Congenital heart defect (CHD) is a significant, rapidly emerging global problem in child health and a leading cause of neonatal and childhood death. Prenatal detection of CHDs with the help of ultrasound allows better perinatal management of such pregnancies, leading to reduced neonatal mortality, morbidity and developmental complications. However, there is a wide variation in reported fetal heart problem detection rates from 34% to 85%, with some low- and middle-income countries detecting as low as 9.3% of cases before birth. Research has shown that deep learning-based or more general artificial intelligence (AI) models can support the detection of fetal CHDs more rapidly than humans performing ultrasound scan. Progress in this AI-based research depends on the availability of large, well-curated and diverse data of ultrasound images and videos of normal and abnormal fetal hearts. Currently, CHD detection based on AI models is not accurate enough for practical clinical use, in part due to the lack of ultrasound data available for machine learning as CHDs are rare and heterogeneous, the retrospective nature of published studies, the lack of multicentre and multidisciplinary collaboration, and utilisation of mostly standard planes still images of the fetal heart for AI models. Our aim is to develop AI models that could support clinicians in detecting fetal CHDs in real time, particularly in nonspecialist or low-resource settings where fetal echocardiography expertise is not readily available. We have designed the Clinical Artificial Intelligence Fetal Echocardiography (CAIFE) study as an international multicentre multidisciplinary collaboration led by a clinical and an engineering team at the University of Oxford. This study involves five multicountry hospital sites for data collection (Oxford, UK (n=1), London, UK (n=3) and Southport, Australia (n=1)). We plan to curate 14 000 retrospective ultrasound scans of fetuses with normal hearts (n=13 000) and fetuses with CHDs (n=1000), as well as 2400 prospective ultrasound cardiac scans, including the proposed research-specific CAIFE 10 s video sweeps, from fetuses with normal hearts (n=2000) and fetuses diagnosed with major CHDs (n=400). This gives a total of 16 400 retrospective and prospective ultrasound scans from the participating hospital sites. We will build, train and validate computational models capable of differentiating between normal fetal hearts and those diagnosed with CHDs and recognise specific types of CHDs. Data will be analysed using statistical metrics, namely, sensitivity, specificity and accuracy, which include calculating positive and negative predictive values for each outcome, compared with manual assessment. We will disseminate the findings through regional, national and international conferences and through peer-reviewed journals. The study was approved by the Health Research Authority, Care Research Wales and the Research Ethics Committee (Ref: 23/EM/0023; IRAS Project ID: 317510) on 8 March 2023. All collaborating hospitals have obtained the local trust research and development approvals.

This study investigates the use of list-mode (LM) maximum a posteriori (MAP) expectation maximization (EM) incorporating prior information predicted by a convolutional neural network for image reconstruction in fast neutron (FN)-based proton therapy range verification.
Approach. A conditional generative adversarial network (pix2pix) was trained on progressively noisier data, where detector resolution effects were introduced gradually to simulate realistic conditions. FN data were generated using Monte Carlo simulations of an 85 MeV proton pencil beam in a computed tomography (CT)-based lung cancer patient model, with range shifts emulating weight gain and loss. The network was trained to estimate the expected two-dimensional (2D) ground truth FN production distribution from simple back-projection images. Performance was evaluated using mean squared error (MSE), structural similarity index (SSIM), and the correlation between shifts in predicted distributions and true range shifts. 
Main results. Our results show that pix2pix performs well on noise-free data but suffers from significant degradation when detector resolution effects are introduced. Among the LM-MAP-EM approaches tested, incorporating a mean prior estimate into the reconstruction process improved performance, with LM-MAP-EM using a mean prior estimate outperforming naïve LM maximum likelihood EM (LM-MLEM) and conventional LM-MAP-EM with a smoothing quadratic energy function in terms of SSIM. 
Significance. Findings suggest that deep learning techniques can enhance iterative reconstruction for range verification in proton therapy. However, the effectiveness of the model is highly dependent on data quality, limiting its robustness in high-noise scenarios.&#xD.

Thyroid nodules are commonly evaluated using ultrasound-based risk stratification systems, which rely on subjective descriptors. Artificial intelligence (AI) may improve assessment, but its effectiveness in non-subspecialist settings is unclear. This study evaluated the impact of an AI-based decision support system (AI-DSS) on thyroid nodule ultrasound assessments by general endocrinologists (GE) without subspecialty thyroid imaging training. A prospective cohort study was conducted on 80 patients undergoing thyroid ultrasound in GE outpatient clinics. Thyroid ultrasound was performed based on clinical judgment as part of routine care by GE. Images were retrospectively analyzed using an AI-DSS (Koios DS), independently of clinician assessments. AI-DSS results were compared with initial GE evaluations and, when referred, with expert evaluations at a subspecialized thyroid nodule clinic (TNC). Agreement in ultrasound features, risk classification by the American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) and American Thyroid Association guidelines, and referral recommendations was assessed. AI-DSS differed notably from GE, particularly assessing nodule composition (solid: 80%vs.36%,p < 0.01), echogenicity (hypoechoic:52%vs.16%,p < 0.01), and echogenic foci (microcalcifications:10.7%vs.1.3%,p < 0.05). AI-DSS classification led to a higher referral rate compared to GE (37.3%vs.30.7%, not statistically significant). Agreement between AI-DSS and GE in ACR TI-RADS scoring was moderate (r = 0.337;p < 0.001), but improved when comparing GE to AI-DSS and TNC subspecialist (r = 0.465;p < 0.05 and r = 0.607;p < 0.05, respectively). In a non-subspecialist setting, non-adjunct AI-DSS use did not significantly improve risk stratification or reduce hypothetical referrals. The system tended to overestimate risk, potentially leading to unnecessary procedures. Further optimization is required for AI to function effectively in low-prevalence environment.

To develop a deep learning model to automatically measure the curvature-related sagittal parameters on cervical spinal X-ray images. This retrospective study collected a total of 700 lateral cervical spine X-ray images from three hospitals, consisting of 500 training sets, 100 internal test sets, and 100 external test sets. 6 measured parameters and 34 landmarks were measured and labeled by two doctors and averaged as the gold standard. A Convolutional neural network (CNN) model was built by training on 500 images and testing on 200 images. Statistical analysis is used to evaluate labeling differences and model performance. The percentages of the difference in distance between landmarks within 4 mm were 96.90% (Dr. A vs. Dr. B), 98.47% (Dr. A vs. model), and 97.31% (Dr. B vs. model); within 3 mm were 94.88% (Dr. A vs. Dr. B), 96.43% (Dr. A vs. model), and 94.16% (Dr. B vs. model). The mean difference of the algorithmic model in labeling landmarks was 1.17 ± 1.14 mm. The mean absolute error (MAE) of the algorithmic model for the Borden method, Cervical curvature index (CCI), Vertebral centroid measurement cervical lordosis (CCL), C₀-C₇ Cobb, C₁-C₇ Cobb, C₂-C₇ Cobb in the test sets are 1.67 mm, 2.01%, 3.22°, 2.37°, 2.49°, 2.81°, respectively; symmetric mean absolute percentage error (SMAPE) was 20.06%, 21.68%, 20.02%, 6.68%, 5.28%, 20.46%, respectively. Also, the algorithmic model of the six cervical sagittal parameters is in good agreement with the gold standard (intraclass correlation efficiency was 0.983; p < 0.001). Our deep learning algorithmic model had high accuracy in recognizing the landmarks of the cervical spine and automatically measuring cervical spine-related parameters, which can help radiologists improve their diagnostic efficiency.

Clinical staging on CT has several biases, and a radiogenomics approach could be proposed. The study aimed to test the performance of a radiogenomics approach in identifying high-risk colon cancer. ATTRACT is a multicentric trial, registered in ClinicalTrials.gov (NCT06108310). Three hundred non-metastatic colon cancer patients were retrospectively enrolled and divided into two groups, high-risk and no-risk, according to the pathological staging. Radiological evaluations were performed by two abdominal radiologists. For 151 patients, we achieved genomics. The baseline CT scans were used to evaluate the radiological assessment and to perform 3D cancer segmentation. One expert radiologist used open-source software to perform the volumetric cancer segmentations on baseline CT scans in the portal phase (3DSlicer v4.10.2). Implementing the classical LASSO with a machine-learning library method was used to select the optimal features to build Model 1 (clinical-radiological plus radiomic feature, 300 patients) and Model 2 (Model 1 plus genomics, 151 patients). The performance of clinical-radiological interpretation was assessed regarding the area under the curve (AUC), sensitivity, specificity, and accuracy. The average performance of Models 1 and 2 was also calculated. In total, 262/300 were classified as high-risk and 38/300 as no-risk. Clinical-radiological interpretation by the two radiologists achieved an AUC of 0.58-0.82 (95% CI: 0.52-0.63 and 0.76-0.85, p < 0.001, respectively), sensitivity: 67.9-93.8%, specificity: 47.4-68.4%, and accuracy: 65.3-90.7%, respectively. Model 1 yielded AUC: 0.74 (95% CI: 0.61-0.88, p < 0.005), sensitivity: 86%, specificity: 48%, and accuracy: 81%. Model2 reached AUC: 0.84, (95% CI: 0.68-0.99, p < 0.005), sensitivity: 88%, specificity: 63%, and accuracy: 84%. The radiogenomics model outperformed radiological interpretation in identifying high-risk colon cancer. Question Can this radiogenomic model identify high-risk stages II and III colon cancer in a preoperative clinical setting? Findings This radiogenomics model outperformed both the radiomics and radiological interpretations, reducing the risk of improper staging and incorrect treatment options. Clinical relevance The radiogenomics model was demonstrated to be superior to radiological interpretation and radiomics in identifying high-risk colon cancer, and could therefore be promising in stratifying high-risk and low-risk patients.

Publications on artificial intelligence (AI) applications have dramatically increased for most medical specialties, including obstetrics. Here, we review the most recent pertinent publications on AI programs in obstetrics, describe trends in AI applications for specific obstetric problems, and assess AI's possible effects on obstetric care. Searches were performed in PubMed (MeSH), MEDLINE, Ovid, ClinicalTrials.gov, Google Scholar, and Web of Science using a combination of keywords and text words related to "obstetrics," "pregnancy," "artificial intelligence," "machine learning," "deep learning," and "neural networks," for articles published between June 1, 2019, and May 31, 2024. A total of 1,768 articles met at least one search criterion. After eliminating reviews, duplicates, retractions, inactive research protocols, unspecified AI programs, and non-English-language articles, 207 publications remained for further review. Most studies were conducted outside of the United States, were published in nonobstetric journals, and focused on risk prediction. Study population sizes ranged widely from 10 to 953,909, and model performance abilities also varied widely. Evidence quality was assessed by the description of model construction, predictive accuracy, and whether validation had been performed. Most studies had patient groups differing considerably from U.S. populations, rendering their generalizability to U.S. patients uncertain. Artificial intelligence ultrasound applications focused on imaging issues are those most likely to influence current obstetric care. Other promising AI models include early risk screening for spontaneous preterm birth, preeclampsia, and gestational diabetes mellitus. The rate at which AI studies are being performed virtually guarantees that numerous applications will eventually be introduced into future U.S. obstetric practice. Very few of the models have been deployed in obstetric practice, and more high-quality studies are needed with high predictive accuracy and generalizability. Assuming these conditions are met, there will be an urgent need to educate medical students, postgraduate trainees and practicing physicians to understand how to effectively and safely implement this technology.

The medical field uses Magnetic Resonance Imaging (MRI) as an essential diagnostic tool which provides doctors non-invasive images of brain structures and pathological conditions. Brain tumor detection stands as a vital application that needs specific and effective approaches for both medical diagnosis and treatment procedures. The challenges from manual examination of MRI scans stem from inconsistent tumor features including heterogeneity and irregular dimensions which results in inaccurate assessments of tumor size. To address these challenges, this paper proposes an Automated Classification and Grading Diagnosis Model (ACGDM) using MRI images. Unlike conventional methods, ACGDM introduces a Multi-Scale Graph Neural Network (MSGNN), which dynamically captures hierarchical and multi-scale dependencies in MRI data, enabling more accurate feature representation and contextual analysis. Additionally, the Spatio-Temporal Transformer Attention Mechanism (STTAM) effectively models both spatial MRI patterns and temporal evolution by incorporating cross-frame dependencies, enhancing the model's sensitivity to subtle disease progression. By analyzing multi-modal MRI sequences, ACGDM dynamically adjusts its focus across spatial and temporal dimensions, enabling precise identification of salient features. Simulations are conducted using Python and standard libraries to evaluate the model on the BRATS 2018, 2019, 2020 datasets and the Br235H dataset, encompassing diverse MRI scans with expert annotations. Extensive experimentation demonstrates 99.8% accuracy in detecting various tumor types, showcasing its potential to revolutionize diagnostic practices and improve patient outcomes.

This study aims to explore whether intratumoral and peritumoral ultrasound radiomics of ultrasound images can predict the low expression status of human epidermal growth factor receptor 2 (HER2) in HER2-negative breast cancer patients. HER2-negative breast cancer patients were recruited retrospectively and randomly divided into a training cohort (n = 303) and a test cohort (n = 130) at a ratio of 7:3. The region of interest within the breast ultrasound image was designated as the intratumoral region, and expansions of 3 mm, 5 mm, and 8 mm from this region were considered as the peritumoral regions for the extraction of ultrasound radiomic features. Feature extraction and selection were performed, and radiomics scores (Rad-score) were obtained in four ultrasound radiomics scenarios: intratumoral only, intratumoral + peritumoral 3 mm, intratumoral + peritumoral 5 mm, and intratumoral + peritumoral 8 mm. An optimal combined nomogram radiomic model incorporating clinical features was established and validated. Subsequently, the diagnostic performance of the radiomic models was evaluated. The results indicated that the intratumoral + peritumoral (5 mm) ultrasound radiomics exhibited the excellent diagnostic performance in evaluated the HER2 low expression. The nomogram combining intratumoral + peritumoral (5 mm) and clinical features showed superior diagnostic performance, achieving an area under the curve (AUC) of 0.911 and 0.869 in the training and test cohorts, respectively. The combination of intratumoral + peritumoral (5 mm) ultrasound radiomics and clinical features possesses the capability to accurately predict the low-expression status of HER2 in HER2-negative breast cancer patients.

To investigate the effects of the deep learning model RaDynPET on fourfold reduced-count whole-body PET examinations. A total of 120 patients (84 internal cohorts and 36 external cohorts) undergoing ¹⁸F-FDG PET/CT examinations were enrolled. PET images were reconstructed using OSEM algorithm with 120-s (G120) and 30-s (G30) list-mode data. RaDynPET was developed to generate enhanced images (R30) from G30. Two experienced nuclear medicine physicians independently evaluated subjective image quality using a 5-point Likert scale. Standardized uptake values (SUV), standard deviations, liver signal-to-noise ratio (SNR), lesion tumor-to-background ratio (TBR), and contrast-to-noise ratio (CNR) were compared. Subgroup analyses evaluated performance across demographics, and lesion detectability were evaluated using external datasets. RaDynPET was also compared to other deep learning methods. In internal cohorts, R30 demonstrated significantly higher image quality scores than G30 and G120. R30 showed excellent agreement with G120 for liver and lesion SUV values and surpassed G120 in liver SNR and CNR. Liver SNR and CNR of R30 were comparable to G120 in thin group, and the CNR of R30 was comparable to G120 in young age group. In external cohorts, R30 maintained strong SUV agreement with G120, with lesion-level sensitivity and specificity of 95.45% and 98.41%, respectively. There was no statistical difference in lesion detection between R30 and G120. RaDynPET achieved the highest PSNR and SSIM among deep learning methods. The RaDynPET model effectively restored high image quality while maintaining SUV agreement for ¹⁸F-FDG PET scans acquired in 25% of the standard acquisition time.

Determining the involvement of specific peripheral nerves (PNs) in the upper limb associated with signs of muscle denervation can be challenging. This study aims to develop, compare, and validate various large language models (LLMs) to automatically identify and establish potential relationships between denervated muscles and their corresponding PNs. We collected 300 retrospective MRI reports in Spanish from upper limb examinations conducted between 2018 and 2024 that showed signs of muscle denervation. An expert radiologist manually annotated these reports based on the affected peripheral nerves (median, ulnar, radial, axillary, and suprascapular). BERT, DistilBERT, mBART, RoBERTa, and Medical-ELECTRA models were fine-tuned and evaluated on the reports. Additionally, an automatic voting system was implemented to consolidate predictions through majority voting. The voting system achieved the highest F1 scores for the median, ulnar, and radial nerves, with scores of 0.88, 1.00, and 0.90, respectively. Medical-ELECTRA also performed well, achieving F1 scores above 0.82 for the axillary and suprascapular nerves. In contrast, mBART demonstrated lower performance, particularly with an F1 score of 0.38 for the median nerve. Our voting system generally outperforms the individually tested LLMs in determining the specific PN likely associated with muscle denervation patterns detected in upper limb MRI reports. This system can thereby assist radiologists by suggesting the implicated PN when generating their radiology reports.