With the rapid growth of the use of computed tomography, advances in artificial intelligence enable opportunistic screening, the systematic extraction of clinically meaningful biomarkers from imaging scans performed for other indications. Modeling studies demonstrate that opportunistic screening can be highly cost effective by enabling early intervention and preventing complications such as osteoporotic fractures. Musculoskeletal radiologists are uniquely positioned to contribute to this paradigm shift because routine examinations frequently include vertebrae, skeletal muscle, adipose tissue, and vasculature, all structures that provide quantitative data on bone mineral density, sarcopenia, adiposity, and cardiovascular risk. However, widespread implementation faces challenges, such as the need for prospective outcomes data, normative reference standards, workflow integration, and clear pathways for clinical follow-up. This review examines the rationale, technical foundations, key applications, and challenges for opportunistic screening in musculoskeletal radiology.
Shear wave elastography (SWE) is an increasingly investigated ultrasound-based technique in musculoskeletal imaging that provides quantitative information on tissue stiffness and biomechanical properties. This narrative review aims to summarize the basic principles, technical considerations, current clinical applications, limitations, and future perspectives of SWE in musculoskeletal imaging. Unlike conventional grayscale and Doppler ultrasonography, which mainly assess morphology and vascularity, SWE may provide additional functional information in major musculoskeletal tissues, including tendons and ligaments, skeletal muscles, peripheral nerves, fibrocartilaginous structures, plantar fascia, and selected soft tissue lesions. Current evidence suggests potential roles for SWE in detecting early biomechanical alterations, assessing disease severity, differentiating symptomatic from asymptomatic tissues, and monitoring response to treatment or rehabilitation. However, musculoskeletal tissues are anisotropic, viscoelastic, and position-dependent; as a result, SWE measurements are influenced by acquisition-related factors, tissue biomechanics, positioning and loading conditions, region of interest (ROI) placement, tissue depth, and device-related variability. For this reason, SWE findings should not be interpreted as standalone diagnostic criteria but should be considered together with clinical findings, conventional ultrasonography, MRI, electrophysiology, histopathology, and patient-centered outcomes when appropriate. This review highlights the need for tissue-specific measurement protocols, standardized reporting, normative reference data, inter-vendor harmonization, and longitudinal validation against clinically meaningful outcomes before SWE can be more reliably integrated into routine musculoskeletal imaging and rehabilitation practice.
Depression and reduced muscle mass are common in patients with lung cancer, and both are associated with adverse clinical outcomes. However, data examining their interrelationship remain limited. This cross-sectional study included 100 patients with newly diagnosed lung cancer undergoing systemic therapy. Depressive symptoms were assessed using the Beck Depression Inventory (BDI). Skeletal muscle mass was evaluated using computed tomography (CT)-derived skeletal muscle index (SMI) at the L3 vertebral level. Clinical and demographic variables were analyzed for associations with depression. Clinically relevant depressive symptoms were identified in 71% of patients, and CT-defined low skeletal muscle mass (LSMM) was present in 67% of patients. Scores on BDI were significantly and inversely correlated with total muscle area (TMA; r = -0.418, p < 0.001) and skeletal muscle index (r = -0.358, p < 0.001). In age-adjusted multivariate analysis, low SMI (OR 3.05, 95% CI 1.15-8.09) and higher Charlson Comorbidity Index (CCI) values (OR 3.27, 95% CI 1.25-8.68) were independently associated with depression. Depressive symptoms are highly prevalent in lung cancer and are independently associated with CT-defined LSMM and a higher comorbidity burden. Routine assessment of SMI on staging CT scans may help identify patients at increased risk and support early multidisciplinary interventions.
99mTc-pyrophosphate (PYP) scintigraphy is widely used for the noninvasive evaluation of transthyretin cardiac amyloidosis. Although interpretation primarily focuses on myocardial uptake, confirmation of appropriate systemic radiotracer biodistribution is essential. We report a case in which an examination presumed to be 99mTc-PYP scintigraphy demonstrated free 99mTc-pertechnetate-like biodistribution. A 75-year-old woman with chronic kidney disease and conduction disturbance underwent 99mTc-PYP scintigraphy for suspected cardiac amyloidosis. The initial study, recorded as the administration of 740 MBq 99mTc-PYP, was imaged 3 h after injection. Planar imaging showed mild apparent activity over the cardiac region; however, SPECT/CT demonstrated no definite myocardial uptake. Instead, intense uptake was observed in the stomach and thyroid gland, with complete absence of skeletal activity. This distribution was inconsistent with correctly administered 99mTc-PYP and suggested free 99mTc-pertechnetate biodistribution, likely due to radiopharmaceutical preparation or administration error. A repeat 99mTc-PYP scan 1.5 months later showed expected skeletal uptake without gastric or thyroid activity and again demonstrated no myocardial uptake. The study was interpreted as negative for cardiac amyloidosis. Gastric and thyroid uptake with absent skeletal activity on presumed 99mTc-PYP scintigraphy should be considered nondiagnostic rather than negative.
Artificial intelligence (AI) is increasingly reshaping musculoskeletal (MSK) imaging across the entire imaging pathway. This narrative review summarizes current AI applications in MSK radiology across four domains: acquisition and reconstruction, detection and triage, characterization and quantification, and prognosis and decision support. AI-based reconstruction has enabled faster MRI acquisitions, improved denoising and artifact reduction, and supported low-dose CT imaging while preserving diagnostic quality. Fracture detection and triage currently represent the most mature clinical applications, particularly in emergency settings. AI is also promoting a shift from qualitative interpretation to quantitative imaging phenotyping through automated assessment of body composition, cartilage, bone density, degenerative spine disease, skeletal maturity, and lesion heterogeneity. Emerging applications in prognostic modeling, implant evaluation, and multimodal risk stratification remain promising but less mature. Broader clinical implementation is still limited by restricted interpretability, dataset bias, insufficient prospective validation, regulatory complexity, and unresolved medico-legal issues. Overall, AI should be viewed as a tool to augment, not replace, radiological expertise.
Though subgroup performance reporting helps ensure the safety of artificial intelligence (AI) products, the extent of this reporting remains unclear. This scoping review identifies studies validating commercially available AI-based products and reports the trends in performance reporting across sex, age, and race/ethnicity demographic subgroups. Peer-reviewed validation studies of commercially available products published after 2010 were collected from the Health AI Register and PubMed on 29 November 2024. Study trends in the reporting of sex, age, and race/ethnicity were mapped with regression analysis. We apply the Wilson confidence interval equation to estimate which tuberculosis detection studies are underpowered for subgroup meta-analysis. Three hundred ninety-two of 545 studies validating 252 products reported subgroup demographic data for any of the three groups. Only 77 of these presented subgroup performance results. Skeletal (20/88) and lung (30/139) studies, and those utilizing chest (24/79) or bone (19/63) radiographs, most often presented subgroup performance data. We found no evidence that more recent studies (OR: 1.039 [95% CI: 0.959-1.127]) or company sponsorship (OR: 1.010 [95% CI: 0.492-1.920]) led to increased subgroup reporting. We show that 14/21 tuberculosis datasets may be underpowered for post-hoc subgroup meta-analysis. This scoping review quantifies how fragmented the commercial validation landscape is, showing that reporting for both the demographics and per-subgroup performance is inadequate for estimating subgroup bias. This systemic problem requires effort from all stakeholders, from researchers to regulatory agencies, encouraging thorough reporting and commercial product validation to support physician and patient trust in medical AI products. Question The number of studies validating the performance of each commercially available radiology AI product for minority subgroup bias is unclear. Findings The currently available commercial AI validation studies often neglect to describe demographic subgroup data, and fewer provide performance results per subgroup, prohibiting algorithmic bias meta-analysis. Clinical relevance Physician and patient trust in the medical AI already used clinically must be built on peer-reviewed literature and meta-analysis. The current literature is insufficient for determining the safety and performance of these products for demographic minorities.
Extraskeletal osteosarcoma (ESOS) is a rare high-grade malignant mesenchymal tumor that produces osteoid in soft tissue without skeletal involvement. Subcutaneous ESOS is exceptionally uncommon and may mimic benign soft tissue lesions. A 45-year-old Samoan man presented with a one-year history of a progressively enlarging, painless left gluteal mass. Contrast-enhanced CT showed a well-circumscribed subcutaneous lesion measuring 9.1 × 10.4 × 8.6 cm, with mild to moderate delayed enhancement and no osseous involvement. The tumor was surgically excised. Histopathology showed a spindle to pleomorphic high-grade sarcoma with osteoid produced by atypical tumor cells. Immunohistochemistry showed SATB2 positivity and negative staining for S100, SOX10, pan-cytokeratin, and ERG. The tumor involved the deep resection margin, consistent with R1 resection; the closest margin distance was not specified. The patient declined adjuvant therapy and was subsequently lost to follow-up. Subcutaneous ESOS is rare and diagnostically challenging. This case highlights the importance of imaging, histopathology, immunohistochemistry, and multidisciplinary management, particularly in resource-limited settings.
To develop a fully automated 2D nnU-Net pipeline for multi-class skeletal muscle segmentation (psoas, paraspinal, and abdominal wall) at the third lumbar (L3) vertebral level, and to quantitatively evaluate its diagnostic performance and reliability compared to manual segmentation. A 2D nnU-Net was trained on 164 axial L3 CT slices from the multi-institutional AMOS22 dataset, spanning diverse abdominal pathologies and multivendor imaging. To assess generalizability under severe anatomical distortion, independent external validation was performed in 50 consecutive patients with advanced liver disease from a single institution (January-December 2025; mean age, 63 ± 15 years; 32 women, 18 men), of whom 88% had moderate-to-severe ascites. Model stability was examined by comparing a five-fold ensemble with the best-performing single-fold model. Intra-observer reliability of the manual reference standard was evaluated in a random subset of 30 cases. Inter-observer agreement was additionally assessed using an independent second reader. Performance metrics included the Dice Similarity Coefficient (DSC), Pearson correlation coefficient (r), and Bland-Altman analysis for cross-sectional areas and mean attenuation. The inference workflow was deployed via a custom Streamlit-based graphical user interface (GUI). In this anatomically complex external validation cohort, the 5-fold ensemble 2D nnU-Net achieved an overall mean DSC of 0.937 ± 0.043 (95% CI, 0.925-0.950), with 80% of cases achieving a mean DSC ≥ 0.90. While the mean DSC was statistically comparable to the best single-fold model (0.937, [95% CI, 0.921-0.952], p = 0.736), the ensemble strategy increased the minimum observed DSC (worst-case performance) from 0.720 to 0.822. Class-specific external validation performance for the 5-fold ensemble was highest for the paraspinal muscles (DSC: 0.960; 95% CI, 0.952-0.967), followed by the psoas muscles (DSC: 0.941; 95% CI, 0.927-0.956), and lowest for the anatomically complex abdominal wall muscles (DSC: 0.911; 95% CI, 0.893-0.929). Comparison between the ensemble model and manual segmentation yielded a Pearson correlation of r = 0.955 (p < 0.001) for total skeletal muscle area, with a mean bias of +7.17 cm2. Intra- and inter-observer agreements for the manual reference standard demonstrated correlation coefficients of r = 0.995 and 0.090 for total areas, respectively. The automated pipeline required 3-5 s per case for inference and quantitative reporting, compared to 3-5 min for manual segmentation. In patients with advanced liver disease and substantial anatomical distortion from ascites, an ensemble-based 2D nnU-Net provides high quantitative agreement with manual L3 skeletal muscle segmentation, while mitigating lower-bound (worst-case) errors relative to single-fold models. Integration with a dedicated GUI enables substantial time savings and supports scalable quantitative body composition measurement.
HACD1-related congenital myopathy is an ultra-rare congenital myopathy recently described in families from the Middle East or Asia. Clinical phenotype has been described, but little is known about the pattern of muscle MRI abnormalities. We describe four Brazilian patients from unrelated families including not only clinico-pathological, but also muscle MRI findings. All patients shared the known pattern of early onset of motor deficits combined with respiratory distress, later followed by remarkable improvement. Muscle biopsy revealed congenital fiber type disproportion. The same pathogenic biallelic HACD1 variant was found in all cases (c.373_375+2delGAGGT). Three out of the 4 patients underwent muscle MRI, which revealed symmetrical fatty infiltration predominantly affecting the pelvic girdle and the posterior compartment of lower limbs; anterior compartment of thighs and legs was essentially preserved. The current description expands the geographical landscape of the disease and refines its phenotypical characterization by presenting the pattern of muscle MRI abnormalities.
Analysis of dynamic phosphorus magnetic resonance spectroscopy (31P MRS) data is often hindered by variability in data quality. A quality control (QC) pipeline developed by Naëgel (2023) introduced six key parameters to ensure reliable 31P MRS results in large clinical datasets. This study tested the transferability of this QC scoring (QCS_REF) to two different research sites equipped with 3T and 7T MR systems and different ergometers. Twelve groups with the focus on frail and elderly subjects and patients with neurodegenerative diseases were included. The application of QCS_REF limits led to the improvement of the statistical power in some patient groups, but to the exclusion of substantial data for all our groups and experimental settings at both 3T and 7T. Only 28% of all recovery and exercise period data at 3T and 21% at 7T passed QCS_REF inclusion criteria. Therefore, two new sets of quality control criteria, QCS1 and QCS2, were proposed, reflecting achieved SNR of individual MR signalsand the patient phenotype included. We showed that the transferability of the QCS_REF did not depend on the magnetic field, the coil, or localization scheme. The new QCS did not significantly influence the mean recovery and exercise time constants of each group compared to QCS_REF. We verified that six proposed key parameters were adequate for an objective assessment of the quality of dynamic 31P MRS measurement at 3T as well as 7T. However, the patient group characteristics and experimental set-up significantly affect the ability to meet dynamic 31P MRS quality control thresholds, supporting the use of flexible QC criteria for robust data acquisition across diverse clinical populations.
Fibrous dysplasia is one of the most common skeletal lesions. The wide spectrum of clinical manifestations ranges from asymptomatic conditions (typical of monostotic forms) to severe skeletal diseases with deformity and fractures for polyostotic fibrous dysplasia. The classical radiological features include: an osteolytic geographic pattern, ground-glass bone matrix, cortical thinning/cortical scalloping, bone deformities and enlargement, concavity of margins (evaluated with MRI), and cystic areas (MRI). All the bones can be affected, and the proximal femur is the most common one (about 30% of cases). Nonetheless, the disease can also affect cranio-facial bones, leading to compression of neural structures, as well as deformation and enlargement of facial bones, leading to the so-called "leontiasis ossea" or "facies leonine". The polyostotic forms of fibrous dysplasia can be associated with multiple soft-tissue myomas (Mazabraud syndrome) or several endocrine diseases (McCune-Albright syndrome). In every diagnostic step of the disease, as well as in different fibrous dysplasia forms, imaging plays a key role. Indeed, radiology is fundamental to assess the suspicion of fibrous dysplasia in classical monostotic forms, representing the sole diagnostic tool needed in many cases. Imaging is also fundamental to staging and following up on more severe polyostotic forms, as well as for detecting complications. In this comprehensive updated review article, we examine every aspect of the disease, with a main focus on imaging presentation. The indications for biopsy are discussed as well. Most importantly, the article details the potential risk of malignant transformation (osteosarcoma, fibrosarcoma, chondrosarcoma, and other rarer sarcomas, all accounting for <1% of cases) underlying the radiological patterns of these conditions. The occurrence of aneurysmal bone cyst-like changes on fibrous dysplasia is also analyzed in the article. This review article aims to be a comprehensive guide for radiologists and clinicians involved in the care of patients affected by various forms of fibrous dysplasia, and a starting point for future research. Many classical and atypical cases are collected as an iconographic comprehensive representation.
To assess nomenclature variability and develop recommendations on standardized MRI reporting of meniscal findings. The Society of Skeletal Radiology identified standardized MRI reporting of knee menisci as an important topic for study and invited all members to serve on a panel to provide consensus recommendations. The Society empaneled 12 musculoskeletal radiologists and 2 orthopaedic surgeons. The panel reviewed published literature (PubMed, Scopus, and Embase) using predetermined criteria for inclusion (peer-reviewed, English-language, human studies) and exclusion (case reports, conference abstracts, book chapters, expert opinions, and commentaries). Literature analysis focused on nomenclature relevant to MRI reporting in two general domains: (I) meniscal anatomy and anatomic variants and (II) meniscal tears and associated findings. Substantial nomenclature variability was identified across both domains. For anatomy and variants, inconsistencies involved root zone definitions, vascular zone descriptors, perimeniscal stabilizer terminology (including eponyms), and diagnostic criteria for the discoid meniscus. For tears and associated findings, variability involved numerical grading systems, the terms "complete" and "incomplete," extrusion measurement methodology, and interpretive labels (e.g., "traumatic," "degenerative," "repairable," "stable"). Ten consensus recommendations for standardized MRI reporting were developed. Descriptive anatomic language-specifying what is seen and where-can be more reproducible and clinically actionable than reporting variable numerical grades and evolving classifications. Adoption of our ten pragmatic recommendations has the potential to reduce miscommunication, improve inter-institutional consistency, and support clinical research.
Esophagectomy and pancreatectomy are invasive oncological surgeries with elevated mortality rates. Preoperative muscle mass deficit and myosteatosis, identifiable on L3 CT-scan, could be associated with poor postoperative outcomes in cancer patients. We aimed to determine their impact on short-term complications following esophageal or pancreatic cancer resection. We conducted a retrospective cohort study in two hospital centers from January 2018 to February 2023. Adult patients undergoing esophagectomy or pancreatectomy for cancer with a preoperative CT scan at L3 level were included. Muscle skeletal mass and quality were assessed using previously published thresholds. Poor postoperative short-term outcomes were defined as the occurrence of sepsis, septic shock, or death within 90 days postoperatively. Of 216 patients, 165 patients were eligible for muscle mass analysis and 143 for muscle quality assessment. Poor short outcome occurred in 55 patients (33.3%). Surprisingly, skeletal muscle depletion was inversely associated with poor outcomes in the multivariate logistic regression model (OR 0.38, 95% CI [0.15-0.97], p = 0.04). Myosteatosis was associated with a significantly 6-fold increased risk of poor short-term outcomes in univariate analysis (OR 6.04, 95% CI [2.35-15.55], p < 0.001), with a persistent trend in the multivariate logistic regression model (OR 3.51, 95% CI [0.99-12.39], p = 0.05). In the ICU subgroup, patients with preoperative skeletal muscle depletion and myosteatosis had higher 28-day mortality than those with preserved muscle mass and quality. Myosteatosis, rather than muscle mass deficit, showed a strong trend toward association with adverse postoperative short-term outcomes following esophagectomy and pancreatectomy. Defining population-specific thresholds for CT scan muscle assessment are necessary to improve the use of L3 scans for short-term outcome risk evaluation in oncologic surgery.
This review examines the role of marrow adipose tissue fatty acid composition in skeletal homeostasis, focusing on osteoporosis. To avoid conceptual confusion, we define evidence tiers: human studies assessing endogenous marrow lipid profiles; mechanistic studies applying exogenous fatty acids in vitro and in vivo; indirect MRI-based surrogates of marrow fat quantity; and direct ex vivo measurements of fatty acid composition via GC-MS, LC-MS, and lipidomics. We also clarify the distinction between marrow fat quantity and fatty acid quality. Human data reveal disease-, age-, and site-related alterations in marrow lipid saturation and unsaturation; however, findings vary by skeletal site, marrow compartment, fracture status, analytical platform, and study population. Experimental evidence demonstrates that saturated fatty acids (e.g., palmitic acid) induce lipotoxicity and osteoblast dysfunction, whereas unsaturated fatty acids (e.g., oleic acid and n-3 polyunsaturated fatty acids) exert protective effects via modulation of mesenchymal stem cell differentiation, osteoclastogenesis, ferroptosis, autophagy, and mitochondrial metabolism. Collectively, current evidence supports an association between marrow fatty acid biology and osteoporotic bone loss. Causal, diagnostic, and therapeutic implications remain preliminary. This review's main contribution is a fatty-acid-centered framework that integrates evidence tiers, molecular categories, and skeletal-site heterogeneity, guiding future research.
Osteoporosis (OP) is a chronic systemic skeletal disorder that predominantly affects the elderly. It is characterized by an imbalance in bone homeostasis, reduced bone mass, microarchitectural deterioration of bone tissue, and increased bone fragility, ultimately leading to a higher risk of fractures and related complications. With the progression of global population aging, the prevalence of OP continues to rise, underscoring the importance of early diagnosis and timely intervention. However, the diagnosis and management of OP-particularly its early detection-remain limited by material constraints such as diagnostic equipment and by subjective factors including clinician experience, which hinder widespread screening. In recent years, artificial intelligence (AI) has emerged as a transformative technology with advantages of efficiency, objectivity, and scalability, and has been increasingly integrated into various medical domains. For example, AI-assisted musculoskeletal measurements on leg and foot radiographs can reduce the measurement time from 166 seconds to 40 seconds, resulting in an overall efficiency improvement of approximately 70%. Applying AI to the diagnosis and treatment of OP can reduce human error, save labor costs, and improve diagnostic accuracy and clinical efficiency. Numerous studies have investigated AI-based approaches in OP-related research and clinical practice. Despite these promising developments, several important limitations should be acknowledged. Considerable heterogeneity exists among published studies regarding patient populations, AI algorithms, and evaluation metrics. Besides, consistent external validation remains insufficient in many studies. Challenges related to data imbalance and potential selection bias further highlight the need for standardized reporting frameworks and multicenter collaborative research to promote safe clinical adoption of AI technologies in osteoporosis management. This review summarizes current AI applications in OP diagnosis, risk prediction and therapy. We highlight key methodological limitations and emerging trends, aiming to guide future research and facilitate safe clinical implementation of AI in OP management.
Hyperbaric oxygen (HBO) therapy increases blood oxygen levels by exposing patients to 100% oxygen under elevated pressure and is used clinically for conditions such as decompression sickness, carbon monoxide poisoning, chronic wounds, and impaired tissue healing. However, real-time precise and quantitative measurement of tissue oxygen tension (pO2) under hyperbaric conditions remains technically very challenging. We developed a specialized small-animal hyperbaric chamber compatible with in vivo electron paramagnetic resonance (EPR) oximetry using customized OxyTrack oxygen detectors. Up to two detectors were implanted simultaneously to measure pO2 in skeletal muscle and tumor tissue in mice. Animals were exposed to air (21% O2) and 100% O2 at both normobaric pressure and 2 atmospheres absolute (ATA) in a stepwise protocol. The chamber operated safely and stably up to 2 ATA, enabling continuous real-time pO2 monitoring without adverse events and issues with the hardware for up to 90 min. In B16-F10 melanoma, tumor pO2 increased from ∼9 mmHg at baseline to ∼26 mmHg at 100% O2 (1 ATA) and ∼50 mmHg at 2 ATA, with partial retention after decompression. In contrast, SCC7 tumors showed minimal responsiveness (∼9-10 mmHg), whereas skeletal muscle demonstrated marked pressure-dependent increases (up to ∼50 mmHg). The small-animal HBO chamber enables real-time, tissue-specific assessment of oxygen dynamics under clinically relevant hyperbaric conditions. The approach could facilitate optimization of HBO protocols and support translational investigation of oxygen-modulated therapies, including radiation and drug responses in solid tumors.
Sarcopenia is a known negative prognostic factor in oncology and is frequently observed in patients with pancreatic ductal adenocarcinoma (PDAC). Computed tomography (CT) enables longitudinal muscle assessment and may provide additional prognostic information. This study aims to assess their association with prognosis in pancreatic cancer and explore the diagnostic and prognostic value of CT radiomics. A retrospective single-center study included 62 patients with primary PDAC who underwent at least three abdominal CT scans: baseline (t0), 3 months (t1), 6 months (t2), and, in 35 patients, 12 months (t3). CT-based sarcopenia was assessed using the psoas muscle index (PMI) based on reference cutoffs and cohort-specific sex-specific quartiles. Skeletal muscles at the L3 level were semi-automatically segmented. Radiomic features of the psoas were extracted and analyzed using k-nearest neighbor, decision tree, and random forest models. Prognostic relevance was evaluated using logistic regression and feature selection via least absolute shrinkage and selection operator (LASSO) regression. Tumor progression was assessed radiologically according to RECIST 1.1 criteria. CT-based sarcopenia prevalence was 45.3% using reference-based PMI thresholds. PMI declined significantly from baseline to t1 and remained stable thereafter, with women exhibiting consistently lower values. Outcome analysis showed a higher proportion of disease progression at t1 in sarcopenic patients using reference cutoffs, whereas cohort-specific quartiles demonstrated no consistent differences. Random forest models predicted sarcopenia with up to 0.73 accuracy and receiver operating characteristics area under the curve (ROC-AUC) of 0.81. LASSO regression identified the psoas short axis and cross-sectional area as the most informative features. Logistic regression using baseline radiomic features predicted disease progression status at 12 months with 0.85 accuracy, weighted F1 0.841, and AUC 0.823. Interobserver agreement for psoas measurements was high (r = 0.86). Longitudinal CT-based assessment of PMI demonstrates progressive sarcopenia within the studied PDAC cohort, with sex-specific declines. Radiomic analysis of skeletal muscle provides complementary information and predictive insights, highlighting their potential to enhance the characterization of muscle status and its association with disease course in patients able to undergo repeated imaging.
Transarterial chemoembolization (TACE) is a standard treatment for patients with unresectable hepatocellular carcinoma (HCC), yet existing models provide limited individualized risk stratification. Automated CT-derived body composition analysis has emerged as an objective marker of patient physiological reserve, but its value in prognostication in TACE patients is insufficiently studied. Therefore, the aim of the study was to evaluate the prognostic value of a fully automated, open-source pipeline for CT-based body composition analysis in predicting overall survival (OS) in patients with HCC undergoing TACE. In this study, we used two independent cohorts of treatment-naive patients undergoing TACE: the WAW-TACE cohort (development; n = 230, OS: 28.6 months) and the HCC-TACE-Seg cohort (validation; n = 100, OS: 24.0 months). Skeletal muscle and fat metrics were extracted from pre-treatment CTs using a standardized deep learning pipeline and normalized by sex. Survival analyses were performed using Cox proportional hazards (CoxPH) models and random survival forests (RSF). Skeletal muscle density (SMD) at the L3 level was the strongest and independent predictor of OS across both cohorts (HR: development, 0.84; p = 0.029; validation, 0.79; p = 0.028). This association remained significant after adjustment for the best-performing clinical composite scores: mHAP-2 in the development (adjusted HR = 0.68; p = 0.049) and CLIP in the validation cohort (adjusted HR = 0.43; p = 0.003). In CoxPH, the addition of SMD metrics resulted in only modest improvements in discrimination (ΔC-index 0.011-0.037) that did not reach statistical significance. In contrast, RSF analysis demonstrated a statistically significant improvement in model discrimination when muscle-based variables were added to clinical features (ΔC-index = 0.023; p < 0.001). In both cohorts, SMD showed a reproducible independent prognostic association with overall survival. While adding SMD to traditional clinical models resulted in only modest, and in Cox-based analyses not statistically significant, improvements in discrimination, SMD provided complementary prognostic information. This suggests that the primary value of these automated CT-derived body composition metrics lies not in their performance as standalone predictors, but in their ability to provide an additional layer of objective biological data that may contribute to risk stratification in a complementary and exploratory manner within multivariable frameworks. Notably, in internally cross-validated RSF analyses, statistically significant increases in model discrimination were observed when muscle-based features were integrated into the model, highlighting their potential complementary value within machine learning frameworks.
Fetiform teratoma (FT) and fetus-in-fetu (FIF) represent a spectrum of rare retroperitoneal masses containing organoid structures. While FIF is classically defined by the presence of a vertebral axis, FT lacks this organized skeletal development. Distinguishing between these entities is critical given the malignant potential associated with FT, estimated at approximately 10%. We report a case of a 5-month-old male presenting with a large (12 cm) retroperitoneal mass and elevated alpha-fetoprotein (AFP 56.8 IU/mL; age-matched reference <7 IU/mL). Macroscopically, the resected tumor featured a distinct rudimentary digitiform projection with a nail bed. Histopathology demonstrated extensive organoid differentiation, including gastrointestinal loops with muscular layers, respiratory epithelium, and well-formed pancreatic parenchyma and adrenal cortex. Despite the complex organogenesis and limb-like morphology, the absence of a vertebral column or ossified long bones supported a diagnosis of mature cystic teratoma with fetiform features (FT) over FIF. This report highlights the diagnostic ambiguity within the "gray zone" of these lesions and emphasizes the role of axial skeletal organization and serum AFP levels as complementary tools for classification and oncologic surveillance.