The cholinergic system is involved in cervical dystonia - An [18F]FEOBV PET study.

This study employed an imaging-decoding strategy to quantitatively characterize intratumoral heterogeneity (ITH) in patients with head and neck squamous cell carcinoma (HNSCC), and further evaluated the capacity of the imaging-based ITH score in prognostic stratification and immunotherapy response prediction. A total of 993 HNSCC patients from three medical centers and one public database were stratified into seven sets. Using an unsupervised radiomics framework integrating local variation and global distribution, tumor regions of interest (ROIs) and volumes of interest (VOIs) were separately analyzed to calculate 2D and 3D ITH scores. The association between the ITH score and patient prognosis was evaluated across independent prognostic sets, and its predictive performance for pathologic complete response (pCR) was assessed in an immunotherapy set. Additionally, histological and molecular characteristics of the ITH score were explored via the pathological set and the genomic set. The ITH score demonstrated robust prognostic value and good predictive performance. Biologically, low-ITH tumors exhibited a higher proportion of inflammatory and connective cells, and were enriched in immune-related pathways, whereas high-ITH tumors exhibited increased heterogeneous tumor cells and upregulation of metabolic pathways. The proposed ITH score represented a reliable, noninvasive, and biologically interpretable imaging biomarker that effectively quantified tumor heterogeneity in HNSCC.

Improving comparability of microCT assessment and reporting of subchondral bone in mouse models of osteoarthritis.

Selective internal radiation therapy (SIRT) increasingly relies on accurate magnetic resonance imaging (MRI) to computed tomography (CT) registration and accurate liver and tumour segmentation, for effective pre-treatment planning. This study evaluates the impact of automatic registration and segmentation techniques on dosimetry calculations for SIRT treatment planning. It compares semi-automatic and automatic registration, as well as manual and automatic deep learning-based segmentation. Pre-treatment data from 90 patients with hepatocellular carcinoma were analysed. The dataset consisted of contrast-enhanced T1-weighted MRI scans with manually delineated liver and tumour volumes of interest (VOIs), as well as single-photon emission computed tomography (SPECT)/CT scans with manually or semi-automatically delineated liver VOI. The clinical routine pipeline, which involves semi-automatic registration and manual/semi-automatic segmentation, was used as the baseline pipeline and compared to experimental pipelines that use intensity-based deformable automatic registration or deep learning-based automatic segmentation. Dosimetric accuracy was assessed via metrics such as the mean absorbed dose, the minimum dose received by 70% of the volume (D70), and inverted cumulative dose-volume histograms. Semi-automatic and automatic MRI to CT liver registration achieved comparable Dice scores of 92%. However, tumour registration varied significantly between registration methods yielding average Dice scores of 79%. Multimodal tumour segmentation approaches outperformed monomodal ones, achieving average Dice scores of 66.6 versus 62.5%. Using the baseline pipeline, the average tumour absorbed dose per patient was 115.6 Gy. Using the fully automatic approach, tumour absorbed doses differed from baseline values by an average of 6.6Gy. Differences ranged from 0.3Gy in the best case to 202.8Gy when the automatic tumour segmentation markedly deviated from the manual delineation. Finally, it was found that a Dice score of at least 80% was required to avoid statistically significant differences in absorbed dose estimates between the clinical and automatic approaches. Semi-automatic and automatic registration show equivalent performance, allowing for complete automation. Automatic segmentation demonstrates promising results, with approximately 40% of patients achieving tumour Dice scores above the 80% threshold. 30% of cases show intermediate performance (60-80% tumour Dice scores), while the remaining 30% are still challenging. Further refinement of segmentation methods is required to enhance dosimetric accuracy.

To investigate the feasibility of non-invasively identifying bone marrow involvement (BMI) in follicular lymphoma (FL) using baseline 18F-FDG PET/CT combined with multidimensional feature fusion, and to compare the impact of different bone marrow volume-of-interest (VOI) frameworks on model performance. This retrospective study included 187 patients with newly diagnosed FL, 93 of whom had BMI. Based on baseline 18F-FDG PET/CT, two bone marrow VOI frameworks were constructed: a pelvic VOI framework and a spine-pelvis VOI framework. Clinical features, conventional imaging features, radiomic features, and deep learning features were extracted. A hierarchical feature screening strategy was employed: clinical and conventional imaging features were screened using univariate logistic regression, Spearman's correlation analysis, and multivariate logistic regression, whereas high-dimensional radiomic and deep learning features were screened using LASSO regression combined with the Boruta algorithm. Based on the selected features, six different modelling schemes were developed. The optimal scheme was selected using the area under the receiver operating characteristic curve (AUC) in the independent validation set as the primary metric. Under the optimal scheme, the performance of seven machine learning models-logistic regression (LR), support vector machine (SVM), gradient boosting machine (GBM), neural network (NN), random forest (RF), k-nearest neighbours (KNN), and adaptive boosting (AdaBoost)-was further compared. SHAP analysis was used to interpret the key features of the final model and the direction of their contributions. Compared with the non-BMI group, the BMI group was more likely to present with widespread regional lymph node involvement, B symptoms, larger lymph node lesions, as well as lower Hb, higher LDH, lower Apo A, lower eGFR, and higher β2-MG levels (all P < 0.05). Under both VOI frameworks, the BMI group exhibited higher bone marrow FDG uptake intensity and metabolic burden, as reflected by higher values of conventional PET/CT features, including SUVmean, Standard Deviation (PET), RMS, 25th Percentile Value, Median, 75th Percentile Value, TLG, Glycolysis Q2-Q4, SAM, and SUVpeak (all P < 0.05). Multivariate logistic regression analysis indicated that regional lymph node involvement and β2-MG consistently remained independent predictors across both VOI frameworks, whereas SUVmean retained statistical significance only within the pelvic VOI framework. A comparison of six modelling schemes revealed that the scheme integrating the spine-pelvis VOI framework with clinical features, conventional imaging features, and radiomic features performed best. Under this scheme, the GBM model achieved the best overall performance on the independent validation set (AUC = 0.906, Accuracy = 0.877, Precision = 0.926, Sensitivity = 0.833, Specificity = 0.926, F1 score = 0.877). SHAP analysis revealed that, in addition to LNr (≥ 5) and β2-MG, first-order statistical features such as PET-Orig-FO-IQR, as well as texture features derived from wavelet/LBP transformations-including PET-Wav-HLL-NGTDM-Strength, PET-Wav-HLL-GLRLM-SRHGLE, CT-LBP3D-m1-GLCM-MCC, and PET-LBP3D-m2-GLSZM-SAHGLE-also made significant contributions. These findings suggest that BMI-associated imaging phenotypes are characterised not only by increased bone marrow metabolism but also by remodelling of the grey-level distribution and spatial heterogeneity within the bone marrow. Bone marrow involvement in follicular lymphoma is associated with higher tumour burden and altered metabolic heterogeneity within the bone marrow. A PET/CT-based radiomic-clinical model showed good performance for non-invasive BMI prediction, and the spine-pelvis VOI framework outperformed the pelvic VOI framework alone. The final GBM model may provide a feasible imaging biomarker for complementary baseline assessment of BMI in FL.

Radiomics models to predict proliferative small hepatocellular carcinoma and its prognosis based on Gd-EOB-DTPA enhanced MRI: A two-center study.

Abstract Myoclonus-dystonia is a hyperkinetic movement disorder and approximately half of myoclonus-dystonia patients have a mutation in the epsilon-sarcoglycan (SGCE) gene, while the remaining cases often have undetermined causative genes. This study aims to assess brain metabolic function in myoclonus-dystonia patients with and without SGCE mutations and compare them to a control group. This study is part of the Next Move in Movement Disorders (NEMO) observational study. We included 23 myoclonus-dystonia patients (11 with SGCE mutations and 12 without) and 23 age-matched controls. Participants underwent 18-Fluordeoxyglucose-PET ([18F]FDG-PET) and anatomical MRI scans. Data were analysed using a voxel-based analysis and a volume of interest (VOI)-based analysis. In the voxel-based analyses, trends towards differences in the supplementary motor area, cingulate gyrus, parietal and occipital lobe were found. When comparing mutation-positive with mutation-negative patients, trends towards differences in the parietal lobe and precentral gyrus were detected. Symptom severity was correlated with changed metabolism in postcentral and supramarginal gyrus, occipital, and frontal lobe, cerebellum, and caudate nucleus. In addition, VOI-based analyses showed statistically significant differences in the supplementary motor area comparing myoclonus-dystonia patients to controls. The identified trends of increased metabolism in the (pre)motor cortex areas fit the model of a more ‘excitable’ state with a lower activation threshold, possibly due to reduced inhibition from the cerebellum and striatum, regions in which we found a negative correlation between symptom severity and metabolism. Differences were also observed in sensory areas such as the parietal lobe and visual cortex. While the phenotype of SGCE mutation-positive and mutation-negative groups is similar, subtle differences suggest distinct endophenotypes.

The 18kDa translocator protein (TSPO) has been a central molecular target for imaging inflammation in the preclinical and clinical research settings across a plethora of applications, including neuroinflammation, cardiovascular inflammation and cancer. Recently, we reported the development of [18F]LW223 as a third-generation TSPO positron emission tomography (PET) radiotracer with binding to human TSPO independent of the rs6971 genetic polymorphism. This study reports the first whole-body human analysis, including biodistribution and dosimetry calculations, following intravenous administration of [18F]LW223. Whole-body PET images were acquired over 250min after intravenous bolus injection of 184.3 ± 20.2 MBq of [18F]LW223 in healthy adult human volunteers. Volumes of interest (VOIs) in different source organs were manually delineated by three independent observers, then time-activity curves were generated for residency times calculations for subsequent quantification of radiation equivalent and effective doses using OLINDA/EXM 2.2 software. The radiotracer biodistribution in humans recapitulated known TSPO expression in various tissues. The main elimination route was found to be hepatobiliary, and the critical organ was the intestine. The cumulated radioactivity excreted by the kidneys was < 10% over the measurement period and no bone uptake suggestive of in vivo defluorination was observed in any of the study subjects. The effective dose ranged between 11.8 ± 0.9 and 12.5 ± 0.9 µSv/MBq. Inter-observer VOI variability had no impact on estimated organ and whole-body effective doses. [18F]LW223 is predominantly excreted by the hepatobiliary route with no evidence of in vivo defluorination but demonstrates marked uptake into tissues with known TSPO expression. It complies with radiation limits and guidelines recommended by regulatory authorities and is in line with previously reported [18F]-labelled radiotracers, such as [18F]fluorodeoxyglucose. [18F]LW223 is suitable for translation into human clinical studies.

BackgroundMycoplasma pneumoniae pneumonia (MPP) is a common cause of pediatric community-acquired pneumonia (CAP). Its computed tomography (CT) manifestations often overlap with those of other pneumonias, making their differentiation challenging. This study aimed to develop and test a radiomics nomogram to distinguish MPP from other pneumonias in children with CAP.MethodsA total of 387 lesions of 325 children with pneumonia were retrospectively analyzed, including 162 MPP and 225 other pneumonias. The radiomic features were extracted from volume of interest (VOI) manually delineated on CT. Through analysis of variance (ANOVA) and least absolute shrinkage and selection operator (LASSO) regression with five-fold cross-validation, optimum radiomic features were screened out. Radiomics signature was calculated through the linear combinations of the screened features. A clinical model was established with clinical independent risk factors selected by univariate and multivariate logistic regression. A radiomics nomogram combining Rad-score and clinical independent risk factors was developed by multivariate logistic regression analysis. Receiver operating characteristic (ROC) curve, decision curve analysis (DCA), and calibration curve were employed to evaluate the performance of the radiomics nomogram.ResultsFourteen radiomic features were selected to calculate the Rad-score. White blood cell (WBC) count and lymphocyte count were independent clinical risk factors. The radiomics nomogram showed good discrimination for the type of pneumonia in the training set with an area under the curve (AUC) of 0.874 [95% confidence interval (CI): 0.836–0.908] and in the testing set with an AUC of 0.841 (95% CI: 0.771–0.893). The DCA and calibration curve demonstrated that the radiomics nomogram had excellent consistency and clinical practicability.ConclusionsThe CT-based radiomics nomogram had good performance in distinguishing MPP from other pneumonias in children with CAP, which can serve as a reference for clinical decision-making.

BackgroundPrevious research has shown that cadmium-zinc-telluride (CZT) crystals, as semiconductor detectors, provide higher spatial and energy resolution than thallium-activated sodium iodide [NaI(Tl)] crystals. This study aimed to evaluate the extent to which a full-ring CZT single-photon emission computed tomography/computed tomography (SPECT/CT) system could improve the image quality of 99mTc-methoxyisobutylisonitrile (99mTc-MIBI) parathyroid imaging, as well as its potential to enhance diagnostic accuracy and optimize imaging protocols.MethodsTwo phantoms were acquired: (I) a uniform phantom to obtain the calibration factor; and (II) a National Electrical Manufacturers Association (NEMA) International Electrotechnical Commission (IEC) phantom containing different-sized spheres to compare the contrast-to-noise ratio (CNR) and recovery coefficients (RCs). A total of 85 patients who were diagnosed with hyperparathyroidism (HPT) were enrolled in this study. Two hours after injecting 99mTc-MIBI, delay-phase parathyroid scanning was performed sequentially on NaI(Tl) SPECT/CT in 8 minutes and full-ring CZT SPECT/CT in 6 minutes. List-mode data were used to reconstruct 3-minute CZT SPECT/CT images. A 10-mm diameter volume of interest (VOI) was used to delineate the lesion and reference regions. The coefficient of variation (CV) and CNR were calculated to quantitatively compare image quality and lesion contrast. The standardized uptake values (SUVs) of the lesions were calculated from the images acquired using the full-ring CZT SPECT/CT. Clinical and pathology information were collected through standardized pre-procedural medical history inquiry and subsequent patient follow-up.ResultsThe NaI(Tl) system exhibited a lower CNR than the full-ring CZT system when using the same injected activity and their respective recommended NEMA IEC phantom parameters. For the images acquired with the full-ring CZT system, those reconstructed using block sequential regularized expectation maximization (BSREM) 20i10s exhibited a higher CNR than those reconstructed using OSEM 4i10s. Compared with OSEM 20i10s, although the difference in the CNR was not statistically significant, the spheres appeared visually clearer, which may be attributed to a more uniform background (CV: OSEM 20i10s, 19.9%; BSREM 20i10s, 17.9%). In the clinical patients, all the three imaging conditions, including 8-minute NaI(Tl), 3-minute CZT, and 6-minute CZT imaging, identified the same positive lesions upon visual assessment. In the image quality comparison, the full-ring CZT system with the standard 6-minute imaging sequence demonstrated better performance than the NaI(Tl) system to varying degrees. Specifically, the CV was lower for the CZT system based on the thyroid (P=0.022) and mediastinal blood pool (P<0.001), but not significantly different based on the cervical muscles (P=0.599). Similarly, the CNR was higher for the CZT system when using the thyroid as the reference region (P=0.033) and mediastinal blood pool as the reference region (P=0.001), but not significantly different when using the cervical muscles as the reference region (P=0.223). The CNR values from the 3-minute CZT imaging were comparable to those from the standard 6-minute CZT imaging (thyroid, P=0.956; cervical muscles, P=0.265; mediastinal blood pool, P=0.182).ConclusionsThe full-ring CZT SPECT/CT system exhibited a higher CNR than the NaI(Tl) crystal-based system in the delay-phase scan of 99mTc-MIBI parathyroid imaging when using the thyroid and mediastinal blood pool as references. Additionally, it enables acquisition within a reduced time frame while maintaining comparable diagnostic sensitivity.

To evaluate the value of integrating habitat radiomics features and deep learning features for predicting occult lymph node metastasis (OLNM) in pancreatic ductal adenocarcinoma (PDAC). Data from 212 eligible PDAC patients across two institutions were analyzed. Cohorts were allocated as follows: training (n = 115), internal validation (n = 50), and external validation (n = 47). Habitat subregion partitioning of the tumor volume of interest (VOI) from portal venous phase computed tomography images was performed using a K-means clustering algorithm, and radiomics features were subsequently extracted. A 2.5D deep learning model based on ResNet18 was used to extract features from the whole VOI. After feature selection, models based on single-feature types and a fusion model integrating habitat radiomics features and deep learning features were developed. Model performance was assessed using receiver operating characteristic curves, decision curve analysis (DCA), and calibration curves. Model interpretability was evaluated via SHapley Additive exPlanations (SHAP). Relative to single-feature-based models, the fusion model achieved superior predictive performance with an area under the curve (AUC) of 0.832 (95% CI: 0.712-0.951) in the external validation cohort. DCA and calibration curves revealed that this model provided greater net clinical benefit compared with other models and demonstrated good calibration. SHAP analysis indicated that deep learning features were the top and third most important predictors. The fusion model exhibited favorable predictive performance for preoperative OLNM diagnosis in PDAC and represents a promising auxiliary tool for personalized therapeutic decision-making.

To develop an interpretable radiomics model based on habitat for predicting lymph node metastasis in oral cancer using Synthetic MRI (SyMRI). A retrospective study of 101 oral cancer patients with pathologically confirmed lymph node status was performed, dividing them into a training cohort (N = 71) and a test cohort (N = 30). All patients underwent MRI, including the MAGiC (SyMRI) sequence. Intra-tumoral volumes of interest (VOIs) were clustered into three spatial habitats using the K-means algorithm. Radiomics models were developed for the entire tumor and clustered intra-tumor regions, based on features from synthetic quantitative T1 and T2 maps. The Shapley Additive Explanations (SHAP) method was used to interpret predictions. A radiomics nomogram was constructed by integrating independent variables, including the radiomics score and MRI-reported status. Predictive performance was evaluated using AUC, DeLong's test, calibration curves, and decision curve analysis (DCA). The habitat with the highest T1 and a moderately low T2 values within the tumor showed the best predictive performance, compared to whole-tumor radiomics (training: 0.87 vs. 0.82; test: 0.74 vs. 0.62). The radiomics nomogram, combining radiomics features and independent clinical variables, outperformed clinical diagnosis alone in both the training cohort (0.92 vs. 0.80; p = 0.0023) and test cohort (0.84 vs. 0.73; p = 0.0362). The habitat-based radiomics signature, integrating SyMRI-derived quantitative T1 and T2 maps, provides enhanced predictive accuracy for preoperative lymph node metastasis in oral cancer compared to clinical diagnosis.

Partial volume effect (PVE) arises from the limited spatial resolution of positron emission tomography (PET) scanners, causing significant quantitative biases that hinder accurate metabolic activity assessment. To address these problems, we proposed an unsupervised deep residual compensation model (U-DRCM) for PET partial volume correction (PVC). U-DRCM first predicted an initial blur kernel for the PVE-affected PET image based on a conditional blind deconvolution module (CBD module). Then, a conditional residual compensation module (CRC module) was introduced to compensate for the error caused by inaccurate blur kernel prediction. The whole model is unsupervised which only needs a single patient's PET image as the training label and the corresponding MR image as the network input. The performance of U-DRCM was evaluated against several established PVC approaches, including Richardson-Lucy (RL), reblurred Van-Cittert (RVC), iterative Yang (IY), neural blind deconvolution (NBD), and deep convolutional neural network (DeepPVC) using both simulated BrainWeb phantom and real clinical datasets. In the simulation study, U-DRCM consistently outperformed competing methods across multiple quantitative metrics, achieved a higher peak signal-to-noise ratio (PSNR), an improved structural similarity index (SSIM), and a lower root mean square error (RMSE). For the real clinical study, U-DRCM delivered substantial improvements in standardized uptake value (SUV) and standardized uptake value ratio (SUVR) across various brain volumes of interest (VOIs). Experimental results show that U-DRCM effectively mitigates the impact of PVE, resulting in high-quality PVC PET images with enhanced brain visualization.

According to the Global Burden of Disease Study 2019, there were approximately 0.9 million new cases of spinal cord injury worldwide. Injury to the spinal cord can lead to significant and often permanent loss of sensory and motor functions. The impairment of sensory and motor functions is a consequence of cellular and molecular events triggered by the injury, resulting in secondary complications. Inflammation and demyelination are two of the primary pathological processes that occur after SCI. Research suggests that these secondary complications are exacerbated in the aged population. This study aimed to assess neuroinflammation and demyelination in a rat model of SCI, comparing young and aged rats using non-invasive positron emission tomography/computed tomography (Positron Emission Tomography (PET)/CT) imaging. Young (3 months) and middle-aged (12 months) male Sprague-Dawley rats were imaged dynamically using inflammation ([18F]DPA714) and demyelination (3[18F]F4AP) PET tracers prior to injury and acutely after a moderate contusion T9 SCI. The tracer uptake was assessed by drawing a volume of interest (VOI), and the mean Standardized Uptake Value (SUVmean) was compared from baseline to post-injury time point for the two radiotracers. Alterations in the tracer SUVmean were also evaluated between the aged and young animals. Kinetic PET scans demonstrated that both injury and age altered the uptake patterns for demyelination and inflammation PET tracers. Compared to young animals, the aged animals showed increased tracer uptake at the injury site for the inflammation ([18F]DPA714) marker only. No change in tracer uptake was observed in the uninjured regions distal to the injury site or baseline scans between age groups. Combined PET scans with histological analyses demonstrated that [18F]DPA714 significantly correlated with gliosis, whereas 3[18F]F4AP correlates with neuronal and white matter markers. The PET/CT imaging using these tracers has the potential to offer valuable insights into prognosis and treatment effectiveness following SCI.

Positron emission tomography (PET) using fibroblast activation protein inhibitors (FAPI) has emerged as a robust imaging tool for solid tumours. The novel ligand [68Ga]BED003 (formerly known as [68Ga]Ga-OncoFAP-DOTAGA) has demonstrated very high affinity for the fibroblast activation protein and favourable biodistribution. This study aimed to assess the in vivo distribution of [68Ga]BED003 across a spectrum of solid tumours as a potential prerequisite for radioligand therapy. In this retrospective analysis, [68Ga]BED003 PET/CT or PET/MRI of 157 patients with 19 different solid malignancies were retrospectively analysed. A spherical volume of interest (VOI) was placed over the primary tumour, the most intense lymph node and distant metastases that were evaluated as at least likely malignant in the written reports, and the maximum standardized uptake value (SUVmax) was derived. Liver and blood pool background mean SUV (SUVmean) were assessed using standardised VOIs. Tumour-to-background ratios (TBRmax) were calculated as the ratio of SUVmax to background SUVmean. SUVmax and TBRmax values of primary tumours, lymph node, and distant metastases were compared using Wilcoxon rank-sum and signed-rank tests, as appropriate. Overall, 115 primary tumours, 70 lymph node metastases, and 116 distant metastases were analysed, yielding median SUVmax of 16.6, 12.3, and 11.9, respectively. No significant difference in SUVmax or TBRmax were observed between lymph node and distant metastases (all P > 0.05). In contrast, primary tumours demonstrated significantly higher uptake than lymph node and distant metastases (all P < 0.001), except for liver TBRmax when comparing primary tumours with lymph node metastases (P > 0.05). Across all 301 lesions, the median SUVmax, liver TBRmax and blood pool TBRmax was 14.7 (range, 4.2–35.9), 21.6 (range, 4.2–62.8) and 11.1 (range, 3.0–33.4), respectively. The highest median SUVmax and TBRmax (both) in cohorts with > 5 lesions were found in medullary thyroid, oesophageal, ovarian, cervical, breast, colorectal, and hepatocellular cancers. [68Ga]BED003-PET demonstrated consistently high uptake across diverse solid malignancies, supporting its potential role as a tool for multi-cancer diagnostic imaging and patient selection for FAP-targeted radioligand therapy.

To develop and validate interpretable models integrating standardized uptake value (SUV), radiomics (Rad), and deep learning (DL) features from 18F-FDG PET/CT for differentiating diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL). This retrospective study included 250 patients from two centers. Volumes of interest (VOIs) were delineated on PET images for SUV, Rad, and DL features extraction. Feature selection was performed using the Mann–Whitney U test, random forest–based recursive feature elimination, and the least absolute shrinkage and selection operator (LASSO). Seven machine learning classifiers were applied to construct diagnostic models, and fused Rad and DL features were further integrated to construct deep learning radiomics (DLR) models. Model interpretability was assessed using SHapley Additive exPlanations (SHAP). Model performance was evaluated in terms of discrimination, calibration, and clinical applicability. The DLR model achieved the best diagnostic performance, with an area under the curve (AUC) of 0.905 and an accuracy of 0.813 in the testing cohort. SHAP analysis identified the Rad feature “original_Maximum” as the most influential predictor for differentiating DLBCL from FL. Calibration curve and decision curve analyses further supported the superiority of the DLR model. Rad and DL features derived from 18F-FDG PET/CT enable effective differentiation between DLBCL and FL. The proposed SHAP-based interpretable model offers superior diagnostic accuracy and potential clinical utility.

Organic anion-transporting polypeptide 2B1 (OATP2B1/SLCO2B1) is an uptake transporter expressed in the liver and in several extrahepatic tissues, including skeletal muscle. Muscular OATP2B1 is thought to facilitate intracellular accumulation of statins, potentially contributing to statin-induced myotoxicity. To investigate functional OATP2B1 activity in vivo in human skeletal muscle, we performed positron emission tomography (PET) with the radiolabelled OATP2B1 substrate [11C]erlotinib. Nine healthy male volunteers (age: 31 ± 9 years) underwent two dynamic 60-min PET scans of the head with concurrent arterial blood sampling following intravenous injection of a microdose of [11C]erlotinib (< 10 µg). The first scan was performed without any pharmacological pre-treatment (baseline scan), whereas the second scan was performed after pre-treatment with a single oral dose of unlabelled erlotinib (650 mg), administered 3.0 ± 0.1 h before the start of the PET scan. Volumes of interest (VOIs) were manually delineated for the right and left temporal muscle surrounding the skull on co-registered PET/magnetic resonance imaging (MRI) data and averaged to generate a global temporal muscle VOI. Time-activity curves for temporal muscle and arterial plasma were analysed using a 1-tissue-2-rate constant (1T2K) compartment model and Logan graphical analysis to estimate the total volume of distribution (VT) of [11C]erlotinib (reflecting the steady-state muscle-to-plasma concentration ratio), as well as the rate constants for transfer of [11C]erlotinib from plasma to muscle (K1) and from muscle to plasma (k2). Both Logan analysis and the 1T2K model demonstrated a significant reduction in VT after erlotinib pre-treatment compared with baseline (VT Logan: baseline: 0.85 ± 0.11 mL/cm3, erlotinib: 0.70 ± 0.08 mL/cm3, -18 ± 8%, p = 0.00047; VT 1T2K: baseline: 0.83 ± 0.11 mL/cm3, erlotinib: 0.67 ± 0.07 mL/cm3, -18 ± 7%, p = 0.00033). K1 showed a trend toward reduction after erlotinib pre-treatment without reaching statistical significance, whereas k2 remained unchanged. Our findings demonstrate saturable distribution of [¹¹C]erlotinib to human skeletal muscle, consistent with functional OATP2B1 activity. These results support a mechanistic role for muscular OATP2B1 in statin-induced myotoxicity and highlight its potential broader relevance for the safety and pharmacology of other OATP2B1 substrate drugs.

Isolated lung cancer (ILC), isolated tuberculoma, and coexistence of tuberculosis with lung cancer (CTBLC) exhibit similarities in computed tomography (CT) imaging features but great differences in pathology, treatment strategy, and prognosis; therefore, accurate differential diagnosis is critical for clinical management and patient safety. The purpose of this study was to develop and validate a habitat-based CT radiomics model that integrates intralesional subregion features with whole-lesion features for reliable differentiation among these three conditions. This study retrospectively included 317 patients with ILC, tuberculoma, or CTBLC from 2018 to 2022. Among these, 239 patients from Beijing Chest Hospital, Capital Medical University (Center 1) formed the training and internal test cohorts, and 78 from Infectious Disease Hospital of Heilongjiang Province (Center 2) constituted an external validation cohort. Volumes of interest (VOIs) were manually outlined by two experienced radiologists on CT images. Then each lesion was partitioned into two subregions using K-means clustering. A total of 1,218 three-dimensional whole-lesion radiomics features and 2,436 habitat features were extracted. Feature selection was performed via least absolute shrinkage and selection operator (LASSO). Six classification algorithms were trained and evaluated. To distinguish ILC, tuberculoma, and CTBLC, three models were developed: (I) a traditional radiomics model using only whole-lesion radiomics features; (II) a habitat model based on intralesional habitat features; and (III) a combined habitat-radiomics model fusing both feature sets. Discrimination was assessed using the area under the curve (AUC), and SHapley Additive exPlanations (SHAP) was used to interpret the optimal model and visualize individual prediction decisions. The combined habitat-radiomics model that integrates habitat and whole-lesion features outperformed the traditional radiomics model. Among them, the extreme gradient boosting (XGBoost)-based fusion model achieved the best performance (mean AUC =0.934) in the internal test cohort, surpassing both the radiomics model (mean AUC =0.910) and the habitat model (mean AUC =0.873). For individual classes, the fusion model yielded AUCs of 0.911 (ILC), 0.955 (tuberculoma), and 0.937 (CTBLC). Compared with the interpretations provided by three radiologists, the combined radiomics-habitat model demonstrated better discriminative performance. SHAP plots revealed key features and presented individual visualizations of each prediction. A habitat-based CT radiomics approach that incorporates intralesional subregion features into whole-lesion signatures improves differentiation among ILC, tuberculoma, and CTBLC. This combined model provides a noninvasive tool to support clinical decision-making.

Traumatic cervical spinal cord injury (TCSCI) often leads to significant patient paralysis. Current clinical diagnosis relies heavily on empirical interpretation of magnetic resonance imaging (MRI) and the American Spinal Injury Association Impairment Scale (AIS) grade, lacking robust quantitative markers to precisely reflect injury severity. This study aimed to build an artificial intelligence (AI) pipeline for AIS grade prediction based on radiomic features extracted from manually defined regions. We included 189 patients with TCSCI who underwent MRI within 48 h post-injury. MRI images from 130 patients were used for developing an AI model encompassing image segmentation. Radiomic features were extracted from manually delineated volumes of interest (VOIs). T2-weighted imaging (T2WI) sagittal images were randomly divided into training (n = 104), validation (n = 13), and test (n = 13) sets for segmentation. A total of 183 patients (excluding AIS E) were included in the AIS grade prediction task. Model performance was evaluated using mean dice similarity coefficient (mDICE), mean intersection over union (mIOU), mean specificity, and mean sensitivity. An optimized UCTransnet network, leveraging a Transformer architecture for formal training, combined with a U-Net++ network for pretraining, achieved promising results in segmenting the spinal cord injury site on T2WI sagittal images (mDICE: 0.777 ± 0.021, mIOU: 0.646 ± 0.025, mean specificity: 0.998 ± 0.001, mean sensitivity: 0.895 ± 0.015). Subsequently, an ensemble model (we named Em-En) constructed using selected radiomic features from the manual VOIs demonstrated superior performance for predicting AIS grades in terms of sensitivity, specificity, accuracy, and clinical decision-making benefit compared to other tested models. This study presents an AI-assisted pipeline for predicting the severity of TCSCI. The developed resources provide a theoretical foundation for the clinical application of AI-assisted diagnostic methods, potentially lowering the interpretation barrier for MRI and offering clinicians preliminary quantitative indicators of injury severity. The source code is publicly available.

BACKGROUNDMetallic artifacts from coaxial needles can severely interfere with the precision of cone-beam computed tomography (CBCT)-guided percutaneous transthoracic needle biopsy (PTNB), particularly in assessing vital anatomical structures around small lesions.AIMTo evaluate the clinical application of a streak metal artifact reduction technique (SMART) in CBCT-PTNB procedures.METHODSWe retrospectively analyzed data from 68 patients (73 CBCT scans) undergoing CBCT-guided PTNB between March 2023 and December 2024. Image quality was compared among original reconstructed CBCT images, those iteratively reconstructed using SMART with a full volume of interest (VOI) (SMART-Full), and those with a small VOI containing only the coaxial needle (SMART-Small). Evaluations focused on artifact types, puncture needle diameter measurements, and density metrics within the region of interest (ROI).RESULTSSMART-Full reconstruction significantly reduced radial, streak, and dark stripe artifacts (P < 0.001) compared to original CBCT, with superior performance in puncture needle diameter measurement and ROI minimum and average density indicators (P < 0.001). The incidence of dark streak artifacts decreased from 71 cases in original CBCT to 26 cases in SMART-Full. Additionally, SMART-Full was more effective than SMART-Small in artifact elimination (P < 0.001).CONCLUSIONSMART technology effectively reduces metal artifacts, enabling clearer visualization of hidden anatomical structures. Through quantitative analysis, this study confirms the clinical value of SMART in CBCT-guided PTNB, providing a technical reference for precise diagnosis and treatment of small pulmonary lesions.