Age estimation, especially in adults, presents substantial challenges in different contexts ranging from forensic to clinical applications. Bone mineral density (BMD), with its distinct age-related variations, has emerged as a critical marker in this domain. This study aims to enhance chronological age estimation accuracy using deep learning (DL) incorporating a multi-modality fusion strategy based on BMD. We conducted a retrospective analysis of 4296 CT scans from a Chinese population, covering August 2015 to November 2022, encompassing lumbar, femur, and pubis modalities. Our DL approach, integrating multi-modality fusion, was applied to predict chronological age automatically. The model's performance was evaluated using an internal real-world clinical cohort of 644 scans (December 2022 to May 2023) and an external cadaver validation cohort of 351 scans. In single-modality assessments, the lumbar modality excelled. However, multi-modality models demonstrated superior performance, evidenced by lower mean absolute errors (MAEs) and higher Pearson's R² values. The optimal multi-modality model exhibited outstanding R² values of 0.89 overall, 0.88 in females, 0.90 in males, with the MAEs of 4.05 overall, 3.69 in females, 4.33 in males in the internal validation cohort. In the external cadaver validation, the model maintained favourable R² values (0.84 overall, 0.89 in females, 0.82 in males) and MAEs (5.01 overall, 4.71 in females, 5.09 in males), highlighting its generalizability across diverse scenarios. The integration of multi-modalities fusion with DL significantly refines the accuracy of adult age estimation based on BMD. The AI-based system that effectively combines multi-modalities BMD data, presenting a robust and innovative tool for accurate AAE, poised to significantly improve both geriatric diagnostics and forensic investigations.

Periodontitis is a severe infection affecting oral and systemic health and is traditionally diagnosed through clinical probing-a process that is time-consuming, uncomfortable for patients, and subject to variability based on the operator's skill. We hypothesized that computer vision can be used to estimate periodontal stability from radiographs alone. At the tooth level, we used intraoral radiographs to detect and categorize individual teeth according to their periodontal stability and corresponding treatment needs: healthy (prevention), stable (maintenance), and unstable (active treatment). At the patient level, we assessed full-mouth series and classified patients as stable or unstable by the presence of at least 1 unstable tooth. Our 3-way tooth classification model achieved an area under the receiver operating characteristic curve of 0.71 for healthy teeth, 0.56 for stable, and 0.67 for unstable. The model achieved an F₁ score of 0.45 for healthy teeth, 0.57 for stable, and 0.54 for unstable (recall, 0.70). Saliency maps generated by gradient-weighted class activation mapping primarily showed highly activated areas corresponding to clinically probed regions around teeth. Our binary patient classifier achieved an area under the receiver operating characteristic curve of 0.68 and an F₁ score of 0.74 (recall, 0.70). Taken together, our results suggest that it is feasible to estimate periodontal stability, which traditionally requires clinical and radiographic examination, from radiographic signal alone using computer vision. Variations in model performance across different classes at the tooth level indicate the necessity of further refinement.

This article proposes a method called DenseNet 121-Mask R-CNN (DN-MRCNN) for the detection and segmentation of brain tumors. The main objective is to reduce the execution time and accurately locate and segment the tumor, including its subareas. The input images undergo preprocessing techniques such as median filtering and Gaussian filtering to reduce noise and artifacts, as well as improve image quality. Histogram equalization is used to enhance the tumor regions, and image augmentation is employed to improve the model's diversity and robustness. To capture important patterns, a gated axial self-attention layer is added to the DenseNet 121 model, allowing for increased attention during the analysis of the input images. For accurate segmentation, boundary boxes are generated using a Regional Proposal Network with anchor customization. Post-processing techniques, specifically nonmaximum suppression, are performed to neglect redundant bounding boxes caused by overlapping regions. The Mask R-CNN model is used to accurately detect and segment the entire tumor (WT), tumor core (TC), and enhancing tumor (ET). The proposed model is evaluated using the BraTS 2019 dataset, the UCSF-PDGM dataset, and the UPENN-GBM dataset, which are commonly used for brain tumor detection and segmentation.

The nasopalatine canal (NPC) is an anatomical formation with varying morphology. NPC can be visualized using the cone-beam computed tomography (CBCT). Also, CBCT has been used in many studies on artificial intelligence (AI). The "You only look once" (YOLO) is an AI framework that stands out with its speed. This study compared the observer and AI regarding the NPC segmentation and assessment of the NPC furcation status in CBCT images. In this study, axial sections of 200 CBCT images were used. These images were labeled and evaluated for the absence or presence of the NPC furcation. These images were then divided into three; 160 images were used as the training dataset, 20 as the validation dataset, and 20 as the test dataset. The training was performed by making 800 epochs using the YOLOv5x-seg model. Sensitivity, Precision, F1 score, IoU, mAP, and AUC values were determined for NPC detection, segmentation, and classification of the YOLOv5x-seg model. The values were found to be 0.9680, 0.9953, 0.9815, 0.9636, 0.7930, and 0.8841, respectively, for the group with the absence of the NPC furcation; and 0.9827, 0.9975, 0.9900, 0.9803, 0.9637, and 0.9510, for the group with the presence of the NPC furcation. Our results showed that even when the YOLOv5x-seg model is trained with the NPC furcation and fewer datasets, it achieves sufficient prediction accuracy. The segmentation feature of the YOLOv5 algorithm, which is based on an object detection algorithm, has achieved quite successful results despite its recent development.

Myocardial computed tomography (CT) late enhancement (LE) allows assessment of myocardial scarring. Super-resolution deep learning image reconstruction (SR-DLR) trained on data acquired from ultra-high-resolution CT may improve image quality for CT-LE. Therefore, this study investigated image noise and image quality with SR-DLR compared with conventional DLR (C-DLR) and hybrid iterative reconstruction (hybrid IR). We retrospectively analyzed 30 patients who underwent CT-LE using 320-row CT. The CT protocol comprised stress dynamic CT perfusion, coronary CT angiography, and CT-LE. CT-LE images were reconstructed using three different algorithms: SR-DLR, C-DLR, and hybrid IR. Image noise, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and qualitative image quality scores are in terms of noise reduction, sharpness, visibility of scar and myocardial boarder, and overall image quality. Inter-observer differences in myocardial scar sizing in CT-LE by the three algorithms were also compared. SR-DLR significantly decreased image noise by 35% compared to C-DLR (median 6.2 HU, interquartile range [IQR] 5.6-7.2 HU vs 9.6 HU, IQR 8.4-10.7 HU; p < 0.001) and by 37% compared to hybrid IR (9.8 HU, IQR 8.5-12.0 HU; p < 0.001). SNR and CNR of CT-LE reconstructed using SR-DLR were significantly higher than with C-DLR (both p < 0.001) and hybrid IR (both p < 0.05). All qualitative image quality scores were higher with SR-DLR than those with C-DLR and hybrid IR (all p < 0.001). The inter-observer differences in scar sizing were reduced with SR-DLR and C-DLR compared with hybrid IR (both p = 0.02). SR-DLR reduces image noise and improves image quality of myocardial CT-LE compared with C-DLR and hybrid IR techniques and improves inter-observer reproducibility of scar sizing compared to hybrid IR. The SR-DLR approach has the potential to improve the assessment of myocardial scar by CT late enhancement.

Large language models (LLMs) have the potential to objectively evaluate radiology resident reports; however, research on their use for feedback in radiology training and assessment of resident skill development remains limited. This study aimed to assess the effectiveness of LLMs in revising radiology reports by comparing them with reports verified by board-certified radiologists and to analyze the progression of resident's reporting skills over time. To identify the LLM that best aligned with human radiologists, 100 reports were randomly selected from 7376 reports authored by nine first-year radiology residents. The reports were evaluated based on six criteria: (1) addition of missing positive findings, (2) deletion of findings, (3) addition of negative findings, (4) correction of the expression of findings, (5) correction of the diagnosis, and (6) proposal of additional examinations or treatments. Reports were segmented into four time-based terms, and 900 reports (450 CT and 450 MRI) were randomly chosen from the initial and final terms of the residents' first year. The revised rates for each criterion were compared between the first and last terms using the Wilcoxon Signed-Rank test. Among the three LLMs-ChatGPT-4 Omni (GPT-4o), Claude-3.5 Sonnet, and Claude-3 Opus-GPT-4o demonstrated the highest level of agreement with board-certified radiologists. Significant improvements were noted in Criteria 1-3 when comparing reports from the first and last terms (Criteria 1, 2, and 3; P < 0.001, P = 0.023, and P = 0.004, respectively) using GPT-4o. No significant changes were observed for Criteria 4-6. Despite this, all criteria except for Criteria 6 showed progressive enhancement over time. LLMs can effectively provide feedback on commonly corrected areas in radiology reports, enabling residents to objectively identify and improve their weaknesses and monitor their progress. Additionally, LLMs may help reduce the workload of radiologists' mentors.

Respiratory motion during PET acquisition may produce lesion blurring. Ultra-fast 20-second breath-hold (U2BH) PET reduces respiratory motion artifacts, but the shortened scanning time increases statistical noise and may affect diagnostic quality. This study aims to denoise the U2BH PET images using a deep learning (DL)-based method. The study was conducted on two datasets collected from five scanners where the first dataset included 1272 retrospectively collected full-time PET data while the second dataset contained 46 prospectively collected U2BH and the corresponding full-time PET/CT images. A robust and data-efficient DL method called mask vision transformer (Mask-ViT) was proposed which, after fine-tuned on a limited number of training data from a target scanner, was directly applied to unseen testing data from new scanners. The performance of Mask-ViT was compared with state-of-the-art DL methods including U-Net and C-Gan taking the full-time PET images as the reference. Statistical analysis on image quality metrics were carried out with Wilcoxon signed-rank test. For clinical evaluation, two readers scored image quality on a 5-point scale (5 = excellent) and provided a binary assessment for diagnostic quality evaluation. The U2BH PET images denoised by Mask-ViT showed statistically significant improvement over U-Net and C-Gan on image quality metrics (p < 0.05). For clinical evaluation, Mask-ViT exhibited a lesion detection accuracy of 91.3%, 90.4% and 91.7%, when it was evaluated on three different scanners. Mask-ViT can effectively enhance the quality of the U2BH PET images in a data-efficient generalization setup. The denoised images meet clinical diagnostic requirements of lesion detectability.

Attention Mechanism (AM) selectively focuses on essential information for imaging tasks and captures relationships between regions from distant pixel neighborhoods to compute feature representations. Accelerated magnetic resonance image (MRI) reconstruction can benefit from AM, as the imaging process involves acquiring Fourier domain measurements that influence the image representation in a non-local manner. However, AM-based models are more adept at capturing low-frequency information and have limited capacity in constructing high-frequency representations, restricting the models to smooth reconstruction. Secondly, AM-based models need mode-specific retraining for multimodal MRI data as their knowledge is restricted to local contextual variations within modes that might be inadequate to capture the diverse transferable features across heterogeneous data domains. To address these challenges, we propose a neuromodulation-based discriminative multi-spectral AM for scalable MRI reconstruction, that can (i) propagate the context-aware high-frequency details for high-quality image reconstruction, and (ii) capture features reusable to deviated unseen domains in multimodal MRI, to offer high practical value for the healthcare industry and researchers. The proposed network consists of a spectral filtering convolutional neural network to capture mode-specific transferable features to generalize to deviated MRI data domains and a dynamic high-pass kernel generation transformer that focuses on high-frequency details for improved reconstruction. We have evaluated our model on various aspects, such as comparative studies in supervised and self-supervised learning, diffusion model-based training, closed-set and open-set generalization under heterogeneous MRI data, and interpretation-based analysis. Our results show that the proposed method offers scalable and high-quality reconstruction with best improvement margins of ∼1 dB in PSNR and ∼0.01 in SSIM under unseen scenarios. Our code is available at https://github.com/sriprabhar/SHFormer.

Objective classification biomarkers that are developed using resting-state functional magnetic resonance imaging (rs-fMRI) data are expected to contribute to more effective treatment for psychiatric disorders. Unfortunately, no widely accepted biomarkers are available at present, partially because of the large variety of analysis pipelines for their development. In this study, we comprehensively evaluated analysis pipelines using a large-scale, multi-site fMRI dataset for major depressive disorder (MDD). We explored combinations of options in four sub-processes of the analysis pipelines: six types of brain parcellation, four types of functional connectivity (FC) estimations, three types of site-difference harmonization, and five types of machine-learning methods. A total of 360 different MDD classification biomarkers were constructed using the SRPBS dataset acquired with unified protocols (713 participants from four sites) as the discovery dataset, and datasets from other projects acquired with heterogeneous protocols (449 participants from four sites) were used for independent validation. We repeated the procedure after swapping the roles of the two datasets to identify superior pipelines, regardless of the discovery dataset. The classification results of the top 10 biomarkers showed high similarity, and weight similarity was observed between eight of the biomarkers, except for two that used both data-driven parcellation and FC computation. We applied the top 10 pipelines to the datasets of other psychiatric disorders (autism spectrum disorder and schizophrenia), and eight of the biomarkers exhibited sufficient classification performance for both disorders. Our results will be useful for establishing a standardized pipeline for classification biomarkers.

Myocardial perfusion imaging (MPI) using single-photon emission computed tomography (SPECT) is a well-established modality for noninvasive diagnostic assessment of coronary artery disease (CAD). However, the time-consuming and experience-dependent visual interpretation of SPECT images remains a limitation in the clinic. We aimed to develop advanced models to diagnose CAD using different supervised and semi-supervised deep learning (DL) algorithms and training strategies, including transfer learning and data augmentation, with SPECT-MPI and invasive coronary angiography (ICA) as standard of reference. A total of 940 patients who underwent SPECT-MPI were enrolled (281 patients included ICA). Quantitative perfusion SPECT (QPS) was used to extract polar maps of rest and stress states. We defined two different tasks, including (1) Automated CAD diagnosis with expert reader (ER) assessment of SPECT-MPI as reference, and (2) CAD diagnosis from SPECT-MPI based on reference ICA reports. In task 2, we used 6 strategies for training DL models. We implemented 13 different DL models along with 4 input types with and without data augmentation (WAug and WoAug) to train, validate, and test the DL models (728 models). One hundred patients with ICA as standard of reference (the same patients in task 1) were used to evaluate models per vessel and per patient. Metrics, such as the area under the receiver operating characteristics curve (AUC), accuracy, sensitivity, specificity, precision, and balanced accuracy were reported. DeLong and pairwise Wilcoxon rank sum tests were respectively used to compare models and strategies after 1000 bootstraps on the test data for all models. We also compared the performance of our best DL model to ER's diagnosis. In task 1, DenseNet201 Late Fusion (AUC = 0.89) and ResNet152V2 Late Fusion (AUC = 0.83) models outperformed other models in per-vessel and per-patient analyses, respectively. In task 2, the best models for CAD prediction based on ICA were Strategy 3 (a combination of ER- and ICA-based diagnosis in train data), WoAug InceptionResNetV2 EarlyFusion (AUC = 0.71), and Strategy 5 (semi-supervised approach) WoAug ResNet152V2 EarlyFusion (AUC = 0.77) in per-vessel and per-patient analyses, respectively. Moreover, saliency maps showed that models could be helpful for focusing on relevant spots for decision making. Our study confirmed the potential of DL-based analysis of SPECT-MPI polar maps in CAD diagnosis. In the automation of ER-based diagnosis, models' performance was promising showing accuracy close to expert-level analysis. It demonstrated that using different strategies of data combination, such as including those with and without ICA, along with different training methods, like semi-supervised learning, can increase the performance of DL models. The proposed DL models could be coupled with computer-aided diagnosis systems and be used as an assistant to nuclear medicine physicians to improve their diagnosis and reporting, but only in the LAD territory. Not applicable.