The study, led by Amira Benattia of Université Paris Cité and the French National Reference Centre for Histiocytoses at Hôpital Saint-Louis in Paris, was published in BMC Medical Imaging as a prospective pilot investigation. Eighteen patients with cystic pulmonary Langerhans cell histiocytosis underwent paired inspiratory and expiratory CT scans alongside a battery of pulmonary function tests performed at the same visit. The researchers then ran the imaging data through three complementary quantitative analysis pipelines: conventional densitometry, parametric response mapping, and a vascular analysis tool known as CALIPER. The results, they report, demonstrate for the first time that these automated biomarkers are not only feasible in this rare disease but agree closely with expert visual scoring.

Pulmonary Langerhans cell histiocytosis is an uncommon interstitial lung disease, seen almost exclusively in smokers, in which abnormal immune cells accumulate in and around the small airways. Over time, the inflammation gives way to a distinctive pattern of nodules, cavities, and eventually thin-walled cysts scattered throughout the lungs. Because the disease is rare, large clinical trials are difficult to mount, and clinicians have lacked validated, objective tools to measure how much lung tissue is affected or how the damage progresses over time. Visual scoring systems exist, but they are inherently reader-dependent: two experienced radiologists looking at the same scan may grade the extent of disease differently, and the same reader may grade it differently on different days.

The Paris team attacked this problem with parametric response mapping, a technique originally developed for chronic obstructive pulmonary disease. The method works by co-registering inspiratory and expiratory CT scans of the same patient, voxel by voxel, and comparing lung density between the two breathing phases. Each tiny cube of lung tissue is then classified into one of three categories: normal density, persistent low-density areas that represent structurally destroyed tissue such as cysts and emphysema, and functional small airways disease, where air is trapped because the small bronchioles are narrowed or obstructed. In a disease like pulmonary Langerhans cell histiocytosis, where cystic destruction and small airway involvement coexist in the same lungs, this three-way classification promised something clinicians had never had: a way to separate the two processes quantitatively.

The findings were striking. Automated densitometry and parametric response mapping measurements of the cystic lesions correlated with expert visual scoring at correlation coefficients of at least 0.80 across the board, and the agreement between the automated and human assessments was excellent, with an intraclass correlation coefficient of 0.96 and a 95 percent confidence interval of 0.91 to 0.99. Just as importantly, the automated metrics tracked with physiology. The extent of persistent low-density tissue on CT correlated significantly with the diffusing capacity of the lungs for carbon monoxide, a standard pulmonary function measure that reflects how efficiently oxygen and carbon dioxide cross the delicate membrane between air and blood. In statistical terms, the numbers on the scan were telling the same story as the numbers on the breathing tests.

Perhaps the most intriguing result emerged when the researchers adjusted their analysis for smoking history, measured in pack-years. Because nearly all patients with this disease are smokers, and smoking itself causes emphysema and small airway damage, disentangling the effects of the histiocytosis from the effects of tobacco is a persistent challenge. In multivariate linear regression models adjusted for pack-years, the persistent low-density fraction measured by parametric response mapping remained significantly associated with all pulmonary function test parameters, independent of the functional small airways disease fraction. This suggests that the cystic destruction characteristic of the disease carries measurable functional consequences of its own, beyond whatever damage smoking has inflicted on the small airways.

The small airways story proved more complicated. The functional small airways disease values varied considerably among patients who had similar forced expiratory volume in one second percentages, a standard measure of airflow obstruction. The authors interpret this variability as evidence of genuine heterogeneity in small airway involvement across the patient population, rather than as a failure of the technique. In other words, two patients with the same overall airflow limitation may arrive there by different routes, one dominated by cystic destruction and another by widespread bronchiolar obstruction. If confirmed in larger cohorts, this kind of quantitative phenotyping could eventually help clinicians understand why some patients deteriorate rapidly while others remain stable for years, and could help stratify patients in future therapeutic trials.

The study also ventured beyond the airways and into the blood vessels. Using CALIPER, or Computer-Aided Lung Informatics for Pathology Evaluation and Rating, the team quantified pulmonary vascular volume, a surrogate measure of the blood vessel bed within the lungs. Vascular involvement is an increasingly recognized dimension of the disease, and the analysis revealed a significant inverse association between the regional extent of cystic destruction and pulmonary vascular volume, with a p-value of 0.002. Where the cysts were most extensive, the vascular tree was most depleted, a finding that fits with the understanding that the disease process destroys not only airspaces but also the capillary networks that supply them, potentially contributing to pulmonary hypertension, a feared complication of advanced disease.

None of this means automated imaging is ready to replace the radiologist tomorrow. The study was a pilot, involving eighteen patients at a single expert centre, and the authors are careful to frame their conclusions as proof of feasibility rather than validated clinical practice. The correlation and agreement statistics are encouraging, but the confidence intervals reflect a small sample, and the technique requires high-quality paired inspiratory-expiratory scans, careful co-registration, and standardized acquisition protocols that are not yet universal. The researchers themselves call for further evaluation of the biomarkers’ potential value for disease phenotyping and longitudinal assessment, the two applications where quantitative imaging could ultimately matter most.

Still, the significance of the work extends well beyond one rare disease. Quantitative CT biomarkers have transformed the study of common conditions like emphysema and idiopathic pulmonary fibrosis, enabling objective endpoints in drug trials and deeper understanding of disease subtypes. Rare diseases have largely been left out of that revolution, precisely because small patient populations make validation difficult. By demonstrating that parametric response mapping, densitometry, and vascular quantification can be applied successfully to pulmonary Langerhans cell histiocytosis, the Paris group has laid the methodological groundwork for the kind of rigorous, measurable, multi-centre research that rare disease patients have long been denied. For a condition where every patient counts, teaching machines to read the lungs with precision may prove to be more than a technical achievement; it may be the first step toward trials that can actually change outcomes.

Subject of Research: Automated quantitative CT imaging biomarkers for assessing pulmonary Langerhans cell histiocytosis

Article Title: Feasibility of automated quantitative CT imaging metrics in pulmonary Langerhans cell histiocytosis: a pilot study

Article References: Feasibility of automated quantitative CT imaging metrics in pulmonary Langerhans cell histiocytosis: a pilot study. (n.d.). https://doi.org/10.1186/s12880-026-02932-w

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02932-w

Keywords: pulmonary Langerhans cell histiocytosis, quantitative CT, parametric response mapping, densitometry, CALIPER, cystic lung disease, lung function, imaging biomarkers, pulmonary vascular volume, small airways disease, rare disease, radiology

The study, registered retrospectively on ClinicalTrials.gov as NCT07162168, enrolled 4,152 participants who had undergone noncontrast routine CT scans. The cohort included 1,947 men, with a median age of 59 years spanning a remarkable range from 18 to 99. Rather than relying on slow, labor-intensive manual tracing of the femur, the team trained a deep learning segmentation model based on the nnU-Net framework, a self-configuring architecture widely regarded as one of the most robust tools for biomedical image segmentation. A subset of proximal femurs was first annotated by hand to teach the model exactly where the bone begins and ends in each volumetric scan. Once trained, the automated system could carve out the femoral volume of interest on its own, achieving a Dice similarity coefficient of 0.963 on the test set, a score indicating near-perfect overlap with expert annotations.

With the femur reliably isolated in thousands of scans, the researchers turned to radiomics, the practice of converting medical images into hundreds of quantitative features that describe texture, shape, intensity and spatial patterns within a defined region. Using the open-source PyRadiomics package, they extracted features from each segmented femur, following standardized conventions designed to make radiomic measurements reproducible across scanners and institutions. The idea behind radiomics is deceptively simple: the microscopic architecture of bone, including trabecular thinning, cortical porosity and the redistribution of mineral, leaves fingerprints in the statistical texture of a CT image. Those fingerprints, aggregated across thousands of features, may encode a person’s skeletal age far more richly than a single density measurement.

To test whether these features actually carry information about aging and sex, the team built machine learning models in a development cohort of 2,969 participants and then evaluated them in a completely separate external validation cohort of 1,183 participants, a design that guards against the optimistic results that plague many single-center AI studies. For age, the best performer was a support vector regression model, which estimated chronological age from femoral radiomics with a coefficient of determination of 0.50 in external validation, a mean absolute error of 8.84 years and a root mean squared error of 11.32 years. In plain terms, the model could roughly halve the uncertainty of a blind guess about a person’s age using nothing but the texture of their hip bone.

Sex classification proved even more striking. Two algorithms, support vector classification and the gradient-boosting method XGBoost, both achieved an area under the receiver operating characteristic curve of 0.94 in external validation, with support vector classification showing slightly higher accuracy and F1-score. An AUC of 0.94 means the model distinguished male from female skeletons with high reliability across the entire cohort, capturing structural sexual dimorphism in the femur that goes well beyond simple size differences. Because the features were extracted automatically from routine scans, the result hints that sex-related skeletal architecture is legible in ordinary clinical imaging without any dedicated bone protocol.

Perhaps the most consequential finding concerns the relationship between the radiomics-derived age index and actual bone mineral biology. In the external validation cohort, the researchers compared their radiomic age index against volumetric bone mineral density of the femoral neck measured by quantitative computed tomography, the gold-standard three-dimensional density technique. The radiomics index tracked vBMD more strongly than chronological age did, with a Spearman correlation of minus 0.546 versus minus 0.443 for age alone, a difference the authors confirmed statistically using Steiger’s test (Z = 4.97, P < 0.001). This comparison matters because it suggests the radiomic signature is not merely a proxy for how many birthdays a person has had, but a reflection of genuine structural deterioration in the bone, the kind of change that underlies osteoporosis and hip fracture risk.

The clinical logic here is what researchers call opportunistic imaging. Dual-energy X-ray absorptiometry, the standard test for osteoporosis, is underused: many people at risk never get scanned, and fractures often arrive as the first sign of disease. But a large fraction of older adults undergo abdominal, chest or pelvic CT for unrelated reasons, and those scans routinely capture the proximal femur. If an automated pipeline can convert that incidental imaging into a skeletal age index or a bone quality flag, clinicians could identify patients with accelerated bone aging who would benefit from confirmatory density testing or preventive treatment, all without a single additional scan, extra radiation dose or added visit. The authors caution that their radiomic patterns still need validation against hard clinical outcomes such as incident fractures before such screening becomes practice, but the biological anchoring to vBMD is an encouraging first step.

The technical achievement should not be undersold either. Automated segmentation at a Dice coefficient above 0.96 removes the bottleneck that has historically limited radiomics to small, single-center studies requiring hours of expert contouring. By combining nnU-Net segmentation, standardized PyRadiomics extraction and rigorous external validation across nationwide cohorts, the study offers a template for how imaging biomarkers can be scaled to population level. The work was supported by the Beijing Natural Science Foundation-Fengtai Joint Innovation Project, the National Natural Science Foundation of China Science Center Program and the Shenzhen Medical Research Fund, and it was approved by the ethics committee of Peking University People’s Hospital with informed consent waived for the retrospective use of de-identified images.

There are, of course, caveats to keep in mind. The cohort was drawn from Chinese hospitals, and radiomic features can be sensitive to scanner vendor, reconstruction settings and dose, so multicenter generalization beyond this dataset remains to be demonstrated. The age model’s mean absolute error of nearly nine years is impressive for image-only prediction but far too coarse to serve as a precise biological clock for any individual. And the study measured associations with density, not with fractures, falls or mortality, so the ultimate clinical value hinges on longitudinal follow-up. Still, the direction of travel is clear: the femur that appears incidentally on a routine scan carries a quantifiable structural biography, written in texture patterns that machines can now read at scale.

If subsequent studies confirm that radiomics-derived bone age predicts fractures and responds to osteoporosis treatment, the implications could extend well beyond orthopedics. Bone is a living archive of hormonal change, nutrition, inflammation and mechanical loading, and a cheap, automated readout of its microarchitecture could inform research on aging itself, from sarcopenia to metabolic bone disease. For now, the study stands as a proof of concept that the skeleton’s quiet remodeling, normally invisible in the reading room, can be decoded from data that hospitals already possess, waiting in the archive.

Subject of Research: Automated CT radiomics of the proximal femur for quantifying age- and sex-related bone structural changes

Article Title: Automated proximal femur CT radiomics identifies age- and sex-related radiomic patterns of bone structure: a nationwide multicohort validation study

Article References: Cheng, H., Zhuang, L., Zhang, Y., Ran, Y., Jiang, B., & Kou, Y. (2026). Automated proximal femur CT radiomics identifies age- and sex-related radiomic patterns of bone structure: a nationwide multicohort validation study. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02836-9

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02836-9

Keywords: radiomics, proximal femur, CT imaging, deep learning, nnU-Net, bone aging, bone mineral density, osteoporosis, machine learning, opportunistic screening, quantitative CT, hip fracture risk