BMC Medical Imaging最新文献

{"title":"Radiomic study of common sellar region lesions differentiation in magnetic resonance imaging based on multi-classification machine learning model.","authors":"Hang Qu, Qiqi Ban, LiangXue Zhou, HaiHan Duan, Wei Wang, AiJun Peng","doi":"10.1186/s12880-025-01690-5","DOIUrl":"https://doi.org/10.1186/s12880-025-01690-5","url":null,"abstract":"<p><strong>Objective: </strong>Pituitary adenomas (PAs), craniopharyngiomas (CRs), Rathke's cleft cysts (RCCs), and tuberculum sellar meningiomas (TSMs) are common sellar region lesions with similar imaging characteristics, making differential diagnosis challenging. This study aims to develop and evaluate machine learning models using MRI-based radiomics features to differentiate these lesions.</p><p><strong>Methods: </strong>Two hundred and fifty-eight pathologically diagnosed sellar region lesions, including 54 TSMs, 81 CRs, 61 RCCs and 63 PAs, were retrospectively studied. All patients underwent conventional MR examinations. Feature extraction and data normalization and balance were performed. Extreme gradient boosting (XGBoost), support vector machine (SVM), and logistic regression (LR) models were trained with the radiomics features. Five-fold cross-validation was used to evaluate model performance.</p><p><strong>Results: </strong>The XGBoost model showed better performance than the SVM and LR models built from contrast-enhanced T1-weighted MRI features (balanced accuracy 0.83, 0.77, 0.75; AUC 0.956, 0.938, 0.929, respectively). Additionally, these models demonstrated significant differences in sensitivity (P = 0.032) and specificity (P = 0.045). The performance of the XGBoost model was superior to that of the SVM and LR models in differentiating sellar region lesions by using contrast-enhanced T1-weighted MRI features.</p><p><strong>Conclusion: </strong>The proposed model has the potential to improve the diagnostic accuracy in differentiating sellar region lesions.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"147"},"PeriodicalIF":2.9,"publicationDate":"2025-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12049783/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143961430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The application of non-enhanced magnetic resonance thoracic ductography combined with magnetic resonance abdominopelvic scanning in the diagnosis of chylous leakage of the female reproductive system.","authors":"Meng Huo, Chunyan Zhang, Ling Li, Jianfeng Xin, Xingpeng Li, Yimeng Zhang, Mingxia Zhang, Ying Sun, Lei Sun, Rengui Wang, Yunlong Yue","doi":"10.1186/s12880-025-01689-y","DOIUrl":"https://doi.org/10.1186/s12880-025-01689-y","url":null,"abstract":"<p><strong>Objective: </strong>To explore the value of non-enhanced magnetic resonance thoracic ductography (NMRTD) combined with MR abdominopelvic scanning in the diagnosis of chylous leakage of the female reproductive system.</p><p><strong>Methods: </strong>A retrospective analysis was conducted on the multimodal imaging data from non-enhanced magnetic resonance thoracic ductography (NMRTD), direct lymphangiography (DLG), and abdominopelvic magnetic resonance imaging (MRI) for 18 female patients with reproductive system chylous leakage. Among these patients, 7 had vaginal chylous leakage, 10 had vulvar cutaneous chylous leakage, and 1 had both conditions.The rate of successful visualization of the thoracic duct, the consistency of the drainage directions of the outlet of the thoracic duct, and the degree of visualization of each segment of the thoracic duct by NMRTD and DLG were analyzed. A retrospective analysis was performed on the abnormal manifestations of abdominopelvic MR.</p><p><strong>Results: </strong>NMRTD had a significant advantage over DLG in terms of successful visualization of the thoracic duct (94.4% vs. 66.7%, P = 0.035). The display of the drainage directions in the outlet area of the thoracic duct by the two methods showed excellent consistency (kappa value = 0.815) in 12 patients whose outlet areas were visualized by both methods. The degrees of visualization of the upper, middle, and lower segments of the thoracic duct in the NMRTD group were significantly greater than those in the DLG group (P values were 0.02, 0.00 and 0.00, respectively). All 18 patients (100%) showed dilatation of the lymph vessels in the pelvic cavity and retroperitoneum on abdominopelvic MR and DLG as well as pelvic perineal reflux or leakage on DLG. MR revealed multiple-site abnormalities that could not be detected by DLG, including multiple long T1 and long T2 lesions of the spleen in 8 patients (44.4%), of the subcutaneous in 7 patients (38.9%), of the bone in 6 patients (33.3%), perineal lymphedema in 18 patients (100%), and abdominopelvic effusion in 10 patients (55.6%).</p><p><strong>Conclusion: </strong>NMRTD combined with abdominopelvic MR has advantages in comprehensively evaluating the thoracic duct and multiple systemic abnormalities in patients with chylous leakage of the female reproductive system.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"146"},"PeriodicalIF":2.9,"publicationDate":"2025-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12049049/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143963319","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A deep learning algorithm for automated adrenal gland segmentation on non-contrast CT images.","authors":"Fanxing Meng, Tuo Zhang, Yukun Pan, Xiaojing Kan, Yuwei Xia, Mengyuan Xu, Jin Cai, Fangbin Liu, Yinghui Ge","doi":"10.1186/s12880-025-01682-5","DOIUrl":"https://doi.org/10.1186/s12880-025-01682-5","url":null,"abstract":"<p><strong>Background: </strong>The adrenal glands are small retroperitoneal organs, few reference standards exist for adrenal CT measurements in clinical practice. This study aims to develop a deep learning (DL) model for automated adrenal gland segmentation on non-contrast CT images, and to conduct a preliminary large-scale study on age-related volume changes in normal adrenal glands using the model output values.</p><p><strong>Methods: </strong>The model was trained and evaluated on a development dataset of annotated non-contrast CT scans of bilateral adrenal glands, utilizing nnU-Net for segmentation task. The ground truth was manually established by two experienced radiologists, and the model performance was assessed using the Dice similarity coefficient (DSC). Additionally, five radiologists provided annotations on a subset of 20 randomly selected cases to measure inter-observer variability. Following validation, the model was applied to a large-scale normal adrenal glands dataset to segment adrenal glands.</p><p><strong>Results: </strong>The DL model development dataset contained 1301 CT examinations. In the test set, the median DSC scores for the segmentation model of left and right adrenal glands were 0.899 and 0.904 respectively, and in the independent test set were 0.900 and 0.896. Inter-observer DSC for radiologist manual segmentation did not differ from automated machine segmentation (P = 0.541). The large-scale normal adrenal glands dataset contained 2000 CT examinations, the graph shows that adrenal gland volume increases first and then decreases with age.</p><p><strong>Conclusion: </strong>The developed DL model demonstrates accurate adrenal gland segmentation, and enables a comprehensive study of age-related adrenal gland volume variations.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"142"},"PeriodicalIF":2.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12046700/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143977862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Functional connectivity across multi-frequency bands in patients with tension-type headache: a resting-state fMRI retrospective study.","authors":"Jili Wang, Hongjie Shen, Qinyan Xu, Shuxian Zhang, Tian Li, Yun Zheng","doi":"10.1186/s12880-025-01599-z","DOIUrl":"https://doi.org/10.1186/s12880-025-01599-z","url":null,"abstract":"<p><strong>Objectives: </strong>Tension-type headache (TTH) is the most common nervous system disorder worldwide. This study aimed to examine abnormal network-level brain functional connectivity (FC) alterations in patients with TTH across multi-frequency bands.</p><p><strong>Methods: </strong>The study enrolled 63 subjects, comprising 32 patients with TTH and 31 healthy controls (HC). According to our team's previous research, the brain regions with abnormal ReHo in the conventional frequency band (0.01-0.08 Hz) and the slow-5 band (0.01-0.027 Hz) were chosen as seed regions of interest (ROIs). Subsequently, the FC between ROIs and the entire brain analysis across various frequency bands was calculated to evaluate network-level alterations, and differences between the TTH and HC were analyzed. Pearson's correlation analysis was conducted to assess the relationship between significantly altered FC values in two frequency bands and visual analog score (VAS) in TTH patients.</p><p><strong>Results: </strong>In the slow-5 band (0.01-0.027 Hz), FC between right medial superior frontal gyrus and right medial temporal pole/right inferior temporal gyrus as well as right middle frontal gyrus and left supramarginal gyrus of TTH patients exhibited significantly higher, compared to the HC group, while FC between right middle frontal gyrus and right lateral occipital cortex reduced. For the correlation results, there was no correlation between abnormal brain regions of FC and VAS score.</p><p><strong>Conclusions: </strong>Changes in FC within brain regions associated with TTH are linked to pain processing. And the altered FC in TTH patients were frequency dependent. These initial observations could enhance our understanding of TTH's pathophysiological mechanism and offer insights for its future diagnosis and treatment.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"145"},"PeriodicalIF":2.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12046950/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143967883","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Preoperative prediction of microvascular invasion and relapse-free survival in hepatocellular Carcinoma ≥3 cm using CT radiomics: Development and external validation.","authors":"Hua Zhong, Yan Zhang, Guanbin Zhu, Xiaoli Zheng, Jinan Wang, Jianghe Kang, Ziying Lin, Xin Yue","doi":"10.1186/s12880-025-01677-2","DOIUrl":"https://doi.org/10.1186/s12880-025-01677-2","url":null,"abstract":"<p><strong>Objective: </strong>To preoperatively predict microvascular invasion (MVI) and relapse-free survival (RFS) in hepatocellular carcinoma (HCC) ≥3 cm by constructing and externally validating a combined radiomics model using preoperative enhanced CT images.</p><p><strong>Methods: </strong>This retrospective study recruited adults who underwent surgical resection between September 2016 and August 2020 in our hospital with pathologic confirmation of HCC ≥3 cm and MVI status. For external validation, adults who underwent surgical resection between September 2020 and August 2021 in our hospital were included. Histopathology was the reference standard. The HCC area was segmented on the arterial and portal venous phase CT images to develop a CT radiomics model. A combined model was developed using selected radiomics features, demographic information, laboratory index and radiological features. Analysis of variance and support vector machine were used as features selector and classifier. Receiver operating characteristic (ROC) curves, calibration curves and decision curve analysis (DCA) were used to evaluate models' performance. The Kaplan-Meier method and log-rank test were used to evaluate the predictive value for RFS.</p><p><strong>Results: </strong>A total of 202 patients were finally enrolled (median age, 59 years, 173 male). Thirteen and 24 features were selected for the CT radiomics model and the combined model, and the area under the ROC curves (AUC) were 0.752 (95 %CI 0.615, 0.889) and 0.890 (95 %CI 0.794, 0.985) in the external validation set, respectively. Calibration curves and DCA showed a higher net clinical benefit of the combined model. The high-risk group (P < 0.001) was an independent predictor for RFS.</p><p><strong>Conclusions: </strong>The combined model showed high accuracy for preoperatively predicting MVI and RFS in HCC ≥3 cm.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"141"},"PeriodicalIF":2.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12046733/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143970011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and assessment of the AE-RADS standardized grid for specifically evaluating adverse events in diagnostic radiology and teleradiology.","authors":"Jean-François Bergerot, Amandine Crombé, Mylène Seux, Basile Porta, Vanessa Fyon, Samuel Le Nivet, Nicolas Lippa, Rémi Peyre, Paul Etchart, Frédérique Gay, Guillaume Gorincour","doi":"10.1186/s12880-025-01670-9","DOIUrl":"https://doi.org/10.1186/s12880-025-01670-9","url":null,"abstract":"<p><strong>Background: </strong>A specific grid for analyzing and grading adverse events in diagnostic radiology is lacking. In France, the standard HAS grid, a generic 5-point scale adapted from the Common Terminology Criteria for Adverse Events (CTCAEs), is criticized for limited applicability in radiology. Our aim was to develop and evaluate a radiology-specific AE grid (AE-RADS) tailored to diagnostic and teleradiological practices and to compare its performance against the CTCAEs-based HAS grid regarding inter-observer reproducibility and agreement with expert consensus.</p><p><strong>Methods: </strong>AE-RADS, structured as a decision tree with 90 items, was developed by four senior radiologists with extensive AE experience. To assess it, 100 AE cases from early 2022 were reviewed by two radiologists and two non-physician support members, all blinded to the initial AE grading. Observers rated AEs using both the HAS and AE-RADS grids, comparing severity, AE frequency per patient, sources, and types for inter-observer reproducibility and expert agreement. Tests included intra-class correlation coefficient (ICC), Fleiss Kappa and Krippendorff alpha for reproducibility and McNemar test for comparing agreement with consensus.</p><p><strong>Results: </strong>Among 100 patients (49 women, median age 66.9 years), 104 AEs were identified. AE-RADS achieved higher inter-observer reproducibility for AE frequency (ICC = 0.690 vs. 0.642 with HAS) and for grading the most serious AE (Krippendorff alpha = 0.519 vs. 0.506 with HAS). Agreement with expert consensus was significantly greater with AE-RADS (63-81%) than with HAS (25-47%; P-value range: 0.0001-0.0051).</p><p><strong>Conclusion: </strong>AE-RADS shows improved, though still imperfect, agreement between evaluators and experts, supporting its potential for more precise AE assessment in diagnostic imaging.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"143"},"PeriodicalIF":2.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12046918/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143961426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artifact estimation network for MR images: effectiveness of batch normalization and dropout layers.","authors":"Tomoko Maruyama, Norio Hayashi, Yusuke Sato, Toshihiro Ogura, Masumi Uehara, Haruyuki Watanabe, Yoshihiro Kitoh","doi":"10.1186/s12880-025-01663-8","DOIUrl":"https://doi.org/10.1186/s12880-025-01663-8","url":null,"abstract":"<p><strong>Background: </strong>Magnetic resonance imaging (MRI) is an essential tool for medical diagnosis. However, artifacts may degrade images obtained through MRI, especially owing to patient movement. Existing methods that mitigate the artifact problem are subject to limitations including extended scan times. Deep learning architectures, such as U-Net, may be able to address these limitations. Optimizing deep learning networks with batch normalization (BN) and dropout layers enhances their convergence and accuracy. However, the influence of this strategy on U-Net has not been explored for artifact removal.</p><p><strong>Methods: </strong>This study developed a U-Net-based regression network for the removal of motion artifacts and investigated the impact of combining BN and dropout layers as a strategy for this purpose. A Transformer-based network from a previous study was also adopted for comparison. In total, 1200 images (with and without motion artifacts) were used to train and test three variations of U-Net.</p><p><strong>Results: </strong>The evaluation results demonstrated a significant improvement in network accuracy when BN and dropout layers were implemented. The peak signal-to-noise ratio of the reconstructed images was approximately doubled and the structural similarity index was improved by approximately 10% compared with those of the artifact images.</p><p><strong>Conclusions: </strong>Although this study was limited to phantom images, the same strategy may be applied to more complex tasks, such as those directed at improving the quality of MR and CT images. We conclude that the accuracy of motion artifact removal can be improved by integrating BN and dropout layers into a U-Net-based network, with due consideration of the correct location and dropout rate.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"144"},"PeriodicalIF":2.9,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12046729/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144061926","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of mental foramen and accessory mental foramen using cone beam computed tomography in a Turkish population.","authors":"Dilek Coban, Zerrin Unal Erzurumlu, Elif Sadik, Yasin Yasa","doi":"10.1186/s12880-025-01589-1","DOIUrl":"https://doi.org/10.1186/s12880-025-01589-1","url":null,"abstract":"<p><strong>Background: </strong>The aim is to assess mental foramen (MF) and anatomical variations using cone beam computed tomography (CBCT) images in a Turkish population.</p><p><strong>Methods: </strong>In this retrospective study, CBCT images of 301 patients (162 females, 139 males) obtained between November 2021 and February 2022 were evaluated. Patients were analyzed in 4 groups according to age (Group 1: 18-30 years, Group 2: 31-45 years, Group 3: 46-55 years, Group 4: 56 years and older). The position of the MF relative to the teeth, vertical (MFV) and horizontal (MFH) dimensions; the distances of the MF to the mandibular midline (MF-MM), ramus posterior border (MF-MP), lower border (MF-ML) and upper border (MF-MU); the presence of an accessory mental foramen (AMF); and if any, the position of the AMF relative to the MF and the distance of the AMF to the MF (MF-AMF) were recorded separately for the right and left sides. Associations with gender and age were evaluated. The independent samples t test was used to determine the relationship between the measurements and gender and the evaluation of the measurement values according to age groups. The evaluation of the location of MF according to gender and age groups was performed using the Chi-Square Test.</p><p><strong>Results: </strong>On the right and left sides, MF was most commonly seen at the apical level of the second premolars (45.4% and 52.1%, respectively). MFV, MFH, MF-MM, MF-MP and MF-ML were significantly higher in males than in females, p < 0.001 on both sides and for each parameter. There was a significant difference between the age groups for MF-MU and MF-ML on the right side, MF-MU on the left side (p < 0.001, p < 0.05, p < 0.001, respectively). A total of 42 AMFs were seen in 39 (13%) of the 301 patients. AMFs were frequently located posteroinferior to the MFs (35.5%). The mean MF-AMF was 7.83 mm.</p><p><strong>Conclusions: </strong>The findings of this study contribute to the existing literature on the anatomy and variations of MF. The results of this study show that the prevalence of AMF in the Turkish population studied is high at 13%.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"140"},"PeriodicalIF":2.9,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12044973/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143963489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of deep learning reconstruction combined with time-resolved post-processing method to improve image quality in CTA derived from low-dose cerebral CT perfusion data.","authors":"Jiajing Tong, Tong Su, Yu Chen, Xiaobo Zhang, Ming Yao, Yanling Wang, Haozhe Liu, Min Xu, Jian Wang, Zhengyu Jin","doi":"10.1186/s12880-025-01623-2","DOIUrl":"https://doi.org/10.1186/s12880-025-01623-2","url":null,"abstract":"<p><strong>Background: </strong>To assess the effect of the combination of deep learning reconstruction (DLR) and time-resolved maximum intensity projection (tMIP) or time-resolved average (tAve) post-processing method on image quality of CTA derived from low-dose cerebral CTP.</p><p><strong>Methods: </strong>Thirty patients underwent regular dose CTP (Group A) and other thirty with low-dose (Group B) were retrospectively enrolled. Group A were reconstructed with hybrid iterative reconstruction (R-HIR). In Group B, four image datasets of CTA were gained: L-HIR, L-DLR, L-DLR_tMIP and L-DLR_tAve. The CT attenuation, image noise, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR) and subjective images quality were calculated and compared. The Intraclass Correlation (ICC) between CTA and MRA of two subgroups were calculated.</p><p><strong>Results: </strong>The low-dose group achieved reduction of radiation dose by 33% in single peak arterial phase and 18% in total compared to the regular dose group (single phase: 0.12 mSv vs 0.18 mSv; total: 1.91mSv vs 2.33mSv). The L-DLR_tMIP demonstrated higher CT values in vessels compared to R-HIR (all P < 0.05). The CNR of vessels in L-HIR were statistically inferior to R-HIR (all P < 0.001). There was no significant different in image noise and CNR of vessels between L-DLR and R-HIR (all P > 0.05, except P = 0.05 for CNR of ICAs, 77.19 ± 21.64 vs 73.54 ± 37.03). However, the L-DLR_tMIP and L-DLR_tAve presented lower image noise, higher CNR (all P < 0.05) and subjective scores (all P < 0.001) in vessels than R-HIR. The diagnostic accuracy in Group B was excellent (ICC = 0.944).</p><p><strong>Conclusion: </strong>Combining DLR with tMIP or tAve allows for reduction in radiation dose by about 33% in single peak arterial phase and 18% in total in CTP scanning, while further improving image quality of CTA derived from CTP data when compared to HIR.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"139"},"PeriodicalIF":2.9,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12042446/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143962139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"¹⁸F-FDG PET/CT-based deep learning models and a clinical-metabolic nomogram for predicting high-grade patterns in lung adenocarcinoma.","authors":"Yue Guo, Xibin Jia, Chuanxu Yang, Chao Fan, Hui Zhu, Xu Chen, Fugeng Liu","doi":"10.1186/s12880-025-01684-3","DOIUrl":"https://doi.org/10.1186/s12880-025-01684-3","url":null,"abstract":"<p><strong>Background: </strong>To develop and validate deep learning (DL) and traditional clinical-metabolic (CM) models based on 18 F-FDG PET/CT images for noninvasively predicting high-grade patterns (HGPs) of invasive lung adenocarcinoma (LUAD).</p><p><strong>Methods: </strong>A total of 303 patients with invasive LUAD were enrolled in this retrospective study; these patients were randomly divided into training, validation and test sets at a ratio of 7:1:2. DL models were trained and optimized on PET, CT and PET/CT fusion images, respectively. CM model was built from clinical and PET/CT metabolic parameters via backwards stepwise logistic regression and visualized via a nomogram. The prediction performance of the models was evaluated mainly by the area under the curve (AUC). We also compared the AUCs of different models for the test set.</p><p><strong>Results: </strong>CM model was established upon clinical stage (OR: 7.30; 95% CI: 2.46-26.37), cytokeratin 19 fragment 21 - 1 (CYFRA 21-1, OR: 1.18; 95% CI: 0.96-1.57), mean standardized uptake value (SUVmean, OR: 1.31; 95% CI: 1.17-1.49), total lesion glycolysis (TLG, OR: 0.994; 95% CI: 0.990-1.000) and size (OR: 1.37; 95% CI: 0.95-2.02). Both the DL and CM models exhibited good prediction efficacy in the three cohorts, with AUCs ranging from 0.817 to 0.977. For the test set, the highest AUC was yielded by the CT-DL model (0.895), followed by the PET/CT-DL model (0.882), CM model (0.879) and PET-DL model (0.817), but no significant difference was revealed between any two models.</p><p><strong>Conclusions: </strong>Deep learning and clinical-metabolic models based on the ¹⁸F-FDG PET/CT model could effectively identify LUAD patients with HGP. These models could aid in treatment planning and precision medicine.</p><p><strong>Clinical trial number: </strong>Not applicable.</p>","PeriodicalId":9020,"journal":{"name":"BMC Medical Imaging","volume":"25 1","pages":"138"},"PeriodicalIF":2.9,"publicationDate":"2025-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12036234/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143953158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}