To evaluate the feasibility of generative pre-trained transformer-4 (GPT-4) in generating structured reports (SRs) from mixed-language (English and Korean) narrative-style CT reports for pancreatic ductal adenocarcinoma (PDAC) and to assess its accuracy in categorizing PDCA resectability. This retrospective study included consecutive free-text reports of pancreas-protocol CT for staging PDAC, from two institutions, written in English or Korean from January 2021 to December 2023. Both the GPT-4 Turbo and GPT-4o models were provided prompts along with the free-text reports via an application programming interface and tasked with generating SRs and categorizing tumor resectability according to the National Comprehensive Cancer Network guidelines version 2.2024. Prompts were optimized using the GPT-4 Turbo model and 50 reports from Institution B. The performances of the GPT-4 Turbo and GPT-4o models in the two tasks were evaluated using 115 reports from Institution A. Results were compared with a reference standard that was manually derived by an abdominal radiologist. Each report was consecutively processed three times, with the most frequent response selected as the final output. Error analysis was guided by the decision rationale provided by the models. Of the 115 narrative reports tested, 96 (83.5%) contained both English and Korean. For SR generation, GPT-4 Turbo and GPT-4o demonstrated comparable accuracies (92.3% [1592/1725] and 92.2% [1590/1725], respectively; <i>P</i> = 0.923). In the resectability categorization, GPT-4 Turbo showed higher accuracy than GPT-4o (81.7% [94/115] vs. 67.0% [77/115], respectively; <i>P</i> = 0.002). In the error analysis of GPT-4 Turbo, the SR generation error rate was 7.7% (133/1725 items), which was primarily attributed to inaccurate data extraction (54.1% [72/133]). The resectability categorization error rate was 18.3% (21/115), with the main cause being violation of the resectability criteria (61.9% [13/21]). Both GPT-4 Turbo and GPT-4o demonstrated acceptable accuracy in generating NCCN-based SRs on PDACs from mixed-language narrative reports. However, oversight by human radiologists is essential for determining resectability based on CT findings.

Intracranial Hemorrhage (ICH) refers to cerebral bleeding resulting from ruptured blood vessels within the brain. Delayed and inaccurate diagnosis and treatment of ICH can lead to fatality or disability. Therefore, early and precise diagnosis of intracranial hemorrhage is crucial for protecting patients' lives. Automatic segmentation of hematomas in CT images can provide doctors with essential diagnostic support and improve diagnostic efficiency. CT images of intracranial hemorrhage exhibit characteristics such as multi-scale, multi-target, and blurred edges. This paper proposes a Multi-scale and Edge Feature Fusion Network (MEF-Net) to effectively extract multi-scale and edge features and fully fuse these features through a fusion mechanism. The network first extracts the multi-scale features and edge features of the image through the encoder and the edge detection module respectively, then fuses the deep information, and employs the multi-kernel attention module to process the shallow features, enhancing the multi-target recognition capability. Finally, the feature maps from each module are combined to produce the segmentation result. Experimental results indicate that this method has achieved average DICE scores of 0.7508 and 0.7443 in two public datasets respectively, surpassing those of several advanced methods in medical image segmentation currently available. The proposed MEF-Net significantly improves the accuracy of intracranial hemorrhage segmentation.

Prediction of severe disease (SVD) in patients with coronavirus disease (COVID-19) pneumonia at an early stage could allow for more appropriate triage and improve patient prognosis. Moreover, the visualization of the topological properties of COVID-19 pneumonia could help clinical physicians describe the reasons for their decisions. We aimed to construct predictive models of SVD in patients with COVID-19 pneumonia at an early stage on computed tomography (CT) images using SVD-specific features that can be visualized on accumulated Betti number (BN) maps. BN maps (b0 and b1 maps) were generated by calculating the BNs within a shifting kernel in a manner similar to a convolution. Accumulated BN maps were constructed by summing BN maps (b0 and b1 maps) derived from a range of multiple-threshold values. Topological features were computed as intrinsic topological properties of COVID-19 pneumonia from the accumulated BN maps. Predictive models of SVD were constructed with two feature selection methods and three machine learning models using nested fivefold cross-validation. The proposed model achieved an area under the receiver-operating characteristic curve of 0.854 and a sensitivity of 0.908 in a test fold. These results suggested that topological image features could characterize COVID-19 pneumonia at an early stage as SVD.

This article aimed to explore topological uncertainty in medical imaging, particularly in assessing coronary artery calcification using artificial intelligence (AI). Topological uncertainty refers to ambiguities in spatial and structural characteristics of medical features, which can impact the interpretation of coronary plaques. The article discusses the challenges of integrating AI with topological considerations and the need for specialized methodologies beyond traditional performance metrics. It highlights advancements in quantifying topological uncertainty, including the use of persistent homology and topological data analysis techniques. The importance of standardization in methodologies and ethical considerations in AI deployment are emphasized. It also outlines various types of uncertainty in topological frameworks for coronary plaques, categorizing them as quantifiable and controllable or quantifiable and not controllable. Future directions include developing AI algorithms that incorporate topological insights, establishing standardized protocols, and exploring ethical implications to revolutionize cardiovascular care through personalized treatment plans guided by sophisticated topological analysis. Recognizing and quantifying topological uncertainty in medical imaging as AI emerges is critical. Exploring topological uncertainty in coronary artery disease will revolutionize cardiovascular care, promising enhanced precision and personalization in diagnostics and treatment for millions affected by cardiovascular diseases.

This study assessed the diagnostic accuracy and prognostic implications of an artificial intelligence (AI) tool for coronary artery calcification (CAC) assessment on nongated, noncontrast thoracic computed tomography (CT). A single-centre retrospective analysis of 75 consecutive patients per age group (<40, 40-49, 50-59, 60-69, 70-79, 80-89, and ≥90 years) undergoing non-gated, non-contrast CT (January-December 2015) was conducted. AI analysis reported CAC presence and generated an Agatston score, and the performance was compared with baseline CT reports and a dedicated radiologist re-review. Interobserver variability between AI and radiologist assessments was measured using Cohen's κ. All-cause mortality was recorded, and its association with AI-detected CAC was tested. A total of 291 patients (mean age: 64 ± 19, 51% female) were included, with 80% (234/291) of AI reports passing radiologist quality assessment. CAC was reported on 7% (17/234) of initial clinical reports, 58% (135/234) on radiologist re-review, and 57% (134/234) by AI analysis. After manual quality assurance (QA) assessment, the AI tool demonstrated high sensitivity (96%), specificity (96%), positive predictive value (95%), and negative predictive value (97%) for CAC detection compared with radiologist re-review. Interobserver agreement was strong for CAC prevalence (κ = 0.92) and moderate for severity grading (κ = 0.60). AI-detected CAC presence and severity predicted all-cause mortality (p < 0.001). The AI tool exhibited feasible analysis potential for non-contrast, non-gated thoracic CTs, offering prognostic insights if integrated into routine practice. Nonetheless, manual quality assessment remains essential. This AI tool represents a potential enhancement to CAC detection and reporting on routine noncardiac chest CT.

Renal disorders are a significant public health concern and a cause of mortality related to renal failure. Manual diagnosis is subjective, labor-intensive, and depends on the expertise of nephrologists in renal anatomy. To improve workflow efficiency and enhance diagnosis accuracy, we propose an automated deep learning model, called EACWNet, which incorporates adaptive channel weighting-based deep convolutional neural network and explainable artificial intelligence. The proposed model categorizes renal computed tomography images into various classes, such as cyst, normal, tumor, and stone. The adaptive channel weighting module utilizes both global and local contextual insights to refine the final feature map channel weights through the integration of a scale-adaptive channel attention module in the higher convolutional blocks of the VGG-19 backbone model employed in the proposed method. The efficacy of the EACWNet model has been assessed using a publicly available renal CT images dataset, attaining an accuracy of 98.87% and demonstrating a 1.75% improvement over the backbone model. However, this model exhibits class-wise precision variation, achieving higher precision for cyst, normal, and tumor cases but lower precision for the stone class due to its inherent variability and heterogeneity. Furthermore, the model predictions have been subjected to additional analysis using the explainable artificial intelligence method such as local interpretable model-agnostic explanations, to visualize better and understand the model predictions.

The image quality of single-energy CT (SECT) limited the accuracy of automatic segmentation. Dual-energy CT (DECT) may potentially improve automatic segmentation yet the performance and strategy have not been investigated thoroughly. Based on DECT-generated virtual monochromatic images (VMIs), this study proposed a novel deep learning model (MIAU-Net) and evaluated the segmentation performance on the head organs-at-risk (OARs). The VMIs from 40 keV to 190 keV were retrospectively generated at intervals of 10 keV using the DECT of 46 patients. Images with expert delineation were used for training, validation, and testing MIAU-Net for automatic segmentation. Theperformance of MIAU-Net was compared with the existingU-Net, Attention-UNet, nnU-Net and TransFuse methods based on Dice Similarity Coefficient (DSC). Correlationanalysis was performed to evaluate and optimize the impact of different virtual energies on the accuracy of segmentation. Using MIAU-Net, average DSCs across all virtual energy levels were 93.78 %, 81.75 %, 84.46 %, 92.85 %, 94.40 %, and 84.75 % for the brain stem, optic chiasm, lens, mandible, eyes, and optic nerves, respectively, higher than the previous publications using SECT. MIAU-Net achieved the highest average DSC (88.84 %) and the lowest parameters (14.54 M) in all tested models. The results suggested that 60 keV-80 keV is the optimal VMI energy level for soft tissue delineation, while 100 keV is optimal for skeleton segmentation. This work proposed and validated a novel deep learning model for automatic segmentation based on DECT, suggesting potential advantages and OAR-specific optimal energy of using VMIs for automatic delineation.

To validate a liver lesion detection and classification model using staging computed tomography (CT) scans of colorectal cancer (CRC) patients. A UNet-based deep learning model was trained on 272 public liver tumour CT scans and tested on 220 CRC staging CTs acquired from a single institution (2014-2019). Performance metrics included lesion detection rates by size (<10 mm, 10-20 mm, >20 mm), segmentation accuracy (dice similarity coefficient, DSC), volume measurement agreement (Bland-Altman limits of agreement, LOAs; intraclass correlation coefficient, ICC), and classification accuracy (malignant vs benign) at patient and lesion levels (detected lesions only). The model detected 743 out of 884 lesions (84%), with detection rates of 75%, 91.3%, and 96% for lesions <10 mm, 10-20 mm, and >20 mm, respectively. The median DSC was 0.76 (95% CI: 0.72-0.80) for lesions <10 mm, 0.83 (95% CI: 0.79-0.86) for 10-20 mm, and 0.85 (95% CI: 0.82-0.88) for >20 mm. Bland-Altman analysis showed a mean volume bias of -0.12 cm³ (LOAs: -1.68 to +1.43 cm³), and ICC was 0.81. Lesion-level classification showed 99.5% sensitivity, 65.7% specificity, 53.8% positive predictive value (PPV), 99.7% negative predictive value (NPV), and 75.4% accuracy. Patient-level classification had 100% sensitivity, 27.1% specificity, 59.2% PPV, 100% NPV, and 64.5% accuracy. The model demonstrates strong lesion detection and segmentation performance, particularly for sub-centimetre lesions. Although classification accuracy was moderate, the 100% NPV suggests strong potential as a CRC staging screening tool. Future studies will assess its impact on radiologist performance and efficiency.

A cardiac digital twin is a virtual replica of a patient-specific heart, mimicking its anatomy and physiology. A crucial step of building a cardiac digital twin is anatomical twinning, where the computational mesh of the digital twin is tailored to the patient-specific cardiac anatomy. In a number of studies, the effect of anatomical variation on clinically relevant functional measurements like electrocardiograms (ECGs) is investigated, using computational simulations. While such a simulation environment provides researchers with a carefully controlled ground truth, the impact of anatomical differences on functional measurements in real-world patients remains understudied. In this study, we develop a biventricular statistical shape model and use it to quantify the effect of biventricular anatomy on ECG-derived and demographic features, providing novel insights for the development of digital twins of cardiac electrophysiology. To this end, a dataset comprising high-resolution cardiac CT scans from 271 healthy individuals, including athletes, is utilized. Furthermore, a novel, universal, ventricular coordinate-based method is developed to establish lightweight shape correspondence. The performance of the shape model is rigorously established, focusing on its dimensionality reduction capabilities and the training data requirements. The most important variability in healthy ventricles captured by the model is their size, followed by their elongation. These anatomical factors are found to significantly correlate with ECG-derived and demographic features. Additionally, a comprehensive synthetic cohort is made available, featuring ready-to-use biventricular meshes with fiber structures and anatomical region annotations. These meshes are well-suited for electrophysiological simulations.

Lung cancer is one of the leading causes of cancer-related mortality worldwide, with most cases diagnosed at advanced stages where curative treatment options are limited. Low-dose computed tomography (LDCT) for lung cancer screening (LCS) of individuals selected based on age and smoking history has shown a significant reduction in lung cancer-specific mortality. The number needed to screen to prevent one death from lung cancer is lower than that for breast cancer, cervical cancer, and colorectal cancer. Despite the substantial impact on reducing lung cancer-related mortality and proof that LCS with LDCT is effective, uptake of LCS has been low and LCS eligibility criteria remain imperfect. While LCS programs have historically faced patient recruitment challenges, research suggests that there are novel opportunities to both identify and improve screening for at-risk populations. In this review, we discuss the global obstacles to implementing LCS programs and strategies to overcome barriers in resource-limited settings. We explore successful approaches to promote LCS through robust engagement with community partners. Finally, we examine opportunities to enhance LCS in at-risk populations not captured by current eligibility criteria, including never smokers and individuals with a family history of lung cancer, with a focus on early detection through novel artificial intelligence technologies.