BMC Medical Imaging最新文献

Background: Artificial intelligence (AI) has shown increasing potential in lung cancer imaging, particularly in detection, staging, prognosis, and recurrence prediction. However, there is limited synthesis of head-to-head comparative evidence between CT, FDG PET/CT, and multimodal fusion models within the same cohorts.

Objectives: To systematically review and critically appraise studies that directly compared CT-only, FDG PET/CT-only, and combined multimodal models in lung cancer, with emphasis on clinical setting, fusion strategy, validation design, and clinical utility.

Methods: This systematic review followed PRISMA 2020 and PRISMA-S guidelines. PubMed, Scopus, IEEE Xplore, and Google Scholar were searched for English-language human studies published between January 1, 2019, and September 8, 2025. Eligible studies reported same-cohort, head-to-head comparisons of CT, PET/CT, or multimodal models for lung cancer screening, staging, or prognosis. Risk of bias was assessed using PROBAST for prediction model studies and SANRA for narrative reviews. Data were extracted in duplicate and synthesized narratively, with meta-analysis performed where ≥ 3 studies were sufficiently homogeneous.

Results: From 2,417 records (PubMed 845, Scopus 920, IEEE Xplore 452, Google Scholar/manual 200), 31 studies met inclusion criteria (20 primary modeling studies, 11 reviews). In screening cohorts, low-dose CT deep-learning models consistently outperformed other modalities, with modest incremental value from clinical covariates. For nodal staging, integrated PET/CT radiomics-clinical models showed superior discrimination, calibration, and net-benefit compared with unimodal approaches. In prognostic and recurrence settings, fused PET/CT models outperformed CT- or PET-only models across institutions, with further improvement from clinical variables. Radiogenomics and pathology integration provided added value but were limited by small samples and lack of external validation.

Conclusions: Comparative evidence demonstrates that modality performance is context-dependent: CT dominates in screening, PET/CT fusion excels in staging and prognosis, and multimodal integration with clinical or biomarker data enhances discrimination and utility. Standardization, harmonization, and rigorous external validation remain critical for generalizability.

Clinical trial number: Not applicable.

Background: Patients undergoing pancreaticoduodenectomy for distal cholangiocarcinoma (dCCA) face a substantial risk of major postoperative cardiac complications (MPCC), which significantly impact mortality and recovery. Existing risk assessment tools lack objective cardiac functional parameters. This study aimed to develop and validate a novel prediction model integrating preoperative cardiac ultrasound parameters to individually predict the risk of MPCC within 30 days after dCCA surgery.

Methods: A retrospective cohort study was conducted on 154 dCCA patients who underwent radical pancreaticoduodenectomy. Univariate and multivariate binary logistic regression analyses were performed to identify independent predictors of MPCC from clinical variables and preoperative transthoracic echocardiography parameters. A nomogram model was constructed based on the identified independent predictors. The model's discrimination, calibration, and clinical utility were assessed using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA), with internal validation via bootstrapping.

Results: The incidence of MPCC was 34.4% (53/154). Multivariate analysis identified preoperative B-type natriuretic peptide (BNP), left ventricular ejection fraction (LVEF), left ventricular mass index (LVMI), and left atrial volume index (LAVI) as independent predictors. A nomogram incorporating these four factors was developed. The model demonstrated excellent discrimination, with an AUC of 0.894 (95% CI: 0.838-0.950). Calibration curves showed good agreement between predicted and observed probabilities. DCA confirmed the model's clinical net benefit across a wide range of threshold probabilities.

Conclusion: This study presents a robust nomogram that effectively integrates cardiac ultrasound parameters (LVEF, LVMI, LAVI) and BNP to preoperatively predict the risk of cardiac complications following dCCA surgery. The model offers superior individualized risk stratification compared to traditional tools, potentially facilitating optimized perioperative management for high-risk patients.

Background: This study investigated the functional connectivity (FC) patterns of cerebellar components of seven intrinsic connectivity networks (ICNs) across different stages of Alzheimer's disease (AD).

Methods: FC between each cerebellar seed region corresponding to one of the seven ICNs and 91 cerebral regions of interest (ROI) corresponding to the cortical parcels defined by Harvard-Oxford atlas was calculated for individuals with clinically diagnosed probable AD dementia (n = 21), mild cognitive impairment (n = 34), and subjective cognitive decline (n = 33). Group differences were assessed using ANOVA with false discovery rate (FDR) correction for multiple ROIs (p_FDR-corr<0.05).

Results: Significant alterations were observed in FC between the frontoparietal network (FPN) and the left superior frontal gyrus (SFG), as well as between the limbic network (LN) and the right superior lateral occipital cortex (sLOC) and temporo-occipital middle temporal gyrus (toMTG). Specifically, FPN-SFG connectivity decreased at the dementia stage, while LN-toMTG and LN-sLOC connectivity decreased during the prodromal stage but increased in the dementia stage.

Conclusions: These results indicate the presence of both decreases and increases in cerebellar-cortical FC across different stages of AD. A detailed examination of cerebellar involvement, an aspect often underexplored in AD research, may be crucial for understanding the neural mechanisms underlying disease progression.

Aim: Thromboangiitis Obliterans (TAO), or Buerger's disease, affects peripheral vessels and is linked to smoking. This study using diffusion tensor imaging (DTI) tractography examines brain microstructural changes in TAO patients compared to healthy smokers and Normal controls, aiming to reveal neurological implications beyond the known peripheral effects.

Methods: The case-control study involved 50 participants aged 27-56 in northeastern Iran, including TAO patients, healthy smokers, and non-smoking controls. MRI scans with DTI were conducted to assess 21 brain tracts for fractional anisotropy (FA) and apparent diffusion coefficient (ADC).

Results: Significant differences in brain tract integrity were observed among the groups. TAO patients showed lower FA values in the minor forceps compared to healthy smokers, while smokers had higher FA values than non-smoking controls. ADC values were notably higher in TAO patients across several tracts, including the minor and major forceps, corticospinal tracts, fornix tracts, and arcuate fasciculus, compared to both healthy smokers and controls.

Conclusion: The study highlights distinct brain tract alterations in TAO patients and suggests potential neurological consequences associated with the disease and smoking habits. DTI proves valuable in understanding microstructural brain changes and could serve as a diagnostic tool for evaluating smoking-related neurologic complications, providing insights into TAO's impact beyond peripheral vessels.

Clinical trial registration: N/A.

Objective: To develop and validate an automated, disc-level deep learning pipeline for quantitative measurement of anteroposterior (AP) thecal sac diameter on mid-sagittal lumbar T2-weighted MRI.

Materials and methods: In this retrospective study, 511 mid-sagittal lumbar T2 MRI examinations were included after screening 758 cases and applying predefined exclusions. The workflow combined YOLOv8 oriented bounding boxes (OBB) for disc-level localization and orientation estimation, homography-based ROI warping, Attention U-Net segmentation, and skeleton-based AP diameter computation in millimeters using DICOM pixel spacing. Validation was performed on internal (50) and external (50; RSNA 2024 lumbar dataset) cohorts with two radiologists providing the reference standard.

Result: Inter-reader agreement was excellent (ICC (2, 1) = 0.967; 711 paired measurements). Against the reader-mean reference, the pipeline achieved an overall MAE of 0.994 mm (711 disc-level measurements). Internal validation showed MAE 0.909 mm (357 measurements) and external validation MAE 1.079 mm (354 measurements). Severity-wise MAE remained ~ 1 mm (mild 0.930 mm; moderate 1.234 mm; severe 1.038 mm). Automatic disc-level labeling was performed, and OBB-derived orientation significantly improved AP measurement-line validity versus axis-aligned detection (acceptable lines 99.02% vs. 77.64%).

Conclusion: An orientation-aware YOLOv8-OBB + Attention U-Net pipeline enables automated, disc-level AP thecal sac diameter quantification on mid-sagittal lumbar MRI with ~ 1 mm error relative to expert reference, supporting standardized morphometric reporting and measurement-driven assessment of lumbar stenosis.