Synovial sarcoma is among the most common non-rhabdomyosarcoma soft tissue sarcomas in adolescents and young adults, yet it remains a rare and formidable diagnosis. Tumors arising in the shoulder region account for approximately 6 percent of all synovial sarcomas, and the disease typically does not originate from inside the joint but from the surrounding soft tissues. The tumor often progresses slowly, taking months to years before patients present for evaluation, which can result in a significant delay in diagnosis. In this case, the girl’s symptoms had been present for several months and had gradually worsened, eventually prompting clinical evaluation when the pain began interfering with daily activities and sports participation. Physical examination revealed localized bulging of the right shoulder, reduced range of motion, and tenderness on palpation of the periarticular region.

The diagnostic journey began with plain radiography and ultrasound of the shoulder. The radiograph revealed a soft tissue mass of increased density adjacent to the joint, without significant calcification or osseous involvement. Subsequent ultrasound demonstrated a large heterogeneous soft tissue mass involving the periarticular region and extending into adjacent muscular compartments. The lesion appeared poorly defined, with marked internal heterogeneity—findings that immediately raised suspicion for an underlying malignant soft tissue tumor. These initial assessments, while not diagnostic on their own, served as the critical first step in a multimodal imaging strategy that would ultimately characterize the tumor from both anatomical and functional perspectives.

Magnetic resonance imaging, performed on a 1.5-T scanner with a standard musculoskeletal protocol including T1-weighted, T2-weighted, and STIR sequences without intravenous contrast medium, demonstrated a large soft tissue mass measuring approximately 18 centimeters in maximum diameter. The lesion showed a multilobulated configuration with marked internal heterogeneity and areas of variable signal intensity on T2-weighted sequences, characterized by both cystic and solid components, with no enhancement in the areas containing cysts. Fluid-fluid levels were noted together with the characteristic triple sign, consisting of the coexistence of low, intermediate, and high signal intensity regions. These features suggested the diagnosis of synovial sarcoma. The tumor involved several muscles of the shoulder girdle, including the deltoid, infraspinatus, and subscapularis muscles, and extended into the surrounding soft tissues. MRI is considered the reference modality for local staging due to its excellent soft tissue contrast and its ability to assess tumor extent and relationship with surrounding structures, information that is essential for surgical planning and for defining target volumes in radiotherapy.

However, MRI findings in synovial sarcoma are not pathognomonic, and additional imaging techniques are often required to fully characterize the disease. In this case, the evaluation of possible bone involvement was initially challenging due to periosteal reaction and inflammatory changes. For this reason, contrast-enhanced computed tomography was performed and confirmed the absence of direct invasion of the humeral head. Moreover, contrast-enhanced CT provided important information on the enhancement pattern of the lesion, contributing to the characterization of tumor heterogeneity. Most synovial sarcomas demonstrate heterogeneous enhancement with reduced or absent enhancement in areas of cystic necrosis or internal septations, and this case was no exception. The CT findings, combined with the MRI data, allowed clinicians to confidently exclude bone invasion while gaining a deeper understanding of the tumor’s complex internal architecture.

To complete staging, a whole-body 18F-FDG PET/CT was performed three weeks after the MRI, acquired 60 minutes after tracer injection according to the institutional pediatric oncology protocol. The examination revealed a highly heterogeneous pattern of metabolic activity within the primary tumor, with a maximum standardized uptake value (SUVmax) of 12.3, a mean activity (SUVmean) of 1.5, and a peak activity (SUVpeak) of 7.7. Areas of intense uptake corresponded to viable tumor tissue, while regions of absent uptake corresponded to necrosis or fluid components. A key diagnostic finding was the identification of metabolically active lymph nodes in the ipsilateral axilla—information that had direct implications for disease staging and therapeutic planning. PET/CT proved particularly valuable in identifying regional lymph node involvement that was not clearly detected on morphological imaging, confirming its well-recognized higher sensitivity in detecting nodal and distant disease in pediatric sarcomas.

Histopathological examination of the excisional biopsy confirmed the diagnosis of monophasic poorly differentiated synovial sarcoma. Molecular testing demonstrated the presence of the characteristic chromosomal translocation t(X;18)(p11;q11) involving the SS18 gene, the genetic hallmark of this tumor type. Following multidisciplinary discussion, the patient started multimodal therapy consisting of combination chemotherapy with ifosfamide and doxorubicin, according to the EpSSG NRSTS 2005 protocol. After completion of three cycles of chemotherapy, which were well tolerated with no relevant treatment-related complications, a follow-up imaging evaluation was performed to assess treatment response.

The follow-up imaging told a fascinating story about the biology of treatment response. Contrast-enhanced CT demonstrated persistent tumor volume without clear evidence of disease progression, but with changes in the enhancement pattern of the solid tumor component. Meanwhile, the whole-body 18F-FDG PET/CT demonstrated a clear reduction in metabolic activity of the primary tumor compared with baseline, consistent with an early metabolic response to therapy. SUVmax decreased from 12.3 at baseline to 6.7 at follow-up, and SUVpeak decreased from 7.7 to 2.6. These structural changes corresponded to the reduction in metabolic activity observed on follow-up PET/CT, suggesting an early biological response to therapy preceding measurable tumor shrinkage. This finding highlights the ability of functional imaging to detect early treatment response before morphological tumor shrinkage becomes evident, a phenomenon that has been reported in high-grade soft tissue sarcomas and supports the role of PET/CT in early response assessment and treatment monitoring. While tumor size remains a widely used parameter for response evaluation, changes in metabolic activity may represent earlier and more sensitive indicators of therapeutic effectiveness—information that would not have been captured by size-based criteria alone.

Semiquantitative metabolic parameters, including SUVmax, SUVmean, and SUVpeak, are commonly used to characterize tumor biology and monitor response to therapy. Among these, SUVpeak, which reflects metabolic activity within a small, standardized tumor volume, may provide a more robust and reproducible measure than single-voxel metrics in heterogeneous lesions. Although considerable variability exists in reported thresholds due to differences in scanners and acquisition protocols, in synovial sarcoma a baseline SUVmax above approximately 5 has been associated with higher histological grade and worse clinical outcomes, supporting its role as a clinically relevant imaging biomarker for risk stratification. In addition, metabolic imaging with 18F-FDG PET/CT can help identify metabolically active tumor regions within heterogeneous masses, which are more likely to correspond to areas of higher cellularity and biological aggressiveness. These regions may represent optimal targets for biopsy and may support biologically guided radiotherapy planning, including potential dose escalation to high-risk tumor subregions while sparing surrounding normal tissues, thereby improving local tumor control and reducing treatment-related toxicity.

Interestingly, the follow-up PET/CT also detected multiple areas of symmetrical intense FDG uptake bilaterally in the neck area, the posterior mediastinum, bilateral costovertebral margins, and bilateral suprarenal fossa, extending to perinephric fat, which anatomically correlated to fat at CT—consistent with functioning brown fat. Although typically considered a physiological finding, recent evidence suggests that brown fat activation on FDG PET/CT may reflect systemic metabolic status and has been associated with improved survival outcomes in oncologic patients. In the present case, this finding was interpreted as a physiological phenomenon; however, its potential prognostic implications deserve further investigation, particularly in pediatric and adolescent populations. The case underscores the importance of an integrated imaging strategy in the management of pediatric soft tissue sarcomas and highlights the clinical value of combining complementary imaging techniques in rare and complex presentations. The main limitation of this report is the relatively short duration of follow-up, which does not allow assessment of long-term outcomes. Nevertheless, individual case reports remain valuable in the context of rare pediatric malignancies, where clinical experience is limited and each well-documented case contributes to improving diagnostic awareness and management strategies. At the time of the most recent follow-up, the patient reported good tolerance of therapy and maintained a positive outlook regarding treatment.

Subject of Research: Multimodal imaging with MRI, CT, and 18F-FDG PET/CT for staging and response assessment of pediatric synovial sarcoma

Article Title: Case Report: Role of MRI, CT, and 18F-FDG PET/CT in staging and response assessment of pediatric synovial sarcoma

Article References: Picori, L., Maccio, L., Meneghello, L., Caldarella, C., Donner, D., & Feraco, P. (2026). Case Report: Role of MRI, CT, and 18F-FDG PET/CT in staging and response assessment of pediatric synovial sarcoma. Frontiers in Nuclear Medicine, 6, Article 1870009. https://doi.org/10.3389/fnume.2026.1870009

Image Credits: AI Generated

DOI: 10.3389/fnume.2026.1870009

Keywords: synovial sarcoma, pediatric oncology, MRI, CT, 18F-FDG PET/CT, SUVmax, tumor staging, treatment response, soft tissue sarcoma, ifosfamide, doxorubicin, triple sign

The study, published in Nature Machine Intelligence by Luyang Cao, Zhijian Wu and colleagues at Nanjing University, Westlake University, Southeast University and the Shenzhen Institute of Advanced Technology, was trained and evaluated on an extraordinary corpus: 3.76 million paired images, in which each degraded clinical image is matched with a high-quality reference. The data were assembled from publicly available repositories including The Cancer Imaging Archive, OpenNeuro, NITRC, Grand Challenge, Synapse, Kaggle and GitHub, and span 34 MRI datasets covering brain, cardiac, abdominal and prostate imaging, 37 CT datasets ranging from thoracic to whole-body acquisitions, six X-ray datasets, five endoscopy datasets, four fundus photography datasets, two ultrasound datasets and one whole-body PET dataset for oncological imaging.

Architecturally, AMIFound follows a philosophy that has transformed natural-language processing but has been slower to arrive in medical image restoration: a shared backbone with adaptive, task-specific routing. A single encoder first maps any low-quality input, regardless of modality, into a common latent feature space that captures both modality cues and the nature of the degradation. A dynamic routing mechanism then selects from a pool of specialised expert modules to construct a restoration pathway tailored to each individual image. The expert pool contains fast Fourier transform attention experts that operate in the frequency domain, spatial attention experts, convolutional local-feature experts and hybrid FFT-CNN fusion experts, allowing the model to combine complementary frequency, spatial and structural information on the fly.

Training such a system required a way to generate realistic degradation across modalities. The researchers built a modality-aware degradation engine that transforms high-quality images into degraded observations using physics-informed simulations. For CT, metal artefacts are synthesised in the projection domain through random metal masks, Radon transformation, nonlinear artefact mapping and inverse reconstruction. MRI aliasing and low-resolution effects are produced by k-space truncation followed by inverse Fourier reconstruction, while PET low-dose degradation is simulated with Poisson noise, Gaussian noise and blur. Endoscopic low-light conditions are modelled by gamma correction and brightness reduction in HSV colour space, X-ray blur by cascaded filtering with anisotropic diffusion, ultrasound artefacts by reverberation echoes, geometric shadowing and depth attenuation, and fundus artefacts by random circular highlights and Gaussian softening. In total, the model handles seven representative degradation patterns.

The benchmark results are striking. Across all seven modalities, AMIFound outperformed 38 competing baselines, including both task-specific restoration networks and generalist models such as AMIR, MoCE-IR and Found-IR, on full-reference image quality metrics such as structural similarity and learned perceptual measures. Radar plots of structural similarity improvements show consistent gains over each competing framework, and detailed per-task comparisons show the model achieving the best or near-best scores against modality-specific methods for every restoration task, from CT metal artefact reduction to PET denoising and cataract-affected fundus enhancement.

Quantitative metrics alone do not convince radiologists, so the team also ran a subjective evaluation with clinicians across all seven modalities. Physicians rated the restored images on sharpness, artefact suppression, naturalness and diagnostic confidence. AMIFound achieved the highest or near-highest average scores in every modality and provided greater diagnostic confidence than competing methods. The model was also stress-tested under combined degradations, where multiple corruptions strike the same image simultaneously, and on naturally degraded clinical images rather than synthetic ones, maintaining its advantage in both settings and preserving anatomical structures and realistic intensity distributions where competitors tended to oversmooth or distort contrast.

Perhaps the most clinically consequential finding is what happens downstream. The researchers fed AMIFound-enhanced images into 44 disease detection tasks and found consistent improvements over detection on the original degraded images. The benefits extended to tumour stage prediction in non-small-cell lung cancer and to survival risk stratification, including improved individual-level prognostic ranking. Attention heatmaps from the CT-based tumour-stage classifier reveal a plausible mechanism: before enhancement, model attention is often diffuse or mislocalized and predictions are less certain, whereas after enhancement, attention concentrates on tumours, lung-parenchyma abnormalities and lesion boundaries, and the highest prediction probability more often agrees with the ground-truth stage.

The significance of this work lies partly in its economics. Hospitals generate enormous volumes of routine imaging that falls short of research-grade quality, and low-dose protocols, deliberately chosen to reduce radiation exposure, or motion artefacts from patients who cannot hold still, are unavoidable facts of clinical life. Previous approaches required a separate, carefully tuned network for each modality-degradation combination, an unsustainable proliferation as imaging technology diversifies. A single foundation model that generalises across modalities and degradations, trained once and deployed everywhere, could lower the barrier for smaller institutions and make image enhancement a routine preprocessing step for any AI diagnostic pipeline.

The authors have released their training and evaluation code publicly on GitHub, with a versioned archive on Zenodo, built on PyTorch and drawing on data-loader pipelines from prior all-in-one restoration frameworks such as PromptIR, AirNet and MoCE-IR. All datasets used were publicly available and accessed under their respective licences, and the team reports that the model preserves clinically relevant information rather than hallucinating plausible-looking detail, a critical concern for any generative approach in medicine. Funded by the National Natural Science Foundation of China and several regional programmes, the work positions AMIFound as a domain-specific foundation model for the unglamorous but essential task of making everyday clinical images good enough, not just for human eyes, but for the growing fleet of AI systems that depend on them.

Subject of Research: A multimodal foundation model for restoring and enhancing degraded medical images across seven imaging modalities

Article Title: An all-in-one foundation model for multimodal medical image restoration and enhancement

Article References: Cao, L., Wu, Z., Zhang, J., Qi, L., Gao, Y., Zheng, H., Zheng, Y., & Shi, Y. (2026). An all-in-one foundation model for multimodal medical image restoration and enhancement. Nature Machine Intelligence. https://doi.org/10.1038/s42256-026-01310-z

Image Credits: AI Generated

DOI: 10.1038/s42256-026-01310-z

Keywords: medical imaging, foundation model, image restoration, deep learning, AMIFound, CT, MRI, PET, ultrasound, endoscopy, lung cancer, diagnostic AI

The research, published as an open-access article in BMC Cancer, set out to systematically characterize the computed tomography (CT) and magnetic resonance imaging (MRI) features of intracranial sarcoma metastases. The team, led by neuroradiologist Giovanni Librizzi and oncologist Benedetta Chiusole, together with senior author Renzo Manara, reviewed the records of patients followed for intracranial sarcoma metastases between January 2011 and April 2024. To keep the analysis rigorous, they excluded patients who lacked pretreatment brain imaging and those whose disease was confined to the skull itself. What remained was a cohort of 22 patients whose pre-treatment scans could be scrutinized in detail, comprising 15 MRI studies and 18 CT examinations, spanning twelve different sarcoma tumor types.

The methodology was deliberately meticulous. Two neuroradiologists, blinded to the pathological diagnosis, independently assessed each patient’s imaging. They recorded the number and location of metastases, the size of the largest lesion, and a battery of morphological features: the presence of perilesional edema, calcification, hemorrhage, cysts, necrosis, and leptomeningeal disease, in which tumor cells spread through the membranes surrounding the brain. They also measured CT density and MRI signal intensity both before and after intravenous contrast administration, quantified Apparent Diffusion Coefficient values on diffusion-weighted imaging, and evaluated perfusion characteristics. This comprehensive approach meant that essentially every dimension of modern neuroradiological assessment was brought to bear on a disease entity that had never been systematically profiled in this way.

The headline finding is heterogeneity, and lots of it. Ten of the 22 patients presented with multiple brain metastases from the outset, and one patient had diffuse leptomeningeal seeding, meaning tumor cells had spread widely through the cerebrospinal fluid spaces. More striking still, the lesions were heterogeneous not only between patients with different tumor types but also between different metastases within the same patient. In practical terms, two tumors sitting in the same brain, originating from the same primary sarcoma, could look dramatically different on the same scan, differing in density, signal intensity, enhancement pattern, and surrounding swelling. For radiologists accustomed to the relatively predictable appearance of metastases from lung, breast, or melanoma primaries, this variability presents a genuine diagnostic challenge.

Yet within that chaos, the researchers identified patterns that appear to be associated with specific sarcoma subtypes. All three patients with chondrosarcoma metastases in the cohort displayed the T2-FLAIR mismatch sign, an imaging phenomenon in which a lesion appears bright on T2-weighted sequences but does not suppress, or darken, on FLAIR sequences. This sign is familiar to neuroradiologists as a hallmark of certain low-grade gliomas, particularly those with specific mutations, but its appearance in chondrosarcoma metastases had not been emphasized before. The sign was absent or uncommon in the other sarcoma subtypes studied, making it a candidate marker that could raise suspicion of chondrosarcoma when a patient with an unknown primary presents with brain lesions.

Ewing sarcoma, an aggressive tumor most often seen in children and young adults, showed an equally distinctive pattern. Every Ewing sarcoma patient in the study had exclusively pachymeningeal disease, meaning the tumor involved the tough outer dura mater rather than forming discrete masses within the brain parenchyma. This dural predilection distinguishes Ewing metastases from the intraparenchymal lesions typical of many other cancers and could help clinicians interpret scans in young patients with known or suspected Ewing sarcoma. Meanwhile, osteosarcoma patients showed yet another signature: their brain metastases exhibited gross calcification, visible as dense, bone-like deposits on CT. This feature makes intuitive biological sense, given that osteosarcoma is a tumor that produces osteoid, or bone matrix, but confirming it radiologically provides a useful diagnostic clue. These three features, the T2-FLAIR mismatch in chondrosarcoma, pachymeningeal disease in Ewing sarcoma, and calcification in osteosarcoma, were absent or uncommon in the other subtypes examined.

The study also documented genuinely unusual presentations that underline how deceptive these tumors can be. One chondrosarcoma patient harbored a mixture of conventional brain metastases and multiple neoplastic emboli, fragments of tumor that had lodged in cerebral blood vessels, producing a picture that could easily be mistaken for stroke or vascular disease. In another remarkable case, a leiomyosarcoma metastasis was invisible on standard sequences and could be detected only on diffusion-weighted imaging, the MRI technique that is sensitive to the random motion of water molecules and is typically used to identify acute stroke. Such cases serve as cautionary tales: a routine contrast-enhanced brain MRI protocol, interpreted without awareness of these possibilities, could theoretically miss disease or mischaracterize it entirely.

Why does this matter clinically? Brain metastases from sarcomas are rare, but their incidence may be rising as systemic therapies extend survival and allow more patients to live long enough for intracranial spread to manifest. Treatment decisions, whether surgery, stereotactic radiosurgery, whole-brain radiotherapy, or systemic agents, often hinge on the number, size, and location of lesions, and increasingly on the biological behavior inferred from imaging. If certain imaging features reliably point to a specific sarcoma subtype, they could accelerate diagnosis in patients whose primary tumor is unknown, inform biopsy planning, and help oncologists anticipate complications such as hemorrhage or leptomeningeal spread. The T2-FLAIR mismatch sign, in particular, is intriguing because it links sarcoma imaging to a well-established concept in neuro-oncology, potentially opening a bridge between two fields that rarely intersect.

The authors are appropriately cautious about overinterpreting their results. With 22 patients and twelve tumor types, the cohort is small, and several subtypes are represented by only one or two individuals. The candidate type-associated imaging features they identified require confirmation in larger, multicenter cohorts before they can be incorporated into diagnostic algorithms or clinical guidelines. Rare-disease research inevitably faces this tension: the very rarity that makes sarcoma metastases poorly understood also makes it difficult to assemble the patient numbers needed for statistically robust conclusions. Single-center studies like this one are best viewed as hypothesis-generating, laying the groundwork for collaborative efforts that pool cases across institutions and countries.

Even so, the study represents a meaningful step forward for a neglected corner of neuro-oncology. By applying a systematic, blinded, quantitative imaging analysis to a group of patients whose scans had never been collectively characterized, the Padua team has transformed scattered anecdotes into a structured spectrum. The message for radiologists and oncologists is twofold: first, assume nothing when interpreting brain lesions in sarcoma patients, because heterogeneity is the rule both within and between individuals; and second, pay attention to specific clues, FLAIR behavior, dural involvement, and calcification, that may whisper the tumor’s identity. As survival improves for sarcoma patients worldwide, the brain is becoming an increasingly important frontier in their care, and studies like this one ensure that the imaging tools needed to meet that challenge are finally beginning to take shape.

Subject of Research: Neuroradiological features of intracranial metastases from sarcoma subtypes

Article Title: Neuroradiological spectrum of sarcoma intracranial metastases: a single center study

Article References: Librizzi, G., Murtas, F., Chiusole, B., Tortorelli, I., Valeggia, S., Vizzaccaro, S., Di Maggio, A., Buzzaccarini, M. S., Ghirelli, A., Brunello, A., & Manara, R. (2026). Neuroradiological spectrum of sarcoma intracranial metastases: a single center study. BMC Cancer. https://doi.org/10.1186/s12885-026-17009-8

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17009-8

Keywords: sarcoma, brain metastases, neuroradiology, MRI, CT, chondrosarcoma, Ewing sarcoma, osteosarcoma, T2-FLAIR mismatch, leptomeningeal disease, cancer imaging, tumor heterogeneity

The clinical problem the Italian team set out to address is well known to anyone who works in pediatric oncology. When a child is diagnosed with a malignant renal mass, the surgical team must decide whether to remove the entire kidney, a procedure called radical nephrectomy, or to attempt nephron-sparing surgery, in which only the tumor and a thin margin of healthy tissue are excised. Nephron-sparing surgery is strongly preferred whenever it is safe, because children who keep functioning kidney tissue face a lower lifetime risk of chronic kidney disease, hypertension, and the long-term consequences of reduced renal reserve. But deciding whether the tumor can be safely shaved away from vital vessels and collecting structures requires an exquisitely detailed understanding of each patient’s unique anatomy — something that flat CT and MRI slices, read one at a time, do not always convey intuitively even to experienced surgeons.

The Turin group, led by pediatric urological surgeon Elisa Cerchia together with colleagues from the hospital’s pediatric surgery, oncology, and radiology units and the urology department of the University of Torino, tested a straightforward technical workflow. Standard abdominal MRI or CT examinations were performed for each patient as part of normal clinical care. The image data, stored in the universal DICOM format that medical scanners produce, were then imported into segmentation software. Segmentation is the computational process of outlining, slice by slice, the boundaries of each structure of interest — the tumor, the renal artery and its branches, the renal vein, the healthy kidney parenchyma, and the urinary collecting system. Once each structure is traced, the software renders it as a colored, fully rotatable three-dimensional object that can be zoomed, turned, made translucent, or viewed from any angle on a screen, effectively letting the surgeon fly through the child’s anatomy before making a single incision.

Eight patients, four boys and four girls, with complex renal masses were included in the study. For each child, the 3D virtual reconstruction was completed before surgery and reviewed by the surgeons and the pediatric radiologist involved in the case. To evaluate the technology rigorously rather than anecdotally, the team designed a modified closed-ended questionnaire administered in two phases, one before the operation and one afterward. Respondents compared the 3D models directly against the conventional imaging studies, rating how well each modality displayed the tumor and its relationships with arteries, veins, and the urinary tract, and assessing whether the reconstruction helped define the criteria for choosing nephron-sparing surgery over radical removal. The questionnaire also asked whether the models were practical to consult during the operation itself.

The statistical results were striking. For the tumor, the arteries, and the veins, the 3D virtual reconstructions were rated significantly better than conventional imaging, with p-values below 0.0001 for each of those structures — a level of significance that indicates the difference is extremely unlikely to be a statistical fluke even in a small case series. The urinary tract was the one exception; here the advantage of the 3D models did not reach statistical significance, with a p-value of 0.0888, suggesting that the collecting system may already be adequately visualized by standard scans or that segmentation of this thin-walled structure is technically harder to render convincingly. Even so, the overall pattern was unmistakable: the interactive models outperformed the flat images that surgeons have relied on for decades.

Perhaps the most clinically meaningful finding concerned surgical decision making. The study found that 3D virtual reconstruction was significantly more effective in defining the elective criteria for nephron-sparing surgery, with a p-value of 0.0003. In practical terms, this means the models helped surgeons judge whether a tumor could be removed while preserving healthy kidney tissue — the single most consequential decision in these operations. The questionnaire results also showed that the models were fundamentally easier to use during surgery than conventional images, with a p-value of 0.0006. A surgeon in the operating room can glance at a rotatable 3D model and instantly re-orient themselves in three-dimensional space, whereas scrolling through stacks of two-dimensional slices mid-operation is cumbersome and mentally taxing.

In the series, the eight children underwent a total of seven recorded operations: four radical nephrectomies and three nephron-sparing procedures. That three of the children could be treated with kidney-preserving surgery at all reflects the kind of complex, carefully selected cases for which preoperative three-dimensional planning is most valuable. In tumors judged too large or too centrally located for partial removal, the models still contributed by clarifying vascular anatomy before a radical nephrectomy, where precise knowledge of the renal artery and vein is essential for safe vessel control. The authors conclude that 3D virtual reconstruction is a genuinely useful tool in the preoperative evaluation of children with complex renal masses and that it facilitates nephron-sparing surgery in selected patients.

The technical pipeline behind these results deserves attention because it is reproducible in any hospital with standard imaging equipment. No additional scanning is required; the same DICOM datasets acquired for diagnosis feed the reconstruction. Segmentation software separates the anatomy into labeled structures, and rendering engines assemble them into interactive volumes. Some centers in the broader field go further and 3D-print physical models from the same data, which can be sterilized and brought into the operating theater, but the Turin study focused on the virtual models themselves, which require no printing hardware and can be shared instantly across the surgical team, from the operating room to the tumor board where treatment decisions are made.

What makes the study especially credible is its multidisciplinary design. The questionnaire was completed not only by the operating surgeons but also by the pediatric radiologist responsible for interpreting the original scans, meaning the comparison between 3D and conventional imaging was evaluated from multiple professional perspectives. The work also drew on collaboration with adult urological surgeons at the Città della Salute e della Scienza hospital in Turin, bringing experience from the larger field of kidney surgery into the pediatric arena. The project received no external funding, and the authors declare no competing interests, with technical support for the reconstruction workflow acknowledged from a collaborating company. The study was approved by the local ethics committee, and informed consent was obtained from each child’s parents or guardian in accordance with hospital rules.

Caveats remain, as they do with any small, single-center case series. Eight patients cannot establish how often 3D planning changes the final operative decision, whether it shortens operative times, reduces complications, or improves long-term kidney function — outcomes that will require larger, ideally multi-center comparative studies. The near-significant result for the urinary tract hints that some structures remain harder to model than others. Still, the consistency and strength of the reported significance levels, particularly for tumors and blood vessels, give the findings real weight. For a specialty in which the stakes are measured in decades of renal health for young patients, the message from Turin is compelling: letting surgeons rehearse a child’s cancer in three dimensions before entering the operating room is no longer a futuristic gimmick but a practical, statistically validated planning tool that may help more children keep the kidney tissue they will need for the rest of their lives.

Subject of Research: Use of 3D virtual reconstruction from MRI and CT scans for preoperative planning of oncologic renal surgery in children

Article Title: 3D virtual reconstruction: usefulness in preoperative planning for oncologic renal surgery in children

Article References: Cerchia, E., Ruggiero, E., Teruzzi, E., Serpentino, M., Cirigliano, L., Quarello, P., Petraz, M., Catti, M., Allasia, M., Gontero, P., Fagioli, F., & Nappo, S. G. (2026). 3D virtual reconstruction: usefulness in preoperative planning for oncologic renal surgery in children. 3D Printing in Medicine. https://doi.org/10.1186/s41205-026-00340-y

Image Credits: AI Generated

DOI: 10.1186/s41205-026-00340-y

Keywords: 3D virtual reconstruction, pediatric surgery, renal tumors, nephron-sparing surgery, surgical planning, 3D Printing in Medicine, DICOM segmentation, pediatric oncology, kidney cancer, preoperative imaging, MRI, CT

The research, led by Jinnian Ge, Yuxin Feng, Xubin Xie, Ruihuan Gao, Dan Sun, Zijian Zhang, Qin Zhou, and senior author Liangfang Shen, set out to stratify patients by their risk of developing nasopharyngeal necrosis within five years of primary intensity-modulated radiotherapy. The team retrospectively assembled a cohort of 170 patients with nasopharyngeal carcinoma, of whom 62 developed the complication. That ratio, roughly one in three, is far higher than the incidence seen in the general treated population, and it reflects a deliberate enrichment of the study sample with positive cases to give the learning algorithms enough examples of the event to learn from. It is a common and defensible strategy in rare-outcome modeling, but as the authors themselves emphasize, it has consequences for how the model’s probability outputs should be interpreted, a point that becomes important later in the analysis.

What distinguishes this work from many prior risk-prediction studies is the sheer richness of the input data. Rather than relying on a handful of clinical variables, the framework ingests four distinct imaging and dosimetric modalities for each patient: the planning computed tomography scan used for treatment preparation, contrast-enhanced T1-weighted magnetic resonance imaging, T2-weighted magnetic resonance imaging, and the full three-dimensional radiation dose distribution computed by the treatment planning system. Each of these data streams carries complementary information. CT delineates bony anatomy and provides the geometric backbone of dose calculation. The contrast-enhanced T1 sequence highlights vascular enhancement and tumor extent, while T2-weighted images are sensitive to edema, mucosal changes, and tissue fluid content. The 3D dose map, meanwhile, encodes exactly how much radiation energy was deposited at every point in space, which is critical because necrosis is fundamentally a dose-dependent injury of tissue that fails to repair itself.

Architecturally, the pipeline is built around four modality-specific three-dimensional encoders, one for each input stream. Each encoder is a convolutional neural network that processes its volumetric data in three dimensions rather than slice by slice, preserving the spatial relationships between structures in the head and neck. The outputs of these encoders are then brought together by a cross-modal Transformer, the attention-based architecture that has transformed fields from language modeling to computer vision. The Transformer’s attention mechanism allows the model to learn which features from one modality are relevant to features in another, for example linking a region of intense contrast enhancement on MRI to the corresponding high-dose region in the radiation map. Finally, rather than simply concatenating all features into one vector, the framework employs an adaptive mixture-of-modality-experts fusion strategy, in which specialized expert subnetworks process different modality combinations and a gating mechanism learns, patient by patient, how much weight to give each expert.

The rationale for adaptive fusion over naive concatenation is one of the study’s most instructive technical findings. When the researchers compared their full framework against a matched baseline that simply concatenated features from all modalities, the adaptive expert fusion achieved an area under the receiver operating characteristic curve of 0.907 compared with 0.798 for the concatenation model, a difference that was statistically significant with a DeLong test p-value below 0.001. The authors also re-implemented four other published baseline models, and the proposed framework outperformed four of the five with p-values of 0.009 or lower. Its margin over the strongest competing baseline, an improvement of 0.036 in AUC, did not reach conventional statistical significance, with p equal to 0.089, a nuance the team reports honestly rather than glossing over. In a field where multimodal fusion is often claimed as a panacea, this careful ablation provides concrete evidence that how modalities are combined matters as much as which modalities are used.

The ablation experiments also quantify the contribution of each data stream. Using only the dose distribution and CT, the model achieved an AUC of 0.695, which is modest and close to what one might expect from dose-volume statistics alone. Adding both MRI sequences lifted performance to 0.907, a dramatic jump that underscores how much predictive signal resides in the soft-tissue appearance of the nasopharynx before treatment. In other words, the pre-existing condition of the mucosa and underlying tissue, as revealed by magnetic resonance, interacts with the radiation dose in ways that determine whether the tissue withstands therapy. This finding aligns with the biological understanding that necrosis risk is not purely a function of dose but of dose delivered to vulnerable tissue, and it suggests that radiomic features extracted from MRI may serve as surrogates for that vulnerability.

Methodological rigor is a second hallmark of the study. Performance was estimated using leakage-controlled five-fold stratified cross-validation, meaning that the split between training and testing folds was designed to prevent any information from a held-out patient from influencing the model’s training, and each fold preserved the overall proportion of necrosis cases. Model selection was performed on an internal validation split within each training fold, so the test fold remained untouched until final evaluation. The authors report 95 percent bootstrap confidence intervals for every metric and use the DeLong test for pairwise comparisons of correlated receiver operating characteristic curves. On the held-out folds, the framework achieved an AUC of 0.907 (95 percent confidence interval 0.861 to 0.948), accuracy of 0.876, sensitivity of 0.888, specificity of 0.872, positive predictive value of 0.808, and negative predictive value of 0.933. The Brier score, which penalizes both poor discrimination and poor calibration, was 0.140, the lowest among all compared models.

That negative predictive value of 0.933 deserves particular attention from a clinical standpoint. In practical terms, it means that when the model tells a clinician that a patient is at low risk of developing nasopharyngeal necrosis within five years, that reassurance is correct in more than nine out of ten cases in this cohort. A reliable low-risk designation could spare patients unnecessary intensive surveillance, repeated endoscopies, and anxiety, while concentrating follow-up resources on the smaller group flagged as high risk. Decision curve analysis, which evaluates the net clinical benefit of acting on a model’s predictions across a range of decision thresholds, showed positive net benefit over a wide span of threshold probabilities in the study sample, suggesting the model could in principle support meaningful triage decisions rather than merely ranking patients.

Yet the authors are notably candid about the study’s limitations, and their honesty is itself a model for the field. Because the cohort was enriched with necrosis cases, the absolute probabilities the model outputs are miscalibrated: a predicted probability from this model does not translate directly into a real-world risk for a typical clinic patient in whom necrosis is much rarer. The team explicitly calls for probability recalibration before any clinical deployment. The study is also retrospective and single-center, drawing on patients treated at Xiangya Hospital, and the model was validated internally rather than on external data from other institutions with different scanners, treatment protocols, and patient demographics. Deep-learning models in medicine are notorious for performance degradation when moved across such domain shifts, and the authors acknowledge that external prospective validation is indispensable. Finally, and perhaps most fundamentally, they note that no evidence yet exists linking risk-guided surveillance strategies to improved clinical outcomes; showing that a model can stratify risk is a necessary but not sufficient step toward showing that acting on that stratification saves lives or reduces morbidity.

Even with those caveats, the study represents a meaningful advance in the application of multimodal artificial intelligence to radiation oncology complications. Post-radiation nasopharyngeal necrosis sits at the intersection of dosimetry, radiology, and tissue biology, and no single data source captures the full picture. By demonstrating that adaptive fusion of CT, dual-sequence MRI, and 3D dose can push prediction performance to an AUC above 0.9 in cross-validation, and by rigorously quantifying the contribution of each modality and each architectural choice, the Xiangya team has laid out a credible technical blueprint for pre-treatment risk stratification. The path from such a blueprint to bedside utility runs through external validation, calibration, and outcome trials, but the destination, a future in which radiation plans are tailored not only to kill tumors but to protect the patients most vulnerable to collateral injury, has come measurably closer.

Subject of Research: Deep learning prediction of post-radiation nasopharyngeal necrosis using multimodal fusion of CT, MRI, and 3D dose distributions

Article Title: Deep learning-based prediction of post-radiation nasopharyngeal necrosis using adaptive multimodal fusion of CT, MRI, and 3D dose

Article References: Ge, J., Feng, Y., Xie, X., Gao, R., Sun, D., Zhang, Z., Zhou, Q., & Shen, L. (2026). Deep learning-based prediction of post-radiation nasopharyngeal necrosis using adaptive multimodal fusion of CT, MRI, and 3D dose. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02894-z

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02894-z

Keywords: nasopharyngeal carcinoma, post-radiation necrosis, deep learning, multimodal fusion, mixture-of-experts, radiotherapy, medical imaging, risk prediction, MRI, CT, Transformer, radiation oncology

Tenosynovial giant cell tumor, or TGCT, is a neoplasm arising from the synovial lining of joints, bursae, or tendon sheaths. Under the 2020 World Health Organization Classification of Soft Tissue and Bone Tumours, the disease is divided by growth pattern into localized and diffuse types. The diffuse variant, historically known as pigmented villonodular synovitis, is a locally aggressive lesion marked by extensive synovial proliferation, nodular growth, and a strong tendency to recur. It most often strikes the knee, but also appears in the finger, ankle, and shoulder. Hip involvement is uncommon, and that rarity is precisely what makes it so dangerous diagnostically: when a deep, inaccessible joint begins to erode bone, the list of possible culprits is long and includes some far more alarming possibilities.

The patient’s story began one week before admission, when he experienced acute chills and high fever and received three days of empirical intravenous cefuroxime at a local hospital, with partial improvement. Then came the hip pain — acute, severe, worst at night, disturbing his sleep, and worsened by movement. Rest and repositioning did not help. On examination he was afebrile at 36.4°C, with marked tenderness over the right hip, pain provoked by internal rotation, adduction, and flexion, and mildly restricted range of motion. There was no redness, warmth, or swelling, no trauma, no weight loss, and no evening fevers. The distal neurovascular examination was unremarkable. In short, nothing about the physical presentation clearly separated infection from tumor.

Imaging told a more revealing story, though not a definitive one. Magnetic resonance imaging of both hips demonstrated abnormal intra-articular synovial tissue in the right hip, with mild bilateral joint effusion that was more pronounced on the right. The lesion, measuring approximately 27 by 35 by 25 millimeters, showed heterogeneous low-to-intermediate signal intensity on T1-weighted images, heterogeneous signal on T2-weighted images, and heterogeneous enhancement after contrast administration — a signal profile characteristic of TGCT, reflecting variable proportions of fibrous tissue, lipid-laden macrophages, fluid, and hemosiderin deposition. The adjacent anterior acetabular column appeared thinned, and mild swelling of the right obturator externus muscle was noted. Computed tomography then showed joint effusion, cystic low-density intra-articular lesions, and clear thinning and erosion of the anterior and inferomedial acetabular wall. Notably, the plain radiograph showed no definite abnormality, underscoring why advanced imaging is essential when plain films fail.

Laboratory testing deepened the ambiguity rather than resolving it. The white blood cell count was 6.46 × 10⁹/L with 68.4% neutrophils, C-reactive protein was elevated at 34.20 mg/L, the erythrocyte sedimentation rate was 25 mm/h, and procalcitonin measured 0.268 ng/mL — a pattern suggestive of low-grade inflammation but far from conclusive. Serum alkaline phosphatase and tumor markers were normal. Whole-body bone scintigraphy showed increased tracer uptake in both hips, sacroiliac joints, and shoulders, which clinicians judged most consistent with degenerative change. With a recent febrile illness, equivocal inflammatory markers, and a destructive-appearing acetabular lesion, the preoperative differential diagnosis spanned infectious or inflammatory synovitis, synovial chondromatosis, and other aggressive synovial or osseous lesions. Because neither imaging nor laboratory work could deliver a verdict, the team proceeded to surgical exploration after multidisciplinary discussion.

What surgeons found inside the joint settled the immediate uncertainty about the lesion’s nature while illustrating why preoperative diagnosis is so difficult. Intraoperative inspection revealed abundant yellowish joint fluid and diffuse proliferation of synovium-like tissue throughout the right hip, with irregular nodular lesions along the anteromedial and inferomedial acetabular wall. The team performed complete excision of the identified intra-articular lesions and involved synovial tissue, curetted the eroded inferomedial acetabular wall, treated it with alcohol ablation, and reconstructed the localized bone defect with an autologous iliac bone graft. Estimated blood loss was approximately 150 milliliters. Although the lesion contained nodular components measuring up to roughly 1.5 by 3.0 centimeters, it was classified as diffuse-type TGCT because of the widespread intra-articular synovial involvement, the acetabular erosion, and the absence of a solitary encapsulated mass.

Histopathologic examination provided the gold-standard confirmation. Microscopy showed proliferative synovium-like tissue with focal cyst wall-like structures lined by flattened to cuboidal cells, a stroma containing proliferative fibrous tissue, histiocytes, prominent congested vessels, diffuse chronic inflammatory cells, scattered neutrophils, and focal coagulative necrosis. Scattered and clustered foamy histiocytes and variable numbers of osteoclast-like multinucleated giant cells were present, along with hemosiderin granules within the cytoplasm of mononuclear cells and histiocytes — the pigment that gave the old disease its name. Immunohistochemistry showed strong CD68 positivity in the multinucleated giant cells and a subset of mononuclear cells, focal S-100 expression, and a Ki-67 labeling index of approximately 5% to 10%, compatible with low proliferative activity. Focal coagulative necrosis was interpreted as a secondary degenerative or ischemic change, and no cytologic atypia, atypical mitotic activity, or sarcomatous overgrowth was identified. Joint-fluid cultures were negative for bacterial and fungal growth, though the authors caution that prior cefuroxime exposure may have reduced culture sensitivity, so infection could never be excluded on microbiologic grounds alone.

The molecular backdrop of this tumor is one of the more fascinating chapters in modern soft-tissue oncology. TGCT is associated with overexpression of colony-stimulating factor 1, or CSF1, which recruits macrophages to the lesion through what researchers call the landscape effect — a minority population of neoplastic cells orchestrating a massive influx of reactive, non-neoplastic immune cells that constitute most of the tumor’s bulk. This mechanism also underpins CSF1/CSF1R-targeted therapies, which are reserved for selected patients with residual, recurrent, or incompletely resectable disease. In this case, however, gross excision and synovectomy were achieved without postoperative radiotherapy or systemic antitumor therapy, with management transitioning instead to clinical and imaging surveillance.

The recovery course was largely favorable, punctuated by one unrelated complication. The patient experienced marked pain relief immediately after surgery, but on postoperative day three he developed fever, mild cough, and fatigue. Chest CT revealed bilateral patchy opacities, pulmonary infection was diagnosed, and intravenous cefuroxime with supportive care normalized his temperature and resolved his systemic symptoms. Follow-up imaging told a reassuring story: CT on the first postoperative day showed satisfactory filling of the acetabular defect by the bone graft, and at two months the graft maintained its position with osseous incorporation and healing. At the two-week mark the incision was well healed and hip motion had improved beyond preoperative levels. By six months, the patient reported sustained pain improvement and had returned to independent daily activities and work, although mild limitation of right hip motion persisted.

The case, reported in Cancer Reports, carries an educational message that resonates well beyond a single patient. Because the hip joint is deeply situated and its capsule has limited capacity to accommodate proliferative synovium, osseous erosion is a recognized feature of hip D-TGCT — yet a destructive-appearing acetabular lesion inevitably raises concern for septic arthritis, inflammatory synovitis, synovial chondromatosis, or a more aggressive neoplasm. The authors emphasize that MRI remains the preferred modality for mapping the extent of synovial disease, while CT more clearly delineates the osseous erosion relevant to operative planning, making the two techniques complementary rather than interchangeable. Ultimately, imaging alone could not reliably distinguish this tumor from infectious or aggressive etiologies; definitive diagnosis depended on integrating the clinical course, intraoperative observations, and characteristic histopathologic features. The authors frame the report primarily as an educational account of diagnostic evaluation and multidisciplinary management rather than evidence of durable local control, noting that the follow-up period remains short and that recurrence is a recognized concern given the complex anatomy of the hip. Longer surveillance will be needed to assess recurrence, joint function, and treatment durability. The broader lesson stands: diffuse-type tenosynovial giant cell tumor deserves a place on the differential diagnosis of any unexplained erosive intra-articular hip lesion, even when the clinical context screams infection.

Subject of Research: Diffuse-type tenosynovial giant cell tumor of the hip with acetabular bone involvement diagnosed by radiologic-pathologic correlation.

Article Title: Diffuse‐Type Tenosynovial Giant Cell Tumor of the Hip With Acetabular Bone Involvement: A Case Report With Radiologic‐Pathologic Correlation

Article References: Quanbing, W., Huanhuan, X., Xing, Z., Tao, Y., Jun, Z., Lei, Z., Kai, L., Lei, L., Feng, A., & Wei, Y. (2026). Diffuse‐Type Tenosynovial Giant Cell Tumor of the Hip With Acetabular Bone Involvement: A Case Report With Radiologic‐Pathologic Correlation. Cancer Reports, 9(9), Article e70696. https://doi.org/10.1002/cnr2.70696

Image Credits: AI Generated

DOI: 10.1002/cnr2.70696

Keywords: tenosynovial giant cell tumor, diffuse-type TGCT, hip, acetabular erosion, MRI, CT, synovectomy, histopathology, CSF1, case report, orthopedic oncology, diagnostic imaging

The review team searched three major biomedical databases, SCOPUS, PubMed, and Web of Science, and identified 129 records for consideration. After screening, 41 studies met the criteria for inclusion and provided usable data on disease type, imaging modalities, and reported radiation dose metrics. The populations covered were diverse: children with congenital heart disease, scoliosis, cystic fibrosis, inflammatory bowel disease, esophageal atresia, osteogenesis imperfecta, spina bifida, hydrocephalus, urological conditions, bleeding disorders, craniosynostosis, cleft palate, asthma, pulmonary hypertension, and organ transplant recipients, among others. This breadth is important, because the authors found that radiation exposure was not uniform across conditions but was strongly influenced by the specific diagnosis, the age at which imaging began, the severity of the disease, and the imaging modalities that each disease trajectory demands.

One of the clearest technical findings of the review is that the modality mix matters enormously. Plain radiography, the conventional X-ray, was by far the most frequently performed examination across nearly all chronic disease groups. However, radiographs deliver relatively small doses per examination, and the review concluded that they were not the dominant contributors to cumulative burden. Instead, computed tomography and fluoroscopy, both of which involve substantially higher dose outputs and, in the case of fluoroscopy, prolonged real-time exposure, accounted for the majority of the cumulative effective dose reported in the included studies. In some patient groups, cumulative doses from these higher-yield modalities exceeded 20 to 50 millisieverts, thresholds that radiation protection specialists regard as significant when accumulated during childhood.

The biological rationale for concern lies in the interaction between ionizing radiation and growing tissue. Effective dose, measured in millisieverts, is a calculated quantity that weights absorbed dose by the radiation sensitivity of the organs exposed, allowing comparison across different types of examinations. Pediatric patients are more sensitive to radiation-induced carcinogenesis than adults for several reasons: their tissues are actively dividing, their longer life expectancy leaves more time for radiation-induced cancers to manifest, and stochastic effects, meaning probabilistic DNA damage that may lead to malignancy decades later, do not have a known safe threshold in the linear no-threshold framework commonly used for protection purposes. A child with a chronic disease diagnosed in infancy may therefore face decades of potential risk following exposures delivered in the first years of life.

Several disease-specific patterns emerged from the included literature. Children with congenital heart disease, particularly those requiring staged surgical palliation or interventional cardiac catheterization, consistently appeared among the most heavily exposed groups, because cardiac fluoroscopy and CT angiography are central to both diagnosis and treatment. Studies cited in the review estimated cumulative doses during staged single-ventricle palliation and documented measurable chromosomal DNA damage in exposed children. In scoliosis management, repeated spinal radiographs required for curve monitoring, combined with intraoperative imaging, produced substantial cumulative exposure, prompting the development of low-dose slot-scanning systems that reduce dose compared with standard radiographs. Children with inflammatory bowel disease frequently underwent CT during acute flare-ups before magnetic resonance enterography became the preferred alternative, and retrospective cohorts documented cumulative doses high enough to raise malignancy concerns.

Other chronic conditions illustrated subtler but still meaningful exposure pathways. Infants with esophageal atresia undergo repeated contrast studies and fluoroscopic procedures in the first months of life, and one French study cited in the review explicitly asked how low these cumulative doses could realistically be pushed. Children with spina bifida and shunt-treated hydrocephalus accumulated exposure through serial imaging of the brain and spine, while pediatric stone disease generated exposure through fluoroscopy-guided procedures such as percutaneous nephrolithotomy and shockwave lithotripsy. Even conditions considered lower risk, such as developmental dysplasia of the hip, appeared in the literature, with one study reassuringly concluding that repeated pelvic radiographs during harness treatment carry very low radiation risk. Meanwhile, pediatric cleft palate patients showed a three- to five-fold increase in cumulative radiation exposure from dental radiology compared with age- and gender-matched peers.

Organ transplant recipients represent another group highlighted by the review. Children receiving heart transplants accumulated considerable exposure within the first post-transplant year through echocardiography-adjacent imaging, catheterization, and CT surveillance for complications such as rejection, infection, and vascular stenosis. A cohort study of pediatric transplant recipients more broadly documented diagnostic imaging exposure that was markedly elevated compared with healthy children. Similarly, children with osteogenesis imperfecta, the brittle bone disorder, required serial skeletal radiographs throughout childhood to monitor fractures and surgical interventions, with one cited study estimating associated lifetime cancer risk from these cumulative exposures.

What emerges from the synthesis is not a reason for alarm or for avoiding medically necessary imaging, the authors emphasize, but rather a roadmap for safer practice. The review calls for evidence-based referral guidelines specific to pediatric chronic disease populations, so that clinicians weigh the diagnostic yield of each examination against its dose contribution within the context of a child’s total imaging history. It also highlights the importance of standardized imaging protocols optimized for children, including weight- and age-based parameter adjustment, substitution of ultrasound or magnetic resonance imaging where diagnostically equivalent, and the use of dose modulation technologies in CT. Equally critical is dose reporting: recording cumulative effective dose in the patient record so that ordering physicians can see the full picture rather than evaluating each request in isolation. Principles such as ALARA, keeping exposure as low as reasonably achievable, and its extensions emphasizing appropriate use and avoiding unnecessary procedures, are framed as essential operational standards rather than abstract ideals.

The authors also point toward the practical infrastructure needed to make dose stewardship routine. Electronic health record integration of dose-tracking systems, standardized dose metrics across institutions, and education of referring clinicians about the relative doses of different modalities all feature in the review’s recommendations. International collaborative efforts, such as the HARMONIC project cohort studies on radiation exposure in children with congenital heart disease cited within the review, exemplify the kind of multinational, disease-stratified data collection the field needs to quantify risk precisely and track the impact of protection measures over time. The review itself was a literature-based analysis, so no individual patient data were collected, and no ethics approval was required.

For families, the message is one of partnership rather than fear. Parents of children with chronic diseases can and should ask whether each proposed imaging examination is necessary, whether a lower-dose alternative is available, and whether the child’s cumulative imaging history has been considered. For the medical community, the review consolidates more than a decade of evidence into a single argument: the child with a chronic disease is not a series of isolated imaging encounters but a single, longitudinally exposed patient whose total radiation burden deserves active management. As imaging technology continues to advance and dose reduction becomes increasingly feasible, the findings serve as both a benchmark of current exposure levels and a challenge to ensure that the children who depend most on medical imaging are also the best protected from its long-term consequences.

Subject of Research: Cumulative ionizing radiation exposure from medical imaging in pediatric patients with chronic diseases

Article Title: Cumulative ionizing radiation exposure in pediatric patients with chronic diseases: a narrative review

Article References: Eimane, A., Francavilla, M., Grigorjevs, A., Granata, C., Limantoro, I., Olteanu, B.-S., Sofia, C., Nievelstein, R. A., Kardos, M., Kasznia-Brown, J., Salerno, S., & Apine, I. (2026). Cumulative ionizing radiation exposure in pediatric patients with chronic diseases: a narrative review. Pediatric Radiology. https://doi.org/10.1007/s00247-026-06785-x

Image Credits: AI Generated

DOI: 10.1007/s00247-026-06785-x

Keywords: pediatric radiology, cumulative radiation dose, ionizing radiation, CT, fluoroscopy, radiation protection, chronic disease, effective dose, congenital heart disease, scoliosis, medical imaging, ALARA