Brain age has emerged as a powerful tool to understand neuroanatomical aging and its link to health outcomes like cognition. However, there remains a lack of studies investigating the rate of brain aging and its relationship to cognition. Furthermore, most brain age models are trained and tested on cross-sectional data from primarily Caucasian, adult participants. It is thus unclear how well these models generalize to non-Caucasian participants, especially children. Here, we tested a previously published deep learning model on Singaporean elderly participants (55-88 years old) and children (4-11 years old). We found that the model directly generalized to the elderly participants, but model finetuning was necessary for children. After finetuning, we found that the rate of change in brain age gap was associated with future executive function performance in both elderly participants and children. We further found that lateral ventricles and frontal areas contributed to brain age prediction in elderly participants, while white matter and posterior brain regions were more important in predicting brain age of children. Taken together, our results suggest that there is potential for generalizing brain age models to diverse populations. Moreover, the longitudinal change in brain age gap reflects developing and aging processes in the brain, relating to future cognitive function.

We conducted a prospective study to evaluate the usefulness of ultralow-dose computed tomography (ULD-CT) with deep-learning reconstruction (DLR) compared with conventional standard-dose CT (SD-CT) for post-endovascular aneurysm repair (EVAR) surveillance. We prospectively performed post-EVAR surveillance using ULD-CT at a single center in 44 patients after they had received SD-CT. The ULD-CT images underwent DLR, whereas the SD-CT images underwent iterative reconstruction. Three radiologists blinded to the patient information and CT conditions independently measured the aneurysmal sac diameter and evaluated the overall image quality. Bland-Altman analysis and a linear mixed-effects model were used to assess and compare the measurement accuracy between SD-CT and ULD-CT. The mean CT dose index volume and dose-length product were significantly lower for ULD-CT (1.0 ± 0.3 mGy and 71.4 ± 26.5 mGy•cm) than that for SD-CT (6.9 ± 0.9 mGy and 500.9 ± 96.0 mGy•cm; p<0.001). The mean short diameters of the aneurysmal sac measured by the 3 observers were 46.7 ± 10.8 mm on SD-CT and 46.3 ± 10.8 mm on ULD-CT. The mean difference in the short diameter of the aneurysmal sac between ULD-CT and SD-CT was -0.37 mm (95% confidence interval, -0.6 to -0.12 mm). The intraobserver limits of agreement (LOA) for measurements by ULD-CT and SD-CT were -3.5 to 2.6, -2.8 to 1.9, and -2.9 to 2.3 for Observers 1, 2, and 3, respectively. The pairwise LOAs for assessing interobserver agreement, such as for the differences between Observers 1 and 2 measurements in SD-CT, were mostly within the predetermined acceptable range. The mean image-quality score was lower for ULD-CT (3.3 ± 0.6) than that for SD-CT (4.5 ± 0.5; p<0.001). Aneurysmal sac diameter measurements by ULD-CT with DLR were sufficiently accurate for post-EVAR surveillance, with substantial radiation reduction versus SD-CT.Clinical ImpactDeep-learning reconstruction (DLR) is implemented as a software-based algorithm rather than requiring dedicated hardware. As such, it is expected to be integrated into standard computed tomography (CT) systems in the near future. The ultralow-dose CT (ULD-CT) with DLR evaluated in this study has the potential to become widely accessible across various institutions. This advancement could substantially reduce radiation exposure in post-endovascular aneurysm repair (EVAR) CT imaging, thereby facilitating its adoption as a standard modality for post-EVAR surveillance.

As cardiovascular disease (CVD) is the leading cause of noncancer mortality in colorectal or gastric cancer patients, it is essential to identify patients at increased CVD risk. Coronary artery calcium (CAC) is an established predictor of atherosclerotic CVD; however, its application is limited in this population. This study evaluates the association between automated CAC scoring using chest computed tomography and atherosclerotic CVD risk in colorectal or gastric cancer patients. A retrospective cohort study was conducted using electronic health records linked to claims data of colorectal or gastric cancer patients who underwent non-ECG-gated chest computed tomography at 2 tertiary hospitals in South Korea between 2011 and 2019. CAC was automatically quantified using deep learning software and used to classify patients into 4 groups (CAC=0, 0<CAC≤100, 100<CAC≤400, CAC>400). The primary outcome was major adverse cardiovascular events (myocardial infarction, stroke, or cardiovascular mortality), and assessed using the multivariable Fine and Gray subdistribution hazard model. A meta-analysis was performed to calculate pooled subdistribution hazard ratios. A total of 3153 patients were included in this study (36.5% female; 36.3% CAC=0; 38.1% 0<CAC≤100; 14.1% 100<CAC≤400; 11.5% CAC>400). The mean follow-up period was 4.1 years. The incidence rate of MACE was 5.28, 8.03, 9.99, and 29.14 per 1000 person-years in CAC=0, 0<CAC≤100, 100<CAC≤400, and CAC>400. Compared with CAC=0, the risk of MACE was not significantly different in patients with 0<CAC≤100 (subdistribution hazard ratio, 1.43 [95% CI, 0.41-5.01]), and 100<CAC≤400 (subdistribution hazard ratio, 0.99 [95% CI, 0.48-2.04]). Patients with CAC>400 had 2.33 (95% CI, 1.24-4.39) times higher risk of MACE compared with those with CAC=0. CAC>400 was associated with an increased risk of MACE compared with CAC=0 among colorectal or gastric cancer patients. CAC quantified on routine chest computed tomography scans provides prognostic information for atherosclerotic CVD risk in this population.

The number of patients referred for and requiring a transthoracic echocardiogram (TTE) has increased over the years resulting in more cardiac sonographers reporting work related musculoskeletal pain. We sought to determine if a scanning protocol that replaced conventional workflows with advanced technologies such as multiplane imaging, artificial intelligence (AI) and automation could be used to optimise conventional workflows and potentially reduce ergonomic risk for cardiac sonographers. The aim was to assess whether this alternate protocol could reduce active scanning time as well as interaction with the ultrasound machine compared to a standard echocardiogram without a reduction in image quality and interpretability. Volunteer participants were recruited for a study that comprised of two TTE's with separate protocols. Both were clinically complete, but Protocol A combined automation, AI assisted acquisition and measurement, simultaneous and multiplane imaging whilst Protocol B reflected a standard scanning protocol without these additional technologies. Keystrokes were significantly reduced with the advanced protocol as compared to the typical protocol (230.9 ± 24.2 vs. 502.8 ± 56.2; difference 271.9 ± 61.3, p < 0.001). Furthermore, there was a reduction in scan time with protocol A compared to protocol B the standard TTE protocol (13.4 ± 2.3 min vs. 18.0 ± 2.6 min; difference 4.6 ± 2.9 min, p < 0.001) as well as a decrease of approximately 27% in the time the sonographers were required to reach beyond a neutral position on the ultrasound console. A TTE protocol that embraces modern technologies such as AI, automation, and multiplane imaging shows potential for a reduction in ultrasound keystrokes and scan time without a reduction in quality and interpretability. This may aid a reduction in ergonomic workload as compared to a standard TTE.

Prostate-specific membrane antigen (PSMA) is an important target for positron emission tomography (PET) with computed tomography (CT) in prostate cancer. In addition to overexpression in prostate cancer cells, PSMA is expressed in healthy cells in the proximal tubules of the kidneys. Consequently, PSMA PET is being explored for renal functional imaging. Left and right renal uptake of PSMA targeted radiopharmaceuticals have shown strong correlations to split renal function (SRF) as determined by other methods. Manual segmentation of kidneys in PET images is, however, time consuming, making this method of measuring SRF impractical. In this study, we designed, trained and validated an artificial intelligence (AI) model for automatic renal segmentation and measurement of SRF in [¹⁸F]PSMA-1007 PET images. Kidneys were segmented in 135 [¹⁸F]PSMA-1007 PET/CT studies used to train the AI model. The model was evaluated in 40 test studies. Left renal function percentage (LRF%) measurements ranged from 40 to 67%. Spearman correlation coefficients for LRF% measurements ranged between 0.98 and 0.99 when comparing segmentations made by 3 human readers and the AI model. The largest LRF% difference between any measurements in a single case was 3 percentage points. The AI model produced measurements similar to those of human readers. Automatic measurement of SRF in PSMA PET is feasible. A potential use could be to provide additional data in investigation of renal functional impairment in patients treated for prostate cancer.

With the advancement of computer technology and imaging equipment, ultrasound has emerged as a crucial tool in breast cancer diagnosis. To gain deeper insights into the research landscape of ultrasound in breast cancer diagnosis, this study employed bibliometric methods for a comprehensive analysis spanning from 2004 to 2024, analyzing 3523 articles from 2176 institutions in 82 countries/regions. Over this period, publications on ultrasound diagnosis of breast cancer showed a fluctuating growth trend from 2004 to 2024. Notably, China, Seoul National University and Kim EK emerged as leading contributors in ultrasound for breast cancer detection, with the most published and cited journals being Ultrasound Med Biol and Radiology. The research spots in this area included "breast lesion", "dense breast" and "breast-conserving surgery", while "machine learning", "ultrasonic imaging", "convolutional neural network", "case report", "pathological complete response", "deep learning", "artificial intelligence" and "classification" are anticipated to become future research frontiers. This groundbreaking bibliometric analysis and visualization of ultrasonic breast cancer diagnosis publications offer clinical medical professionals a reliable research focus and direction.

Carotid web (CaW) is a risk factor for ischemic stroke, mainly in young patients with stroke of undetermined etiology. Its detection is challenging, especially among non-experienced physicians. We included patients with CaW from six international trials and registries of patients with acute ischemic stroke. Identification and manual segmentations of CaW were performed by three trained radiologists. We designed a two-stage segmentation strategy based on a convolutional neural network (CNN). At the first stage, the two carotid arteries were segmented using a U-shaped CNN. At the second stage, the segmentation of the CaW was first confined to the vicinity of the carotid arteries. Then, the carotid bifurcation region was localized by the proposed carotid bifurcation localization algorithm followed by another U-shaped CNN. A volume threshold based on the derived CaW manual segmentation statistics was then used to determine whether or not CaW was present. We included 58 patients (median (IQR) age 59 (50-75) years, 60% women). The Dice similarity coefficient and 95th percentile Hausdorff distance between manually segmented CaW and the algorithm segmented CaW were 63.20±19.03% and 1.19±0.9 mm, respectively. Using a volume threshold of 5 mm³, binary classification detection metrics for CaW on a single artery were as follows: accuracy: 92.2% (95% CI 87.93% to 96.55%), precision: 94.83% (95% CI 88.68% to 100.00%), sensitivity: 90.16% (95% CI 82.16% to 96.97%), specificity: 94.55% (95% CI 88.0% to 100.0%), F1 measure: 0.9244 (95% CI 0.8679 to 0.9692), area under the curve: 0.9235 (95%CI 0.8726 to 0.9688). The proposed two-stage method enables reliable segmentation and detection of CaW from head and neck CT angiography.

Synthesizing high quality CT images remains a signifi-cant challenge due to the limited availability of annotat-ed data and the complex nature of CT imaging. In this work, we present PRO, a novel framework that, to the best of our knowledge, is the first to perform CT image synthesis in the projection domain using latent diffusion models. Unlike previous approaches that operate in the image domain, PRO learns rich structural representa-tions from raw projection data and leverages anatomi-cal text prompts for controllable synthesis. This projec-tion domain strategy enables more faithful modeling of underlying imaging physics and anatomical structures. Moreover, PRO functions as a foundation model, capa-ble of generalizing across diverse downstream tasks by adjusting its generative behavior via prompt inputs. Experimental results demonstrated that incorporating our synthesized data significantly improves perfor-mance across multiple downstream tasks, including low-dose and sparse-view reconstruction, even with limited training data. These findings underscore the versatility and scalability of PRO in data generation for various CT applications. These results highlight the potential of projection domain synthesis as a powerful tool for data augmentation and robust CT imaging. Our source code is publicly available at: https://github.com/yqx7150/PRO.

Dental diagnosis relies on two primary imaging modalities: panoramic radiographs (PX) providing 2D oral cavity representations, and Cone-Beam Computed Tomography (CBCT) offering detailed 3D anatomical information. While PX images are cost-effective and accessible, their lack of depth information limits diagnostic accuracy. CBCT addresses this but presents drawbacks including higher costs, increased radiation exposure, and limited accessibility. Existing reconstruction models further complicate the process by requiring CBCT flattening or prior dental arch information, often unavailable clinically. We introduce ViT-NeBLa, a vision transformer-based Neural Beer-Lambert model enabling accurate 3D reconstruction directly from single PX. Our key innovations include: (1) enhancing the NeBLa framework with Vision Transformers for improved reconstruction capabilities without requiring CBCT flattening or prior dental arch information, (2) implementing a novel horseshoe-shaped point sampling strategy with non-intersecting rays that eliminates intermediate density aggregation required by existing models due to intersecting rays, reducing sampling point computations by $52 \%$, (3) replacing CNN-based U-Net with a hybrid ViT-CNN architecture for superior global and local feature extraction, and (4) implementing learnable hash positional encoding for better higher-dimensional representation of 3D sample points compared to existing Fourier-based dense positional encoding. Experiments demonstrate that ViT-NeBLa significantly outperforms prior state-of-the-art methods both quantitatively and qualitatively, offering a cost-effective, radiation-efficient alternative for enhanced dental diagnostics.

Multimodal medical imaging integrates diverse data types, such as structural and functional neuroimaging, to provide complementary insights that enhance deep learning predictions and improve outcomes. This study focuses on a neuroimaging prediction framework based on both structural and functional neuroimaging data. We propose a next-generation prediction model, \textbf{MultiViT2}, which combines a pretrained representative learning base model with a vision transformer backbone for prediction output. Additionally, we developed a data augmentation module based on the latent diffusion model that enriches input data by generating augmented neuroimaging samples, thereby enhancing predictive performance through reduced overfitting and improved generalizability. We show that MultiViT2 significantly outperforms the first-generation model in schizophrenia classification accuracy and demonstrates strong scalability and portability.