Severity classification of Tenosynovial Giant Cell Tumours on MR imaging

AIM

Current development of novel systemic agents requires identification and monitoring of extensive Tenosynovial Giant Cell Tumours (TGCT). This study defines TGCT extension on MR imaging to classify severity.

METHODS

In part one, six MR parameters were defined by field-experts to assess disease extension on MR images: type of TGCT, articular involvement, cartilage-covered bone invasion, and involvement of muscular/tendinous tissue, ligaments or neurovascular structures. Inter- and intra-rater agreement were calculated using 118 TGCT MR scans. In part two, the previously defined MR parameters were evaluated in 174 consecutive, not previously used, MR-scans. TGCT severity classification was established based on highest to lowest Hazard Ratios (HR) on first recurrence.

RESULTS

In part one, all MR parameters showed good inter- and intra-rater agreement (Kappa≥0.66). In part two, cartilage-covered bone invasion and neurovascular involvement were rarely appreciated (<13%) and therefore excluded for additional analyses. Univariate analyses for recurrent disease yielded positive associations for type of TGCT HR12.84(95%CI4.60-35.81), articular involvement HR6.00(95%CI2.14-16.80), muscular/tendinous tissue involvement HR3.50(95%CI1.75-7.01) and ligament-involvement HR4.59(95%CI2.23-9.46). With these, a TGCT severity classification was constructed with four distinct severity-stages. Recurrence free survival at 4 years (log rank p < 0.0001) was 94% in mild localized (n56, 1 recurrence), 88% in severe localized (n31, 3 recurrences), 59% in moderate diffuse (n32, 12 recurrences) and 36% in severe diffuse (n55, 33 recurrences).

CONCLUSION

The proposed TGCT severity classification informs physicians and patients on disease extent and risk for recurrence after surgical treatment. Definition of the most severe subgroup attributes to a universal identification of eligible patients for systemic therapy or trials for novel agents.

Differential diagnosis and mutation stratification of desmoid-type fibromatosis on MRI using radiomics

PURPOSE

Diagnosing desmoid-type fibromatosis (DTF) requires an invasive tissue biopsy with β-catenin staining and CTNNB1 mutational analysis, and is challenging due to its rarity. The aim of this study was to evaluate radiomics for distinguishing DTF from soft tissue sarcomas (STS), and in DTF, for predicting the CTNNB1 mutation types.

METHODS

Patients with histologically confirmed extremity STS (non-DTF) or DTF and at least a pretreatment T1-weighted (T1w) MRI scan were retrospectively included. Tumors were semi-automatically annotated on the T1w scans, from which 411 features were extracted. Prediction models were created using a combination of various machine learning approaches. Evaluation was performed through a 100x random-split cross-validation. The model for DTF vs. non-DTF was compared to classification by two radiologists on a location matched subset.

RESULTS

The data included 203 patients (72 DTF, 131 STS). The T1w radiomics model showed a mean AUC of 0.79 on the full dataset. Addition of T2w or T1w post-contrast scans did not improve the performance. On the location matched cohort, the T1w model had a mean AUC of 0.88 while the radiologists had an AUC of 0.80 and 0.88, respectively. For the prediction of the CTNNB1 mutation types (S45 F, T41A and wild-type), the T1w model showed an AUC of 0.61, 0.56, and 0.74.

CONCLUSIONS

Our radiomics model was able to distinguish DTF from STS with high accuracy similar to two radiologists, but was not able to predict the CTNNB1 mutation status.

A Bayesian approach for diagnostic accuracy of malignant peripheral nerve sheath tumors: a systematic review and meta-analysis

Background

Malignant peripheral nerve sheath tumors (MPNST) carry a dismal prognosis and require early detection and complete resection. However, MPNSTs are prone to sampling errors and biopsies or resections are cumbersome and possibly damaging in benign peripheral nerve sheath tumor (BPNST). This study aimed to systematically review and quantify the diagnostic accuracy of noninvasive tests for distinguishing MPNST from BPNST.

Methods

Studies on accuracy of MRI, FDG-PET (fluorodeoxyglucose positron emission tomography), and liquid biopsies were identified in PubMed and Embase from 2000 to 2019. Pooled accuracies were calculated using Bayesian bivariate meta-analyses. Individual level-patient data were analyzed for ideal maximum standardized uptake value (SUV_max) threshold on FDG-PET.

Results

Forty-three studies were selected for qualitative synthesis including data on 1875 patients and 2939 lesions. Thirty-five studies were included for meta-analyses. For MRI, the absence of target sign showed highest sensitivity (0.99, 95% CI: 0.94-1.00); ill-defined margins (0.94, 95% CI: 0.88-0.98); and perilesional edema (0.95, 95% CI: 0.83-1.00) showed highest specificity. For FDG-PET, SUV_max and tumor-to-liver ratio show similar accuracy; sensitivity 0.94, 95% CI: 0.91-0.97 and 0.93, 95% CI: 0.87-0.97, respectively, specificity 0.81, 95% CI: 0.76-0.87 and 0.79, 95% CI: 0.70-0.86, respectively. SUV_max ≥3.5 yielded the best accuracy with a sensitivity of 0.99 (95% CI: 0.93-1.00) and specificity of 0.75 (95% CI: 0.56-0.90).

Conclusions

Biopsies may be omitted in the presence of a target sign and the absence of ill-defined margins or perilesional edema. Because of diverse radiological characteristics of MPNST, biopsies may still commonly be required. In neurofibromatosis type 1, FDG-PET scans may further reduce biopsies. Ideal SUV_max threshold is ≥3.5.

Diagnostic accuracy of grayscale, power Doppler and contrast-enhanced ultrasound compared with contrast-enhanced MRI in the visualization of synovitis in knee osteoarthritis

PURPOSE

To assess the diagnostic accuracy of grayscale (GSUS), power Doppler (PDUS) and contrast-enhanced ultrasound (CEUS) for detecting synovitis in knee osteoarthritis (OA).

METHOD

Patients with different degrees of radiographic knee OA were included prospectively. All underwent GSUS, PDUS, CEUS, and contrast-enhanced magnetic resonance imaging (CE-MRI), on which synovitis was assessed semi-quantitatively. Correlations of synovitis severity on ultrasound based techniques with CE-MRI were determined. Receiver operating characteristic (ROC) analysis was performed to assess diagnostic performance of GSUS, PDUS, and CEUS, for detecting synovitis, using CE-MRI as reference-standard.

RESULTS

In the 31 patients included, synovitis scoring on GSUS and CEUS was significantly correlated (ρ = 0.608, p < 0.001 and ρ = 0.391, p = 0.033) with CE-MRI. For detecting mild synovitis, the area under the curve (AUC) was 0.781 (95 %CI 0.609-0.953) for GSUS, 0.788 (0.622-0.954) for PDUS, and 0.653 (0.452-0.853) for CEUS. Sensitivity and specificity were 0.667 (0.431-0.845) and 0.700 (0.354-0.919) for GSUS, 0.905 (0.682-0.983) and 0.500 (0.201-0.799) for PDUS, and 0.550 (0.320-0.762) and 0.700 (0.354-0.919) for CEUS, respectively. The AUC of GSUS increased to 0.862 (0.735-0.989), 0.823 (0.666-0.979), and 0.885 (0.767-1.000), when combined with PDUS, CEUS, or both, respectively. For detecting moderate synovitis, the AUC of GSUS was higher (0.882 (0.750-1.000)) and no added value of PDUS and CEUS was observed.

CONCLUSIONS

GSUS has limited overall accuracy for detecting synovitis in knee OA. When GSUS is combined with PDUS or CEUS, overall diagnostic performance improves for detecting mild synovitis, but not for moderate synovitis.

Preoperative Classification of Peripheral Nerve Sheath Tumors on MRI Using Radiomics

Malignant peripheral nerve sheath tumors (MPNSTs) are aggressive soft-tissue tumors prevalent in neurofibromatosis type 1 (NF1) patients, posing a significant risk of metastasis and recurrence. Current magnetic resonance imaging (MRI) imaging lacks decisiveness in distinguishing benign peripheral nerve sheath tumors (BPNSTs) and MPNSTs, necessitating invasive biopsies. This study aims to develop a radiomics model using quantitative imaging features and machine learning to distinguish MPNSTs from BPNSTs. Clinical data and MRIs from MPNST and BPNST patients (2000-2019) were collected at a tertiary sarcoma referral center. Lesions were manually and semi-automatically segmented on MRI scans, and radiomics features were extracted using the Workflow for Optimal Radiomics Classification (WORC) algorithm, employing automated machine learning. The evaluation was conducted using a 100× random-split cross-validation. A total of 35 MPNSTs and 74 BPNSTs were included. The T1-weighted (T1w) MRI radiomics model outperformed others with an area under the curve (AUC) of 0.71. The incorporation of additional MRI scans did not enhance performance. Combining T1w MRI with clinical features achieved an AUC of 0.74. Experienced radiologists achieved AUCs of 0.75 and 0.66, respectively. Radiomics based on T1w MRI scans and clinical features show some ability to distinguish MPNSTs from BPNSTs, potentially aiding in the management of these tumors.

Osteosclerotic Metaphyseal Dysplasia Due to a Likely Pathogenic LRRK1 Variant as a Cause of Recurrent Long Bone Fractures

Osteosclerotic metaphyseal dysplasia (OSMD) is a very rare autosomal-recessive disease caused by mutations in the leucine-rich repeat kinase 1 (LRRK1) gene. It is a sclerosing skeletal dysplasia characterized by osteosclerosis of the long bones, predominantly at the metaphyses and vertebrae. Phenotypic features can be short stature, pathological fractures, delayed development, and hypotonia, but they are not uniformly present, and relatively few cases are known from the literature. A 40-year-old man was seen at our bone center because of nonspontaneous multiple peripheral low-energy trauma fractures since puberty. He had no other complaints and his family history was negative. Except for a relatively short stature (167 cm; -1.5 SD), there were no abnormalities on examination, including laboratory tests. Initially, a suspicion was raised of osteogenesis imperfecta, but bone mineral density was high and X-rays of the whole skeleton showed osteosclerosis of the metaphyses of long bones and vertebrae. Whole-exome sequencing showed a homozygous, likely pathogenic, variant (American College of Medical Genetics and Genomics criteria class 4) in the LRRK1 gene, fitting the diagnosis of OSMD. In conclusion, we described a 40-year-old patient with osteosclerotic metaphyseal dysplasia caused by a homozygous variant in the LRRK1 gene, resulting in multiple fractures of the long bones without other features of the disease, adding to the phenotypic variation of OSMD. © 2023 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.

Reliability and agreement of manual and automated morphological radiographic hip measurements

Objective

To determine the reliability and agreement of manual and automated morphological measurements, and agreement in morphological diagnoses.

Methods

Thirty pelvic radiographs were randomly selected from the World COACH consortium. Manual and automated measurements of acetabular depth-width ratio (ADR), modified acetabular index (mAI), alpha angle (AA), Wiberg center edge angle (WCEA), lateral center edge angle (LCEA), extrusion index (EI), neck-shaft angle (NSA), and triangular index ratio (TIR) were performed. Bland-Altman plots and intraclass correlation coefficients (ICCs) were used to test reliability. Agreement in diagnosing acetabular dysplasia, pincer and cam morphology by manual and automated measurements was assessed using percentage agreement. Visualizations of all measurements were scored by a radiologist.

Results

The Bland-Altman plots showed no to small mean differences between automated and manual measurements for all measurements except for ADR. Intraobserver ICCs of manual measurements ranged from 0.26 (95%-CI 0-0.57) for TIR to 0.95 (95%-CI 0.87-0.98) for LCEA. Interobserver ICCs of manual measurements ranged from 0.43 (95%-CI 0.10-0.68) for AA to 0.95 (95%-CI 0.86-0.98) for LCEA. Intermethod ICCs ranged from 0.46 (95%-CI 0.12-0.70) for AA to 0.89 (95%-CI 0.78-0.94) for LCEA. Radiographic diagnostic agreement ranged from 47% to 100% for the manual observers and 63%-96% for the automated method as assessed by the radiologist.

Conclusion

The automated algorithm performed equally well compared to manual measurement by trained observers, attesting to its reliability and efficiency in rapidly computing morphological measurements. This validated method can aid clinical practice and accelerate hip osteoarthritis research.

AI in radiological imaging of soft-tissue and bone tumours: a systematic review evaluating against CLAIM and FUTURE-AI guidelines

BACKGROUND

Soft-tissue and bone tumours (STBT) are rare, diagnostically challenging lesions with variable clinical behaviours and treatment approaches. This systematic review aims to provide an overview of Artificial Intelligence (AI) methods using radiological imaging for diagnosis and prognosis of these tumours, highlighting challenges in clinical translation, and evaluating study alignment with the Checklist for AI in Medical Imaging (CLAIM) and the FUTURE-AI international consensus guidelines for trustworthy and deployable AI to promote the clinical translation of AI methods.

METHODS

The systematic review identified literature from several bibliographic databases, covering papers published before 17/07/2024. Original research published in peer-reviewed journals, focused on radiology-based AI for diagnosis or prognosis of primary STBT was included. Exclusion criteria were animal, cadaveric, or laboratory studies, and non-English papers. Abstracts were screened by two of three independent reviewers to determine eligibility. Included papers were assessed against the two guidelines by one of three independent reviewers. The review protocol was registered with PROSPERO (CRD42023467970).

FINDINGS

The search identified 15,015 abstracts, from which 325 articles were included for evaluation. Most studies performed moderately on CLAIM, averaging a score of 28.9 ± 7.5 out of 53, but poorly on FUTURE-AI, averaging 5.1 ± 2.1 out of 30.

INTERPRETATION

Imaging-AI tools for STBT remain at the proof-of-concept stage, indicating significant room for improvement. Future efforts by AI developers should focus on design (e.g. defining unmet clinical need, intended clinical setting and how AI would be integrated in clinical workflow), development (e.g. building on previous work, training with data that reflect real-world usage, explainability), evaluation (e.g. ensuring biases are evaluated and addressed, evaluating AI against current best practices), and the awareness of data reproducibility and availability (making documented code and data publicly available). Following these recommendations could improve clinical translation of AI methods.

FUNDING

Hanarth Fonds, ICAI Lab, NIHR, EuCanImage.

The Association Between the Development of Cam Morphology During Skeletal Growth in High-Impact Athletes and the Presence of Cartilage Loss and Labral Damage in Adulthood: A Prospective Cohort Study With a 12-Year Follow-up

BACKGROUND

Cam morphology develops during skeletal growth, but its influence on cartilage and the labrum in high-impact athletes later in life is unknown.

PURPOSE

To (1) explore the association between the presence and duration of cam morphology during adolescence and the cartilage and labral status 7 to 12 years later and (2) report the prevalence of cartilage loss and labral damage in a population of young male athletes (<32 years old) who played professional soccer during skeletal growth.

STUDY DESIGN

Cohort study (Prognosis); Level of evidence, 2.

METHODS

A total of 89 healthy male academy soccer players from the Dutch soccer club Feyenoord (aged 12-19 years) were included at baseline. At baseline and 2.5- and 5-year follow-ups, standardized supine anteroposterior pelvis and frog-leg lateral radiographs of each hip were obtained. At 12-year follow-up, magnetic resonance imaging of both hips was performed. Cam morphology was defined by a validated alpha angle ≥60° on radiographs at baseline or 2.5- or 5-year follow-up when the growth plates were closed. Hips with the presence of cam morphology at baseline or at 2.5-year follow-up were classified as having a "longer duration" of cam morphology. Hips with cam morphology only present since 5-year follow-up were classified as having a "shorter duration" of cam morphology. At 12-year follow-up, cartilage loss and labral abnormalities were assessed semiquantitatively. Associations were estimated using logistic regression, adjusted for age and body mass index.

RESULTS

Overall, 35 patients (70 hips) with a mean age of 28.0 ± 2.0 years and mean body mass index of 24.1 ± 1.8 participated at 12-year follow-up. Cam morphology was present in 56 of 70 hips (80%). The prevalence of cartilage loss was 52% in hips with cam morphology and 21% in hips without cam morphology (adjusted odds ratio, 4.52 [95% CI, 1.16-17.61]; P = .03). A labral abnormality was present in 77% of hips with cam morphology and in 64% of hips without cam morphology (adjusted odds ratio, 1.99 [95% CI, 0.59-6.73]; P = .27). The duration of cam morphology did not influence these associations.

CONCLUSION

The development of cam morphology during skeletal growth was associated with future magnetic resonance imaging findings consistent with cartilage loss in young adults but not with labral abnormalities.

Asymptomatic female softball pitchers have altered hip morphology and cartilage composition

Few studies have explored hip morphology and cartilage composition in female athletes or the impact of asymmetric repetitive loading, such as occurs during softball pitching. The current cross-sectional study assessed bilateral bony hip morphology on computed tomography imaging in collegiate-level softball pitchers ('Pitch1', n = 25) and cross-country runners ('Run', n = 13). Magnetic resonance imaging was used to assess cartilage relaxation times in a second cohort of pitchers ('Pitch2', n = 10) and non-athletic controls ('Con', n = 4). Pitch1 had 52% greater maximum alpha angle than Run (p < 0.001) and were 21.3 (95% CI 2.4 to 192.0) times more likely to have an alpha angle ≥ 60° within at least one hip. Pitch2 had longer T2 relaxation times in the superior femoral cartilage of the drive leg (same side as the throwing arm) and stride leg than Con (all p < 0.02). The drive leg in Pitch2 had longer T1ρ and T2 relaxation times in the superior femoral cartilage compared to the stride leg (all p ≤ 0.03). Asymptomatic softball pitchers exhibit altered bony hip morphology and cartilage composition compared to cross-country runners and non-athletic controls, respectively. They also exhibit asymmetry in cartilage composition. Further studies with larger sample sizes are warranted and any potential long-term consequences of the changes in terms of symptom and osteoarthritis development requires investigation.

Minimally interactive segmentation of soft-tissue tumors on CT and MRI using deep learning

BACKGROUND

Segmentations are crucial in medical imaging for morphological, volumetric, and radiomics biomarkers. Manual segmentation is accurate but not feasible in clinical workflow, while automatic segmentation generally performs sub-par.

PURPOSE

To develop a minimally interactive deep learning-based segmentation method for soft-tissue tumors (STTs) on CT and MRI.

MATERIAL AND METHODS

The interactive method requires the user to click six points near the tumor's extreme boundaries in the image. These six points are transformed into a distance map and serve, with the image, as input for a convolutional neural network. A multi-center public dataset with 514 patients and nine STT phenotypes in seven anatomical locations, with CT or T1-weighted MRI, was used for training and internal validation. For external validation, another public dataset was employed, which included five unseen STT phenotypes in extremities on CT, T1-weighted MRI, and T2-weighted fat-saturated (FS) MRI.

RESULTS

Internal validation resulted in a dice similarity coefficient (DSC) of 0.85 ± 0.11 (mean ± standard deviation) for CT and 0.84 ± 0.12 for T1-weighted MRI. External validation resulted in DSCs of 0.81 ± 0.08 for CT, 0.84 ± 0.09 for T1-weighted MRI, and 0.88 ± 0.08 for T2-weighted FS MRI. Volumetric measurements showed consistent replication with low error internally (volume: 1 ± 28 mm3, r = 0.99; diameter: - 6 ± 14 mm, r = 0.90) and externally (volume: - 7 ± 23 mm3, r = 0.96; diameter: - 3 ± 6 mm, r = 0.99). Interactive segmentation time was considerably shorter (CT: 364 s, T1-weighted MRI: 258s) than manual segmentation (CT: 1639s, T1-weighted MRI: 1895s).

CONCLUSION

The minimally interactive segmentation method effectively segments STT phenotypes on CT and MRI, with robust generalization to unseen phenotypes and imaging modalities.

KEY POINTS

Question Can this deep learning-based method segment soft-tissue tumors faster than can be done manually and more accurately than other automatic methods? Findings The minimally interactive segmentation method achieved accurate segmentation results in internal and external validation, and generalized well across soft-tissue tumor phenotypes and imaging modalities. Clinical relevance This minimally interactive deep learning-based segmentation method could reduce the burden of manual segmentation, facilitate the integration of imaging-based biomarkers (e.g., radiomics) into clinical practice, and provide a fast, semi-automatic solution for volume and diameter measurements (e.g., RECIST).