBN1- patients. ROC analysis for the identification of FBN1+ patients showed AUCs of 0.797 for total lumbosacral dural sac volume, 0.660 for L3 volume ratio, 0.694 for L4 volume ratio, 0.766 for L5 volume ratio, 0.786 for S1 dural sac volume, and 0.753 for S1 to L4 dural sac volume ratio. Combined sensitivity, specificity, and positive and negative predictive values for identification of FBN1+ patients were 69%, 93%, 84%, and 85%.
Conclusion: Fibrillin-1 gene mutations were associated with a significantly altered three-dimensional phenotype of the lumbosacral spine and dural sac. Shape-based machine learning allowed identification of patients with Fibrillin-1 gene mutations with a high specificity and moderate sensitivity.
Limitations: Potential sources of bias are the retrospective study design and secondary exclusion of those patients without genetic testing.
Ethics: An ethics committee approved the study and waived written informed consent.
Funding: No funding was received for this work.
RPS 1010a-9
Radiology improving the diagnosis and management of patients suffering with Ehlers-Danlos syndrome. An upright biomechanical MRI study
Purpose: The incidence of the hypermobile Ehlers-Danlos syndrome (hEDS) is larger than is generally understood, with many patients being misdiagnosed as suffering from ME or fibromyalgia. It is important to make the correct diagnosis to enable appropriate treatment.
Methods: 112 consecutive patients with hEDS had MRI scans of the cervical spine and craniocervical junction performed in an upright MRI scanner, where the cervical spine angle was measured with the neck in neutral flexion and extension, and compared with an age-sex matched control group. MRI of the atlantoaxial joint was performed looking forwards, to the right, and to the left to assess for evidence of laxity of the craniocervical ligamentous complex. The cerebellar tonsillar station was also recorded.
Results: Patients with hEDS showed a decrease in their cervical spine angle in neutral 20% less than the control group (16º vs 20º), a 50% larger ability to flex the neck (-20º vs -11º), and a 45% greater angle in extension (55º vs 32º). 72% showed ligamentous laxity/instability of the cervical spine. 80% showed evidence of atlantoaxial instability.
43.5% demonstrated hindbrain herniation of greater than 5 mm, with no evidence of syringomyelia in a pattern distinctly different from the Chiari malformation.
Conclusion: It is important to be able to make accurate measurements of increased mobility and laxity of both the cervical spine and atlantoaxial joints in patients suffering from hEDS to allow for referral for the appropriate treatment, including surgery where indicated. It is also important for the reassurance of these patients, who have a physical disability, that they do have hEDS and are not psychologically disturbed or suffering from ME or fibromyalgia.
Limitations: n/a
Ethics: n/a
Funding: No funding was received for this work.
RPS 1010a-10
Can sacrum height predict body height, age, and sex? A large population-based MRI study
Purpose: Post-mortem identification of unknown deceased persons and age estimation in the living are central tasks of forensic scientists. Body height is traditionally derived from measurements of long bones. The sacrum is a centrally located bone that is relatively unsusceptible to environmental factors. The aim of the study was to investigate if sacrum height can be used to estimate body height, weight, age, and sex using datasets from a population-based magnetic resonance imaging (MRI) study.
Methods: A total of 2,499 whole-body MRIs were evaluated. Sacrum height was measured in the median plane of sagittal T2-weighted images of the spine and associations with age, sex, and body height were analysed. Linear regression was used to derive formulas for body height estimation.
Results: Male participants had significantly (p<0.001) greater sacrum height (11.4 cm, standard deviation 1.1 cm, range: 7.9-14.6 cm) compared to female participants (10.9 cm, standard deviation 1.0 cm, range: 7.3-14.5 cm). The Pearson-correlation was r=0.44 for body height. The mean absolute difference (MAD) between calculated and measured body height was 4.9 cm for females and 4.8 cm for males. ROC cut-off to classify sex had a 60% sensitivity and 57.3% specificity.
Conclusion: Contrary to previous reports, sacrum height alone cannot be recommended to predict body height, age, or sex. It is not as reliable as predictions based on large bones.
Limitations: A population-based MR-study from a specific region of Germany. Not all participants in the study were included in the final analysis. There was no correlation with established techniques for age, sex, and height determination based on the analysis of long bones.
Ethics: Approved by the local ethics committee and all participants gave their written informed consent.
Funding: No funding was received for this work.
RPS 1010a-11
The use of whole-body MRI in chronic recurrent multifocal osteomyelitis in children: our experience in 29 patients
Purpose: To describe the most common locations of involvement in chronic recurrent multifocal osteomyelitis (CRMO) in children using whole-body magnetic resonance imaging (WB-MRI).
Methods: We retrospectively reviewed 29 patients (12 girls, 17 boys) with a diagnosis of CRMO from 2014-May 2019. All patients underwent WB-MRI using STIR sequences in axial and coronal planes, and sagittal planes to evaluate the spine. Imaging data was evaluated by a paediatric radiologist with more than 10 years of experience and a radiology fellow with 3 years of experience.
Results: The mean age at diagnosis was 11 years (3-16 years), but the mean age of clinical presentation was 10 years.
A total of 175 lesions were investigated. The lesions were multifocal in 28/29 cases (96.5%) and the number of lesions encountered per patient ranged from 1-28 lesions.
The most common sites of involvement were the pelvis and femur (21%), followed by tibia (19.3%), and the spine and humerus (7.4%). Other less frequent locations included clavicles (4.5%), tarsus (4%), foot (3.4%), radius and cuboid (2.8%), fibula (2.3%), and the sternum (1.7%). The involvement of the patella, ribs, carpus, and scapula was exceedingly rare, accounting for 0.6%.
The metaphysis of long bones was most frequently involved (56.4%), followed by epiphysis (13.6%). Bilateral lesions were 42.9%.
Conclusion: WB-MRI is the diagnostic modality of choice in patients with a clinical diagnosis of CRMO. Bilateral involvement of metaphysis and epiphysis of long bones, and the pelvis, are the most frequent areas of involvement.
Limitations: n/a
Ethics: n/a
Funding: No funding was received for this work.
RPS 1010a-12
Clinical validation of a deep learning-based bone age software: a feasibility study (recorded)
Purpose: To evaluate the clinical performance of deep learning (DL)-based software for bone age assessment in a paediatric endocrinology clinic in South Korea.
Methods: We compared the bone age estimated from commercial DL-based software (BoneAge, Vuno, Seoul, Korea) with the bone ages independently estimated by 3 physicians (1 musculoskeletal radiologist, 1 paediatric endocrinologist, and 1 radiology resident) using a Greulich-Pyle atlas. In total, 109 children (8 boys, 101 girls) aged between 4 and 17 years who visited the paediatric endocrinology clinic from December 2018 were enrolled. Fisher?s exact test (two sides), Pearson?s correlation coefficient, and root mean squared error (RMSE) were used to compare the bone ages of DL software and those of the 3 physicians. An intraclass correlation coefficient (ICC) evaluated inter-rater variation.
Results: Fisher?s exact test showed significant differences between DL bone age and all 3 physicians? bone age (p<0.025). There were good correlations (r=0.93, 0.92, and 0.9, p<0.05), however, the RMSE values were 9.7, 13.1, and 9.9 months between the bone age of DL software and those of the musculoskeletal radiologist, paediatric endocrinologist, and radiology resident, respectively. RMSE values between the 3 physicians were 12.3, 18, and 14.8 months, with relatively good correlation with each other (r=0.88, 0.89, and 0.88, p<0.05). ICC values between the 3 physicians were 0.93, 0.84, and 0.87 each, with an overall ICC value of 0.92.
Conclusion: Bone age estimated by DL-based software showed good correlation with the bone ages estimated using a traditional Greulich-Pyle atlas by 3 physicians but showed statistically different values.
Limitations: A retrospective, single-centre study with a small number of patients
Ethics: n/a
Funding: No funding was received for this work.


Luca Maria Sconfienza (Italy)

Reto Sutter (Switzerland)

European Society of Radiology

Watch ECR 2020 live

This session is part of ECR 2020 Live. Please register for ECR 2020 Live in order to get access.

    WITHOUT CME €39.00
    FROM HOME €350.00