Search Results (2,144)

Background/Objectives: Trauma management in emergency departments (EDs) requires rapid and reliable diagnostic tools. The Focused Assessment with Sonography in Trauma (FAST) is a bedside ultrasound examination used for the early detection of free fluid in the intraperitoneal cavity, pericardium, and pleural spaces. Expanding FAST use to trained emergency nurses may support timely bedside evaluation in high-demand settings. However, data on agreement with physicians remains limited. This study aimed to evaluate the inter-rater agreement between a trained emergency nurse and physicians in performing FAST and to explore the diagnostic accuracy of nurse-performed FAST compared with computed tomography (CT). Methods: A prospective pilot observational agreement study was conducted between October and December 2023 in the ED of a general hospital in Cyprus. FAST examinations were independently performed by a nurse trained in FAST and by physicians from the radiology department. Four anatomical areas were assessed: right upper quadrant (RUQ), left upper quadrant (LUQ), subxiphoid-pericardial area (SUPH), and suprapubic area (BLADDER). Findings were recorded independently to promote blinding. Diagnostic performance of nurse-performed FAST was explored in a subset of patients undergoing CT. Results: The sample included 68 trauma patients, of whom 58 underwent FAST by both the nurse and the radiologists and were included in the inter-rater agreement analysis. Fluid was detected in four patients (6.9%) in the RUQ area and in one patient (1.7%) in both the LUQ and SUPH regions, while no positive findings were recorded in the BLADDER area. Agreement in the RUQ area was 98.3% (Cohen’s kappa = 0.85, p < 0.001) while agreement was observed in all cases in the SUPH region (100%, Cohen’s kappa = 1.00, p < 0.001), although this finding was based on a single positive case. High observed agreement was also noted in LUQ (98.3%) and BLADDER regions; however, Cohen’s kappa could not be reliably estimated in these regions due to limited variability and the very small number of positive cases. In a subgroup of patients who underwent CT (n = 23), as well as in an additional Trauma Team subgroup (n = 10), diagnostic accuracy estimates were 100% for sensitivity and specificity; however, these estimates were based on a very small number of positive cases (only two positive cases in each subgroup) and were associated with wide confidence intervals. Conclusions: This pilot study suggests that, under specific training conditions, a trained emergency nurse may achieve a high level of agreement with physician assessments when performing FAST. The findings regarding diagnostic accuracy are preliminary and should be interpreted with caution due to the small sample size and low number of positive cases. Further studies with larger samples and multiple operators are required to confirm these findings and to evaluate their clinical implications. Future research is also needed to determine whether nurse-performed FAST may contribute to improved patient safety and emergency department workflow. Full article

Structured radiology reporting can improve clinical decision support by standardizing clinical findings into hierarchical formats. However, thousands of questions in structured report templates about clinical findings are prohibitively time-consuming, which can limit clinical adoption. Furthermore, early medical VQA datasets primarily focused on free-text and independent question–answer pairs while a recent dataset, Rad-ReStruct, introduced a hierarchical VQA, but the accompanying model still relies heavily on flattened embedding representations and single-path text–image fusion mechanisms that inadequately handle complex hierarchical dependencies in responses. In this paper, we propose DPA-HiVQA (Dual-Path Cross-Attention for Hierarchical VQA), addressing these limitations through two key contributions: (1) multi-scale image embedding representing global semantic embeddings with patch-level spatial features from domain-specific BioViL encoder; (2) dual-path cross-attention mechanism enabling simultaneous holistic semantic understanding and fine-grained spatial reasoning. Evaluated on the Rad-ReStruct benchmark, the model substantially outperforms the established benchmark baseline with an overall F1-score and Level 3 F1-score improvement by 21.2% and 31.9%, respectively. The proposed model demonstrates that dual-path cross-attention architectures can effectively connect holistic semantic understanding and fine-grained spatial detail, paving the way for practical AI-assisted structured reporting systems that reduce radiologist burden while maintaining diagnostic accuracy. Full article

Electrocardiogram (ECG)-gated contrast-enhanced computed tomography angiography (CTA) is the reference method for follow-up of ascending aortic aneurysms but delivers substantially higher radiation doses than ECG-gated non-contrast CT (NCCT). NCCT can be acquired at a lower dose while enabling measurements of the aortic outer diameter. This study aimed to quantify the radiation dose of both techniques and determine whether a significant difference exists in ascending thoracic aorta diameter measurements between NCCT and CTA. Eighty patients who underwent ECG-gated cardiac CT for suspected coronary artery disease were retrospectively analyzed. Three observers measured the ascending aortic diameter at the level of the pulmonary artery in a plane perpendicular to the aorta on both NCCT and CTA images. Inter-rater reliability was assessed using intraclass correlation coefficients, and paired samples t-tests were used to evaluate measurement differences. Dose-length products (DLP) were collected. Median DLP values were 16.1 mGy·cm (interquartile range 11.8–25.1) for NCCT and 190.3 mGy·cm (interquartile range 120.5–298.9) for CTA. NCCT measurements were consistently larger than CTA measurements, with mean differences of 2.1 ± 0.8 mm, 2.6 ± 0.96 mm, and 2.9 ± 1.09 mm for the senior radiologist, junior radiologist, and resident, respectively (all p < 0.001). Inter-observer agreement was excellent (ICC = 0.99, p < 0.001). NCCT delivered an 11.8-fold lower radiation dose than CTA. NCCT may replace CTA for ascending aortic diameter follow-up if measurements are adjusted by approximately 2–3 mm relative to CTA-derived inner-diameter thresholds. Full article

Background: The goal of automated radiology report generation is to help radiologists in their task of creating descriptive reports from chest radiographs. However, the process of creating coherent and contextually accurate reports has been challenging, mainly due to the intricacies of medical language and the need to correlate visual data with textual descriptions. Methods: This study presents LLaMA-XR, a novel framework that integrates Meta LLaMA 3.1 Large Language Model with DenseNet-121-based image embeddings and Quantized Low-Rank Adaptation (QLoRA) fine-tuning. Results: The experiment conducted on the IU X-ray dataset demonstrates that LLaMA-XR outperforms a range of state-of-the-art methods. It achieves an ROUGE-L score of 0.433 and a METEOR score of 0.336, establishing new performance benchmarks in the domain. Conclusions: These results underscore LLaMA-XR’s potential as an effective artificial intelligence system for automated radiology reporting, offering enhanced performance. Full article

Breast mastitis is a common condition that can be found during clinical practice, challenging the clinician, who must reach the correct diagnosis among the many differentials, to properly treat the underlying pathology. In this review, we aim to provide clinicians and radiologists with an overview of the various forms of mastitis, focusing on clinical presentation, etiological subtypes, imaging appearances across modalities (e.g., ultrasound, mammography/tomosynthesis, contrast enhanced techniques, MRI), related complications, and the typical imaging takeaways. Our goal is also to provide tools for the correct differential diagnosis between various forms of mastitis, breast cancer and other inflammatory breast pathologies. A computerized literature search using PubMed and Google Scholar was performed by authors, entering various keywords (e.g., “mastitis”, “breast infections”, “breast abscess”, “breast cancer mimickers”, “lactational mastitis”, “non lactational mastitis”, “mastitis imaging”, “rare forms of mastitis”). Articles published between 2002 and 2025 were taken into consideration. The authors selected various eligible studies, scientific articles and extracted data to cover the whole spectrum of mastitis clinical presentation and underlying pathology. Authors divided the mastitis spectrum into “lactational” and “non-lactational” forms. Between the second group, periductal mastitis, idiopathic granulomatous mastitis, and rarer forms are taken into consideration. Our review has several limitations: it is a narrative and not systematic review and has limited generalizability of rare subtypes because of the case report driven evidence, heterogeneity of selected studies and potential selection bias. It supplies imaging from various clinical cases, which can be useful to familiarize with the pathology spectrum. In conclusion, breast mastitis is a challenge for breast radiologists and clinicians, familiarity with this condition is crucial to make a correct differential diagnosis. Further studies are needed on rarer subtypes. Full article

Background/Objectives: Accurate quantification of alveolar cleft defects for bone grafting remains difficult due to inconsistent anatomical boundaries. This study established an expert consensus on boundary landmarks for alveolar bone graft (ABG) planning and validated the accuracy and reliability of digital volumetric measurement methods. Methods: Three cleft specialists performed repeated simulated graft procedures in seven patient-specific 3D-printed models, first according to the operator’s clinical judgment, and subsequently according to panel-derived consensus boundaries. Two radiologists independently conducted digital volumetric assessments in 3D X-ray imaging using four measurement approaches (axial tracing, interpolated axial tracing, landmark-based mirroring, and mesh-based mirroring), generating 56 independent digital segmentations to be evaluated against the consensus-based physical reference standard. Volumes of the defects were recorded, intra- and inter-rater reliabilities were calculated using the intraclass correlation coefficient (ICC), and differences among methods were analyzed. Results: Operator-defined plans showed significant inter-operator differences (p < 0.001) with poor-to-excellent reliability (intra-rater ICC 0.060–0.967; inter-rater ICC 0.300–0.635). Consensus established standardized boundaries: tilted plane from base of anterior nasal spine to hard palate, cemento-enamel junctions, incisive canal, and alveolar contour. Consensus-based filling showed non-significant inter-rater differences (p = 0.139) and substantially improved reliability (intra-rater ICC 0.904–0.988; inter-rater ICC 0.622–0.861). Among the four digital methods evaluated, axial tracing demonstrated excellent reliability (intra-rater ICC 0.971–0.99; inter-rater ICC 0.965) and high accuracy (mean difference 0.001–0.026 cm³), with no significant difference (p = 0.999) from the physical reference standard. Conclusions: These proposed consensus-based boundary definitions and validated volumetric measurement methods improved the accuracy and reproducibility of personalized alveolar bone graft planning. Full article

Background: Because of the increased awareness of the clinical importance of ILAs on chest CT, this study aims to determine the prevalence of ILAs in an Italian health cohort undergoing CT for a lung cancer screening (LCS) program and quantify ILA using both visual and deep learning-based analyses. Methods: In this observational, retrospective monocentric study, 500 participants (ITALUNG2, n = 100; RISP, n = 400) underwent low-dose CT (CTDI < 2mGy). Two radiologists retrospectively reviewed the images and determined the presence and extent of ILAs, classifying them as fibrotic or non-fibrotic, while a lung texture analysis was performed by using commercially available deep learning–based software. Results: ILAs were present in 34 patients (11 females and 23 males), with a prevalence of 6,8%, with similar rates across both screening cohorts taken individually. Interobserver agreement between radiologists was almost perfect, whereas concordance between visual and automated quantification was substantial. Visual assessment tended to yield higher estimates of ILA extent compared with software-based analysis. Conclusions: These findings confirm that ILAs are relatively common in LCS populations and highlight the importance of their detection and characterization, particularly for fibrotic patterns. Differences between visual and automated approaches underline the need for further refinement and validation of quantitative tools. Full article

Accurate detection of prostate cancer suspicious areas in biparametric MRI (bpMRI) remains challenging because of severe lesion-to-background imbalance, limited lesion contrast, and inter-reader variability in lesion delineation. Unlike prior approaches that collapse inter-reader disagreement into a single consensus label, this study makes three contributions: (1) an adapted nnU-Net framework with prostate-centered preprocessing to reduce voxel-level class imbalance; (2) a class-imbalance-aware composite loss combining Dice, binary cross-entropy, and tailored focal loss to improve sensitivity to small and low-contrast lesions; and (3) a multi-expert learning strategy that preserves reader-specific annotations as separate supervision targets and aggregates predictions at the ensemble level. The method was developed on a single-center dataset of 378 bpMRI studies independently annotated by three board-certified radiologists. Of these, 323 studies were used for model development with patient-level 5-fold cross-validation, and 55 studies were reserved as a fixed independent test set. Compared with our previously published U-Net baseline, the proposed consensus-based nnU-Net improved Average Precision (AP) from 0.69 to 0.75, AUROC from 0.92 to 0.96, and the PI-CAI score from 0.81 to 0.85 on the independent test set. In addition, the multi-expert approach further improved AP to 0.81 versus 0.76 (+6.6%, p < 0.01), AUROC to 0.99 versus 0.95 (+4.2%, p < 0.01), and the PI-CAI score to 0.90 versus 0.86 (+4.7%). These findings demonstrate that explicitly preserving expert disagreement as a training signal, combined with anatomically targeted preprocessing and tailored loss design, substantially improves prostate lesion detection in bpMRI, providing a strong basis for future multi center external validation. Full article

Objective: Existing deep learning (DL) studies on brain metastasis have largely focused on algorithm or reader performance in controlled settings, whereas its role in routine teleradiology quality support remains unestablished. We evaluated the report-level impact of DL assistance on brain metastasis interpretation in a real-world teleradiology workflow using dual-sequence MRI. Materials and Methods: In this retrospective study, 600 patients who underwent contrast-enhanced dual-sequence brain MRI during two consecutive 3-month periods before (pre-DL, n = 286) and after (post-DL, n = 314) DL integration into teleradiology workflow were analyzed. Ten board-certified teleradiologists interpreted all the cases with or without DL-generated overlays. Report-level diagnostic metrics were assessed against a consensus reference standard established by faculty neuroradiologists. Subsequently, exploratory case-level stratified sensitivity analyses were performed for metastasis-positive examinations based on lesion multiplicity and the largest lesion size. Teleradiologists’ perceptions were assessed using a post-interpretation survey. Results: Compared with the pre-DL group, the post-DL group showed higher sensitivity (77.7% vs. 90.8%, p < 0.001), specificity (82.3% vs. 90.8%, p = 0.002), accuracy (80.8% vs. 90.8%, p < 0.001), positive predictive value (68.2% vs. 85.7%, p < 0.001), and negative predictive value (88.3% vs. 94.2%, p = 0.011). False-positive and false-negative rates were lower after DL implementation (11.9% vs. 5.7%, p = 0.009; 7.3% vs. 3.5%, p = 0.045). Sensitivity gains were most pronounced for cases with single metastasis (74.6% vs. 91.2%, p = 0.007) and with the largest lesion ≤ 5 mm (74.3% vs. 92.0%, p = 0.004), whereas sensitivity was similar for multiple metastases and for cases with a largest lesion > 5 mm. Survey responses suggested favorable usability and diagnostic support. Conclusions: In this real-world teleradiology workflow, DL implementation was associated with higher report-level diagnostic metrics and fewer false interpretations. DL assistance may help support quality control for brain metastasis interpretation, particularly in more subtle and diagnostically challenging cases, although radiologist judgment remains essential for subtle or borderline lesions. Full article

Background/Objectives: Slipped Capital Femoral Epiphysis (SCFE) is a common, serious hip disorder in children and adolescents. Two-dimensional (2D) radiography is the gold standard for diagnosis but may not fully capture the deformity’s complexity, and it is vulnerable to positioning errors. Advances in three-dimensional (3D) imaging, such as computed tomography and magnetic resonance imaging (MRI), enable more accurate assessments. This study aimed to (1) assess the inter-rater reliability of 2D radiographic and 3D MRI measurements, and (2) evaluate the correlations and agreements between these outcomes. Methods: Patients were randomly selected from a cohort of patients aged under 16 years old and diagnosed with SCFE between January 2000 and December 2024. Southwick angles and posterior epiphyseal slip angles on 2D radiographs were independently measured by two orthopaedic surgeons. Posterior epiphyseal slip angles on 3D MRI were independently measured by two orthopaedic surgeons and two paediatric radiologists. Relationships between the three outcomes were evaluated using the Pearson correlation coefficient (r). Inter-rater reliability and agreements between the three outcomes were evaluated using the intraclass correlation coefficient (ICC) and the standard error measurement (SEM). Results: A total of 35 patients (35 hips) were recruited, with a mean age of 11.8 (1.2) years old and 19/35 (54%) females. Radiographic outcomes were moderately correlated (r < 0.75, p < 0.01) with MRI posterior epiphyseal slip angles. MRI posterior epiphyseal slip angles were systematically greater (16° on average) than both radiographic outcomes, regardless of whether contralateral correction was applied. The inter-rater reliability of radiographic outcomes was excellent (ICC > 0.85, SEM > 5.0°) and almost perfect (ICC > 0.95, SEM = 2.5°) for the MRI posterior epiphyseal slip angles measured by the paediatric radiologists. Conclusions: Findings suggest that while both diagnostic methods are reliable, radiographic measurements systematically underestimate epiphyseal slip severity by approximately 16° compared to MRI. This discrepancy could impact the accuracy of disease staging, leading to potential misclassifications. This highlights the need for a more standardised approach to evaluating SCFE, especially regarding the type of imaging used for angle measurement. Full article

Background: Magnetic resonance imaging (MRI) is the reference standard for evaluating temporomandibular joint (TMJ) disorders, particularly for assessing disc position, joint effusion, and degenerative changes. With increasing imaging demands and advances in deep learning, artificial intelligence (AI) has emerged as a potential adjunct to expert interpretation. This systematic review aimed to compare the diagnostic performance of AI-based models with that of human experts in TMJ MRI analysis. Methods: This review was conducted in accordance with the PRISMA 2020 guidelines and prospectively registered in PROSPERO (CRD420251174127). A systematic search of PubMed/MEDLINE, ScienceDirect, Wiley Online Library, and Springer Nature Link was performed for studies published between 2020 and 2026. Eligible studies included human participants undergoing TMJ MRI and evaluated AI, machine learning, or deep learning models against human expert interpretation. Extracted outcomes included sensitivity, specificity, accuracy, area under the receiver operating characteristic curve (AUC), and agreement metrics. Risk of bias was assessed using QUADAS-2. Due to substantial heterogeneity, a narrative synthesis was conducted. Results: Five retrospective diagnostic accuracy studies were included, comprising sample sizes ranging from 118 to 1474 patients. Target conditions included anterior disc displacement, joint effusion, osteoarthritis, and disc perforation. AI models demonstrated strong discriminative performance, with reported AUC values ranging from 0.79 to 0.98. In direct comparisons, AI achieved diagnostic accuracy comparable to experienced radiologists. AI systems frequently demonstrated higher specificity and similar overall accuracy, whereas human experts often showed higher sensitivity. In osteoarthritis assessment, AI performance approached expert level and exceeded that of less experienced readers. All studies were retrospective and predominantly single-center, with heterogeneous reference standards and limited external validation. Conclusions: AI achieves diagnostic performance comparable to experienced clinicians in TMJ MRI interpretation and shows promise as a decision-support tool. Nevertheless, it should be regarded as complementary to, rather than a replacement for, expert radiological assessment pending further rigorous validation. Full article

Background/Objectives: Intraoral radiographs acquired using photostimulable phosphor (PSP) plates are inherently susceptible to a wide spectrum of artifacts that can compromise diagnostic reliability and lead to unnecessary repeat exposures. Although structured taxonomies describing these artifacts have been proposed, automated methods capable of detecting and localizing multiple artifact types at the pixel level remain limited, particularly under realistic multi-class conditions. In this study, we address the problem of fine-grained, multi-class PSP artifact segmentation by systematically evaluating a deep learning-based framework and establishing a realistic baseline for this inherently challenging task. Methods: A retrospective, multi-center dataset comprising 1497 intraoral PSP radiographs (bitewing and periapical) collected from three institutions was analyzed. Pixel-level annotations were generated by expert oral and maxillofacial radiologists according to a standardized taxonomy consisting of four major artifact groups and 29 artifact classes, together with a background class. A 2D nnU-Net v2 architecture was employed as a baseline segmentation model. Model development was performed using 5-fold cross-validation, and performance was evaluated on an independent test set using Dice coefficient, Intersection over Union (IoU), Precision, and Recall. Results: Across all classes, the model achieved a mean Dice score of 0.0894 ± 0.0084 in cross-validation and 0.0952 on the independent test set, reflecting the intrinsic complexity of the task. Class-wise analysis revealed substantial variability, with higher performance in larger and visually distinctive artifacts, whereas small-scale, low-contrast, and underrepresented classes exhibited markedly reduced performance. Notably, several artifact categories were absent from the training data, resulting in a zero-shot scenario that directly constrained model generalization. Furthermore, segmentation performance demonstrated a strong dependency on class frequency, measured in terms of pixel distribution, underscoring the impact of severe class imbalance. Group-based evaluation showed relatively higher performance for pre-exposure and exposure-related artifacts compared to post-exposure and scanner-related categories. Conclusions: These findings demonstrate that large-scale, multi-class pixel-level segmentation of PSP artifacts represents a fundamentally challenging problem shaped by the combined effects of class imbalance, small object size, heterogeneous artifact morphology, and incomplete training representation. While the proposed framework confirms the feasibility of automated artifact localization, its current performance suggests greater immediate value as a quality control or screening support tool rather than a fully autonomous diagnostic system. By providing a comprehensive baseline and systematic analysis, this study establishes a benchmark for future research and highlights the critical need for imbalance-aware learning strategies, hierarchical modeling, and data-centric approaches to advance this field. Full article

Background/Objectives: Stroke generates 157 million disability-adjusted life-years (DALYs) annually, making it the leading neurological cause of global disease burden. Artificial intelligence (AI) and machine learning (ML) have emerged as transformative technologies across the stroke care continuum. This narrative review maps the trajectory of AI in stroke medicine over the decade from 2015 to 2025. Methods: We conducted a narrative review with a structured, pre-specified search strategy across eight pre-specified thematic clusters using PubMed/MEDLINE (January 2015–December 2025), identifying 8549 records and including 1335 studies after screening. Inclusion criteria encompassed primary research articles, systematic reviews, meta-analyses, and RCTs reporting quantitative performance metrics or clinical outcome data for AI/ML in stroke. Results: Stroke imaging AI is the most commercially mature domain, with over 30 FDA-cleared tools. Automated ASPECTS scoring reduced radiologist reading time by 74.8% (AUC 84.97%; 95% CI: 83.1–86.8%). The only triage AI RCT demonstrated an 11.2 min reduction in door-to-groin time without significant improvement in 90-day functional independence (OR 1.3, 95% CI 0.42–4.0). Brain–computer interface rehabilitation showed significant upper limb recovery in a 17-center RCT (FMA-UE mean difference +3.35 points, 95% CI 1.05–5.65; p = 0.0045). AF detection AI is FDA-cleared and RCT-validated. LLMs and federated learning are pre-regulatory but growing exponentially. Conclusions: AI in stroke has achieved diagnostic maturity but therapeutic immaturity. Bridging algorithmic performance to patient outcomes, addressing equity gaps, and building the economic evidence base for scalable deployment are the defining challenges of the next decade. Full article

Background/Objectives: Coronary computed tomography angiography (CCTA) is a modern method for assessing the total burden of atherosclerotic lesions. The perivascular fat attenuation index (PFAI) is a reliable predictor of major adverse cardiovascular events (MACE). Radiomics extracts substantially more information from images than visual assessment by radiologists. However, the relationships between quantitative parameters of coronary atherosclerosis, the PFAI, and radiomic features of pericoronary adipose tissue (PCAT) in patients with coronary artery disease (CAD) remain unclear. The study aimed to evaluate the associations between PCAT characteristics, including radiomic features, and quantitative parameters of coronary atherosclerosis in stable CAD patients. Methods: The study included 79 patients with stable CAD who underwent CCTA. The patients were divided into two groups: nonobstructive CAD (NOCAD, stenosis < 50%; n = 61) and obstructive CAD (OCAD, stenosis ≥ 50%; n = 18). The CCTA data were analyzed to quantify coronary atherosclerosis parameters (plaque volume and burden), the PFAI, PCAT volume, and radiomic features of PCAT in the proximal segments of major coronary arteries. Results: The study included 79 patients: NOCAD group = 61 patients (age 57.00 (50.00–65.00) years) and OCAD group = 18 patients (age 60.5 (55.75–65.75) years). The OCAD patients exhibited higher plaque volume and burden across all components. No significant between-group differences were observed in PFAI or PCAT volume for any vessel. However, 50% (46/92) of PCAT radiomic features in the proximal right coronary artery (RCA) differed significantly between groups, 42 of which were textural. The PFAI correlated most strongly with soft tissue (ST) plaque volume (ρ = −0.22), and burden (ρ = −0.21) of the soft tissue component of plaques (p < 0.001). The PCAT volume significantly correlated (p < 0.001) with plaque volume (ρ = 0.30) and with individual components—soft tissue (ρ = 0.30), fibrous–fatty (ρ = 0.27), fibrous (ρ = 0.30), calcified (ρ = 0.22), and non-calcified (ρ = 0.30)—as well as with the burden of the soft tissue component (ρ = 0.26). Conclusions: The radiomic features of RCA PCAT differed significantly between the NOCAD and OCAD groups. Quantitative coronary atherosclerosis parameters showed significant associations with the PCAT radiomic features in CAD patients, potentially serving as independent predictors of the MACE risk. In contrast, the PFAI values did not differ between groups and neither PFAI nor PCAT volume associated with atherosclerosis parameters. Full article

Myositis ossificans (MO) is a benign, self-limiting heterotopic ossification process that typically develops within soft tissues following trauma, although non-traumatic forms have also been described. Despite its benign nature, MO frequently represents a diagnostic challenge, particularly in its early stages when imaging findings may mimic aggressive soft-tissue tumors, leading to unnecessary biopsies or surgical interventions. This narrative review provides an updated overview of the classification, pathophysiology, and imaging features of myositis ossificans, with a specific focus on the time-dependent evolution of radiologic appearances across different imaging modalities. Radiologic findings are discussed according to disease stage, highlighting key diagnostic clues such as the zonal phenomenon and peripheral maturation pattern. In addition, the main entities included in the differential diagnosis are reviewed, with particular emphasis on imaging features that help distinguish myositis ossificans from soft-tissue sarcomas and other calcified or ossified lesions. Finally, current management strategies and the role of imaging in patient follow-up are summarized. A thorough understanding of the evolving imaging spectrum of myositis ossificans is essential for radiologists and clinicians to achieve an accurate diagnosis, guide appropriate management, and avoid overtreatment. Full article