Search

The accurate classification of bone tumours is crucial for guiding clinical decisions regarding treatment and follow-up. However, differentiating between various tumour types is challenging due to the rarity of certain entities, high intra-class variability, and limited training data in clinical practice. This study proposes a multimodal deep learning model that integrates clinical metadata and X-ray imaging to improve the classification of primary bone tumours. The dataset comprises 1,785 radiographs from 804 patients collected between 2000 and 2020, including metadata such as age, affected bone site, tumour position, and gender. Ten tumour types were selected, with histopathology or tumour board decisions serving as the reference standard., Methods: Our model is based on the NesT image classification model and a multilayer perceptron with a joint fusion architecture. Descriptive statistics included incidence and percentage ratios for discrete parameters, and mean, standard deviation, median, and interquartile range for continuous parameters., Results: The mean age of the patients was 33.62 ± 18.60 years, with 54.73% being male. Our multimodal deep learning model achieved 69.7% accuracy in classifying primary bone tumours, outperforming the Vision Transformer model by five percentage points. SHAP values indicated that age had the most substantial influence among the considered metadata., Conclusion: The joint fusion approach developed in this study, integrating clinical metadata and imaging data, outperformed state-of-the-art models in classifying primary bone tumours. The use of SHAP values provided insights into the impact of different metadata on the model's performance, highlighting the significant role of age. This approach has potential implications for improving diagnostic accuracy and understanding the influence of clinical factors in tumour classification., (© 2024. The Author(s).)

Objectives: To develop an algorithm to link undiagnosed patients to previous patient histories based on radiographs, and simultaneous classification of multiple bone tumours to enable early and specific diagnosis., Materials and Methods: For this retrospective study, data from 2000 to 2021 were curated from our database by two orthopaedic surgeons, a radiologist and a data scientist. Patients with complete clinical and pre-therapy radiographic data were eligible. To ensure feasibility, the ten most frequent primary tumour entities, confirmed histologically or by tumour board decision, were included. We implemented a ResNet and transformer model to establish baseline results. Our method extracts image features using deep learning and then clusters the k most similar images to the target image using a hash-based nearest-neighbour recommender approach that performs simultaneous classification by majority voting. The results were evaluated with precision-at-k, accuracy, precision and recall. Discrete parameters were described by incidence and percentage ratios. For continuous parameters, based on a normality test, respective statistical measures were calculated., Results: Included were data from 809 patients (1792 radiographs; mean age 33.73 ± 18.65, range 3-89 years; 443 men), with Osteochondroma (28.31%) and Ewing sarcoma (1.11%) as the most and least common entities, respectively. The dataset was split into training (80%) and test subsets (20%). For k = 3, our model achieved the highest mean accuracy, precision and recall (92.86%, 92.86% and 34.08%), significantly outperforming state-of-the-art models (54.10%, 55.57%, 19.85% and 62.80%, 61.33%, 23.05%)., Conclusion: Our novel approach surpasses current models in tumour classification and links to past patient data, leveraging expert insights., Clinical Relevance Statement: The proposed algorithm could serve as a vital support tool for clinicians and general practitioners with limited experience in bone tumour classification by identifying similar cases and classifying bone tumour entities., Key Points: • Addressed accurate bone tumour classification using radiographic features. • Model achieved 92.86%, 92.86% and 34.08% mean accuracy, precision and recall, respectively, significantly surpassing state-of-the-art models. • Enhanced diagnosis by integrating prior expert patient assessments., (© 2024. The Author(s).)

Background: Chondrosarcomas are rare malignant bone tumors diagnosed by analyzing radiological images and histology of tissue biopsies and evaluating features such as matrix calcification, cortical destruction, trabecular penetration, and tumor cell entrapment., Methods: We retrospectively analyzed 16 cartilaginous tumor tissue samples from three patients (51-, 54-, and 70-year-old) diagnosed with a dedifferentiated chondrosarcoma at the femur, a moderately differentiated chondrosarcoma in the pelvis, and a predominantly moderately differentiated chondrosarcoma at the scapula, respectively. We combined a hematein-based x-ray staining with high-resolution three-dimensional (3D) microscopic x-ray computed tomography (micro-CT) for nondestructive 3D tumor assessment and tumor margin evaluation., Results: We detected trabecular entrapment on 3D micro-CT images and followed bone destruction throughout the volume. In addition to staining cell nuclei, hematein-based staining also improved the visualization of the tumor matrix, allowing for the distinction between the tumor and the bone marrow cavity. The hematein-based staining did not interfere with further conventional histology. There was a 5.97 ± 7.17% difference between the relative tumor area measured using micro-CT and histopathology (p = 0.806) (Pearson correlation coefficient r = 0.92, p = 0.009). Signal intensity in the tumor matrix (4.85 ± 2.94) was significantly higher in the stained samples compared to the unstained counterparts (1.92 ± 0.11, p = 0.002)., Conclusions: Using nondestructive 3D micro-CT, the simultaneous visualization of radiological and histopathological features is feasible., Relevance Statement: 3D micro-CT data supports modern radiological and histopathological investigations of human bone tumor specimens. It has the potential for being an integrative part of clinical preoperative diagnostics., Key Points: • Matrix calcifications are a relevant diagnostic feature of bone tumors. • Micro-CT detects all clinically diagnostic relevant features of x-ray-stained chondrosarcoma. • Micro-CT has the potential to be an integrative part of clinical diagnostics., (© 2024. The Author(s).)

Objectives: To develop and validate machine learning models to distinguish between benign and malignant bone lesions and compare the performance to radiologists., Methods: In 880 patients (age 33.1 ± 19.4 years, 395 women) diagnosed with malignant (n = 213, 24.2%) or benign (n = 667, 75.8%) primary bone tumors, preoperative radiographs were obtained, and the diagnosis was established using histopathology. Data was split 70%/15%/15% for training, validation, and internal testing. Additionally, 96 patients from another institution were obtained for external testing. Machine learning models were developed and validated using radiomic features and demographic information. The performance of each model was evaluated on the test sets for accuracy, area under the curve (AUC) from receiver operating characteristics, sensitivity, and specificity. For comparison, the external test set was evaluated by two radiology residents and two radiologists who specialized in musculoskeletal tumor imaging., Results: The best machine learning model was based on an artificial neural network (ANN) combining both radiomic and demographic information achieving 80% and 75% accuracy at 75% and 90% sensitivity with 0.79 and 0.90 AUC on the internal and external test set, respectively. In comparison, the radiology residents achieved 71% and 65% accuracy at 61% and 35% sensitivity while the radiologists specialized in musculoskeletal tumor imaging achieved an 84% and 83% accuracy at 90% and 81% sensitivity, respectively., Conclusions: An ANN combining radiomic features and demographic information showed the best performance in distinguishing between benign and malignant bone lesions. The model showed lower accuracy compared to specialized radiologists, while accuracy was higher or similar compared to residents., Key Points: • The developed machine learning model could differentiate benign from malignant bone tumors using radiography with an AUC of 0.90 on the external test set. • Machine learning models that used radiomic features or demographic information alone performed worse than those that used both radiomic features and demographic information as input, highlighting the importance of building comprehensive machine learning models. • An artificial neural network that combined both radiomic and demographic information achieved the best performance and its performance was compared to radiology readers on an external test set., (© 2022. The Author(s).)

Objectives: To evaluate the performance and reproducibility of MR imaging features in the diagnosis of joint invasion (JI) by malignant bone tumors., Methods: MR images of patients with and without JI (n = 24 each), who underwent surgical resection at our institution, were read by three radiologists. Direct (intrasynovial tumor tissue (ITT), intraarticular destruction of cartilage/bone, invasion of capsular/ligamentous insertions) and indirect (tumor size, signal alterations of epiphyseal/transarticular bone (bone marrow replacement/edema-like), synovial contrast enhancement, joint effusion) signs of JI were assessed. Odds ratios, sensitivity, specificity, PPV, NPV, and reproducibilities (Cohen's and Fleiss' κ) were calculated for each feature. Moreover, the diagnostic performance of combinations of direct features was assessed., Results: Forty-eight patients (28.7 ± 21.4 years, 26 men) were evaluated. All readers reliably assessed the presence of JI (sensitivity = 92-100 %; specificity = 88-100%, respectively). Best predictors for JI were direct visualization of ITT (OR = 186-229, p < 0.001) and destruction of intraarticular bone (69-324, p < 0.001). Direct visualization of ITT was also highly reliable in assessing JI (sensitivity, specificity, PPV, NPV = 92-100 %), with excellent reproducibility (κ = 0.83). Epiphyseal bone marrow replacement and synovial contrast enhancement were the most sensitive indirect signs, but lacked specificity (29-54%). By combining direct signs with high specificity, sensitivity was increased (96 %) and specificity (100 %) was maintained., Conclusion: JI by malignant bone tumors can reliably be assessed on preoperative MR images with high sensitivity, specificity, and reproducibility. Particularly direct visualization of ITT, destruction of intraarticular bone, and a combination of highly specific direct signs were valuable, while indirect signs were less predictive and specific., Key Points: • Direct visualization of intrasynovial tumor was the single most sensitive and specific (92-100%) MR imaging sign of joint invasion. • Indirect signs of joint invasion, such as joint effusion or synovial enhancement, were less sensitive and specific compared to direct signs. • A combination of the most specific direct signs of joint invasion showed best results with perfect specificity and PPV (both 100%) and excellent sensitivity and NPV (both 96 %)., (© 2022. The Author(s).)

Background An artificial intelligence model that assesses primary bone tumors on radiographs may assist in the diagnostic workflow. Purpose To develop a multitask deep learning (DL) model for simultaneous bounding box placement, segmentation, and classification of primary bone tumors on radiographs. Materials and Methods This retrospective study analyzed bone tumors on radiographs acquired prior to treatment and obtained from patient data from January 2000 to June 2020. Benign or malignant bone tumors were diagnosed in all patients by using the histopathologic findings as the reference standard. By using split-sample validation, 70% of the patients were assigned to the training set, 15% were assigned to the validation set, and 15% were assigned to the test set. The final performance was evaluated on an external test set by using geographic validation, with accuracy, sensitivity, specificity, and 95% CIs being used for classification, the intersection over union (IoU) being used for bounding box placements, and the Dice score being used for segmentations. Results Radiographs from 934 patients (mean age, 33 years ± 19 [standard deviation]; 419 women) were evaluated in the internal data set, which included 667 benign bone tumors and 267 malignant bone tumors. Six hundred fifty-four patients were in the training set, 140 were in the validation set, and 140 were in the test set. One hundred eleven patients were in the external test set. The multitask DL model achieved 80.2% (89 of 111; 95% CI: 72.8, 87.6) accuracy, 62.9% (22 of 35; 95% CI: 47, 79) sensitivity, and 88.2% (67 of 76; CI: 81, 96) specificity in the classification of bone tumors as malignant or benign. The model achieved an IoU of 0.52 ± 0.34 for bounding box placements and a mean Dice score of 0.60 ± 0.37 for segmentations. The model accuracy was higher than that of two radiologic residents (71.2% and 64.9%; P = .002 and P < .001, respectively) and was comparable with that of two musculoskeletal fellowship-trained radiologists (83.8% and 82.9%; P = .13 and P = .25, respectively) in classifying a tumor as malignant or benign. Conclusion The developed multitask deep learning model allowed for accurate and simultaneous bounding box placement, segmentation, and classification of primary bone tumors on radiographs. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Carrino in this issue.

The diagnosis of tumors and tumorlike lesions of bone is a routine part of both general and specialist radiologic practices. The spectrum of disorders ranges from the small incidental lesion to the potentially life-limiting malignancies whether primary or secondary. In this review, authored by experts from several European orthopaedic oncology centers, we present a collection of pieces of advice in the form of 10 commandments. Adherence in daily practice to this guidance should help minimize adverse patient experiences and outcomes., Competing Interests: None declared., (Thieme. All rights reserved.)

Chordomas are rare bone malignancies that are thought to arise from remnants of the notochord and usually are located in the axial skeleton. Immunophenotypical matching neoplasms primarily found in appendicular locations, referred to as extra-axial chordoma, are rarely encountered by radiologists, surgeons, and pathologists. Only a few of these cases have been described in the literature with only one intra-articular case with involvement of the knee joint. We present the first case of an intra-articular extra-axial chordoma of the wrist. Diagnostic imaging patterns were initially ambiguous and histopathological reprocessing was crucial in order to determine the diagnosis of an intra-articular neoplasm with co-expression of cytokeratins, S-100 protein, and brachyury.

Objectives: To evaluate the diagnostic value of MR-derived CT-like images and simulated radiographs compared with conventional radiographs in patients with benign and malignant bone tumors., Methods: In 32 patients with a benign or malignant bone lesion (mean age 33.9 ± 18.5 years, 17 females), 3-T MR imaging was performed including a 3D T1-weighted gradient echo sequence as the basis for the CT-like images. From these, intensity-inverted MR image volumes were converted into 2D images via a forward projection to obtain simulated radiographs. Two radiologists assessed these images as well as conventional radiographs for the type of periosteal reaction, matrix mineralization and destruction pattern. Agreement between the modalities was calculated using Cohen's κ., Results: The agreement between conventional radiographs and MR-derived CT-like images in combination with simulated radiographs was substantial (periosteal reaction, κ = 0.67; destruction pattern, κ = 0.75), and the sensitivity of both modalities for the final diagnosis of the lesion (aggressive vs. nonaggressive) was high (MR-derived CT-like images, 86.2% vs. conventional radiographs, 90.0%). Additional information on soft tissue extension (MR-derived CT-like images, 21.9% vs. conventional radiographs, 12.5%; p = 0.009) and lobulation (9.4% vs. 0%; p < 0.001) was significantly more often found on MR-derived CT-like images compared with conventional radiographs., Conclusions: The assessment of the destruction patterns, periosteal reaction and distinction between aggressive and nonaggressive tumors was feasible using MR-derived CT-like images and simulated radiographs and is comparable to that of conventional radiographs. Moreover, MR-derived CT-like images provided additional information on soft tissue extension and tumor architecture., Key Points: • CT-like images and simulated radiographs can be generated from 3D MRI. • Evaluation of bone tumors is feasible with MR-derived images. • CT-like images and simulated radiographs provide additional information on bone tumors.

Background: The purpose of this study was to investigate whether pathological fractures (PF) influence the prognosis of patients with osteosarcoma (OS) or Ewing tumor (ET) regarding 5-year survival, occurrence of metastases, and local recurrence., Methods: We retrospectively analyzed 205 patients with metastatic and nonmetastatic OS or ET. Survival analysis was performed for all patients and differentiated for patients with OS (n = 127) and ET (n = 78) as well as for adults (n = 101) and children (n = 104)., Results: Patients with PF showed survival rates of 64% compared to 83% for those without PF (p = 0.023). Local recurrence occurred in 7% of the patients without and in 24% of those with PF (p = 0.023). In patients with ET and in children, survival analysis showed no significant difference between patients with and without PF in survival and local recurrence rates. In patients with OS, survival rate decreased from 83 to 59% (p = 0.024) and local recurrence rate increased from 13 to 30% (p = 0.042). In adults, survival rate decreased from 78 to 51% (p = 0.004) and local recurrence rate increased from 13 to 42% (p < 0.001). In multivariate analysis, age and PF were associated with inferior survival., Conclusion: This study suggests that the occurrence of PF has a negative impact on survival and implicates an increased risk of local recurrence. In children and in patients with ET, PF did not have a prognostic impact., (© 2018 S. Karger AG, Basel.)

Fifty-nine cases of lesions presenting in the patella were identified after review of the databases of four European bone tumour registries. Of the 59 cases, 46% were non neoplastic, 39% were benign and 15% were malignant. The commonest benign neoplasm was giant cell tumour (GCT) (11 cases). Younger patients were more likely to have a benign neoplasm. Lesions in patients less than 40 years of age included giant cell tumour, chondroblastoma, aneurysmal bone cyst (ABC), osteomyelitis, osteoid osteoma and solitary bone cyst. In patients older than 40 years, the following were common lesions: intra-osseous gout, metastasis and intra-osseous ganglion. Expansion of the patella with thinning of cortex was seen more commonly in GCT and brown tumour in hyperparathyroidism. There was associated soft tissue extension in gout and malignant lesions.

Epithelioid hemangioendothelioma (EH) of bone is a rare vascular neoplasm characterized by epithelioid endothelial cells and a variable biologic behaviour. The principal sites of occurrence of this osteolytic tumor are the lower extremity and the axial skeleton. Approximately half of the cases present with multifocal disease. The latter feature can be helpful in suggesting the diagnosis of a vascular tumor; on the other hand, it strengthens the need for a skeletal survey or whole-body MRI/CT. We report on the clinical, histologic and radiologic features-including CT and MRI findings-of EH in a case of multifocal disease of the phalanges of the hand, a very uncommon anatomic site of affliction.

This article reviews the role of CT and MR imaging in the diagnosis of benign bone tumors and tumor-like lesions of bone with with regard to differential diagnosis, the assessment of tumor-related complications, and the detection of postoperative recurrence. Indications for cross-sectional imaging of specific lesions, including osteoid osteoma, osteoblastoma, enchondroma, osteochondroma, intraosseous lipoma, hemangioma, giant cell tumor, aneurysmal bone cyst, simple bone cyst, and eosinophilic granuloma, are discussed, and advantages and disadvantages of the different imaging modalities are illustrated on the basis of pathologically confirmed cases.

Purpose: To evaluate computed tomography (CT)-guided radiofrequency (RF) ablation as a minimally invasive therapy for osteoid osteoma with regard to technical and clinical success and immediate and delayed complications., Materials and Methods: Forty-seven patients (age range, 8-41 y; mean age, 19.6 y) with osteoid osteomas (femur, n = 25; tibia, n = 15; pelvis, n = 2; humerus, n = 1; ulna, n = 1; talus, n = 1; calcaneus, n = 1; vertebral body, n = 1) were treated with CT-guided RF ablation in 15 cases after one (n = 10) or two (n = 5) unsuccessful attempts at open surgical resection. Percutaneous therapy was performed with use of general or spinal anesthesia. After localization of the nidus with 1-3-mm CT sections, osseous access was established with either a 2-mm coaxial drill system or an 11-gauge Jamshidi needle. RF ablation was performed at 90 degrees C for a period of 4-5 minutes with use of a rigid RF electrode with a diameter of 1 mm. The procedures were regarded as technically successful if the tip of the RF electrode could be placed within the center of the nidus and could be heated to the desired temperature. Clinical success of treatment was defined as permanent relief of pain and return to normal function without additional treatment. In case of persistence or recurrence of symptoms after RF ablation, treatment was regarded as secondarily successful if permanent relief of symptoms could be achieved in a second procedure., Results: All procedures were technically successful. Clinical success was achieved in 94% of patients (44 of 47). Three patients had recurrence of pain 3, 5, and 7 months after treatment, respectively (mean observation interval, 22 mo). All recurrences were treated successfully in a second procedure (secondary success rate, 100%). No immediate or delayed complications were observed., Conclusion: CT-guided percutaneous RF ablation is a simple, minimally invasive, safe and highly effective technique for treatment of osteoid osteoma.

Purpose: The purpose of this study was to describe the MR characteristics of periosteal chondroma., Method: MR images of 12 proven cases of periosteal chondroma were analyzed with reference to tumor morphology and size. MR features were correlated with radiographic and pathologic findings., Results: Tumor size ranged from 1 to 7 cm in maximum diameter with a mean value of 2.6 cm. On MR images, a soft tissue mass at the bone surface with pressure erosion of adjacent cortical bone could be identified in all cases. All lesions were bordered by a hypointense rim (100%) and frequently showed a lobulated configuration (75%). Edema of medullary bone or soft tissues was not observed in any of the cases. Signal intensity of cartilaginous tumor tissue was typically hypo-or isointense relative to muscle on T1-weighted (100%) and hyperintense relative to fat on T2-weighted (92%) and T2*-weighted (100%) MR images. Radiographically significant calcifications of the tumor matrix, present in half of the cases, caused focal signal loss on MR images of all pulse sequences. Contrast enhancement was observed predominantly at the periphery of the lesions (100%), which on pathologic examinations typically contained fibrovascular bundles, surrounding the cartilage lobules., Conclusion: Periosteal chondroma appears to have a relatively typical MR appearance, which reflects the histologic composition of the lesion. In addition to radiography, MRI therefore can substantially aid in the preoperative diagnosis of this rare bone lesion.

Osteochondromas can be complicated by mechanical irritation, compression or injury of adjacent structures, fracture, malignant transformation, and postoperative recurrence. Magnetic resonance imaging represents the most valuable imaging modality in symptomatic cases, because it can demonstrate typical features of associated soft tissue pathology, which can be differentiated from malignant transformation. Reactive bursae formation presents as an overlying fluid collection with peripheral contrast enhancement. Dislocation, deformation, and signal alterations of adjacent soft tissue structures can be observed in different impingement syndromes caused by osteochondromas. Magnetic resonance imaging provides excellent demonstration of arterial and venous compromise and represents the method of choice in cases with compression of spinal cord, nerve roots, or peripheral nerves, depicting changes in size, position, and signal intensity of the affected neural structures. Malignant transformation as the most worrisome complication occurs in approximately 1% of solitary and 5-25% of multiple osteochondromas. Magnetic resonance imaging is the most accurate method in measuring cartilage cap thickness, which represents an important criterion for differentiation of osteochondromas and exostotic (low-grade) chondrosarcomas. Cartilage cap thickness exceeding 2 cm in adults and 3 cm in children should raise the suspicion for malignant transformation. Finally, MR imaging can detect postoperative recurrence by depiction of a recurrent mass presenting typical morphological features of a cartilage-forming lesion.

Objective: To evaluate the conventional X-ray and MR imaging features of malignant fibrous histiocytoma (MFH) of bone., Design: MRI examinations and conventional radiographs were reviewed in 39 patients with biopsy-proven MFH. Imaging characteristics were analyzed and the differential diagnoses assessed in a masked fashion by two experienced radiologists., Results: Typical X-ray features included aggressive, destructive tumor growth centrally located in the metaphysis of long bones. Periosteal reactions and expansive growth were rarely seen. On MR images extraosseous tumor spread was frequently noted. On T2-weighted images and contrast-enhanced T1-weighted images most of the tumors displayed an inhomogeneous, nodular signal pattern with peripheral Gd-DTPA enhancement., Conclusions: Although several MR imaging criteria were typical for MFH none of them was specific. X-ray diagnosis of MFH may also prove difficult, with the main differential diagnosis being metastasis in the older and osteosarcoma in the younger population.