Liver cancer ranks among the most lethal cancers. Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer and better diagnostic tools are needed to diagnose patients at risk. The aim is to develop a machine learning algorithm that enhances the sensitivity and specificity of the Contrast-Enhanced Ultrasound Liver Imaging Reporting and Data System (CEUS-LIRADS) in classifying indeterminate at-risk liver nodules (LR-M, LR-3, LR-4) as HCC or non-HCC. Our study includes patients at risk for HCC with untreated indeterminate focal liver observations detected on US or contrast-enhanced CT or MRI performed as part of their clinical standard of care from January 2018 to November 2022. Recursive partitioning was used to improve HCC diagnosis in indeterminate at-risk nodules. Demographics, blood biomarkers, and CEUS imaging features were evaluated as potential predictors for the algorithm to classify nodules as HCC or non-HCC. We evaluated 244 indeterminate liver nodules from 224 patients (mean age 62.9 y). Of the nodules, 73.2% (164/224) were from males. The algorithm was trained on a random 2/3 partition of 163 liver nodules and correctly reclassified more than half of the HCC liver nodules previously categorized as indeterminate in the independent 1/3 test partition of 81 liver nodules, achieving a sensitivity of 56.3% (95% CI: 42.0%, 70.2%) and specificity of 93.9% (95% CI: 84.4%, 100.0%). Machine learning was applied to the multicenter, multinational study of CEUS LI-RADS indeterminate at-risk liver nodules and correctly diagnosed HCC in more than half of the HCC nodules.

Neonatal neuroimaging is essential in research and clinical practice, offering important insights into brain development and neurologic injury mechanisms. Visualizing the brain enables researchers and clinicians to improve neonatal care and parental counselling through better diagnosis and prognostication of disease. Common neuroimaging modalities used in the neonatal intensive care unit (NICU) are cranial ultrasonography (cUS) and magnetic resonance imaging (MRI). Between these modalities, conventional MRI provides the optimal image resolution and detail about the developing brain, while advanced MRI techniques allow for the evaluation of tissue microstructure and functional networks. Over the last two decades, medical imaging techniques using brain MRI have rapidly progressed, and these advances have facilitated high-quality extraction of quantitative features as well as the implementation of novel devices for use in neurological disorders. Major advancements encompass the use of low-field dedicated MRI systems within the NICU and trials of ultralow-field portable MRI systems at the bedside. Additionally, higher-field magnets are utilized to enhance image quality, and ultrafast brain MRI is employed to decrease image acquisition time. Furthermore, the implementation of advanced MRI sequences, the application of machine learning algorithms, multimodal neuroimaging techniques, motion correction techniques, and novel modalities are used to visualize pathologies that are not visible to the human eye. In this narrative review, we will discuss the fundamentals of these neuroimaging modalities, and their clinical applications to explore the present landscape of neonatal neuroimaging from bench to bedside.

This clinically oriented review explores the technical advancements of simultaneous PET/magnetic resonance (MR) imaging to provide an overview of the addressed obstacles over time, current challenges, and future trends in the field. In particular, advanced attenuation and motion correction techniques and MR-guided PET reconstruction frameworks were reviewed, and the state-of-the-art PET/MR systems and their strengths were discussed. Overall, PET/MR holds great potential in various clinical applications, including oncology, neurology, and cardiology. However, it requires continued optimization in hardware, algorithms, and clinical protocols to achieve broader adoption and be included in the routine clinical standards.

The bone volume drilled by surgeons during mastoidectomy is determined by the need to localize the position, optimize the view, and reach the surgical endpoint while avoiding critical structures. Predicting the volume of bone removed before an operation can significantly enhance surgical training by providing precise, patient-specific guidance and enable the development of more effective computer-assisted and robotic surgical interventions. Single institution, cross-sectional. VR simulation. We developed a deep learning pipeline to automate the prediction of bone volume removed during mastoidectomy using data from virtual reality mastoidectomy simulations. The data set included 15 deidentified temporal bone computed tomography scans. The network was evaluated using fivefold cross-validation, comparing predicted and actual bone removal with metrics such as the Dice score (DSC) and Hausdorff distance (HD). Our method achieved a median DSC of 0.775 (interquartile range [IQR]: 0.725-0.810) and a median HD of 0.492 mm (IQR: 0.298-0.757 mm). Predictions reached the mastoidectomy endpoint of visualizing the horizontal canal and incus in 80% (12/15) of temporal bones. Qualitative analysis indicated that predictions typically produced realistic mastoidectomy endpoints, though some cases showed excessive or insufficient bone removal, particularly at the temporal bone cortex and tegmen mastoideum. This study establishes a foundational step in using deep learning to predict bone volume removal during mastoidectomy. The results indicate that learning-based methods can reasonably approximate the surgical endpoint of mastoidectomy. Further refinement with larger, more diverse data sets and improved model architectures will be essential for enhancing prediction accuracy.

This study aims to integrate CT imaging with occupational health surveillance data to construct a multimodal model for preclinical CWP identification and individualized risk evaluation. CT images and occupational health surveillance data were retrospectively collected from 874 coal workers, including 228 Stage I and 4 Stage II pneumoconiosis patients, along with 600 healthy and 42 subcategory 0/1 coal workers. First, the YOLOX was employed for automated 3D lung extraction to extract radiomics features. Second, two feature selection algorithms were applied to select critical features from both CT radiomics and occupational health data. Third, three distinct feature sets were constructed for model training: CT radiomics features, occupational health data, and their multimodal integration. Finally, five machine learning models were implemented to predict the preclinical stage of CWP. The model's performance was evaluated using the receiver operating characteristic curve (ROC), accuracy, sensitivity, and specificity. SHapley Additive exPlanation (SHAP) values were calculated to determine the prediction role of each feature in the model with the highest predictive performance. The YOLOX-based lung extraction demonstrated robust performance, achieving an Average Precision (AP) of 0.98. 8 CT radiomic features and 4 occupational health surveillance data were selected for the multimodal model. The optimal occupational health surveillance feature subset comprised the Length of service. Among 5 machine learning algorithms evaluated, the Decision Tree-based multimodal model showed superior predictive capacity on the test set of 142 samples, with an AUC of 0.94 (95% CI 0.88-0.99), accuracy 0.95, specificity 1.00, and Youden's index 0.83. SHAP analysis indicated that Total Protein Results, original shape Flatness, diagnostics Image original Mean were the most influential contributors. Our study demonstrated that the multimodal model demonstrated strong predictive capability for the preclinical stage of CWP by integrating CT radiomic features with occupational health data.

Generative artificial intelligence models facilitate open-data sharing by proposing synthetic data as surrogates of real patient data. Despite the promise for healthcare, some of these models are susceptible to patient data memorization, where models generate patient data copies instead of novel synthetic samples, resulting in patient re-identification. Here we assess memorization in unconditional latent diffusion models by training them on a variety of datasets for synthetic data generation and detecting memorization with a self-supervised copy detection approach. We show a high degree of patient data memorization across all datasets, with approximately 37.2% of patient data detected as memorized and 68.7% of synthetic samples identified as patient data copies. Latent diffusion models are more susceptible to memorization than autoencoders and generative adversarial networks, and they outperform non-diffusion models in synthesis quality. Augmentation strategies during training, small architecture size and increasing datasets can reduce memorization, while overtraining the models can enhance it. These results emphasize the importance of carefully training generative models on private medical imaging datasets and examining the synthetic data to ensure patient privacy.

Medicine is one of the most sensitive fields in which artificial intelligence (AI) is extensively used, spanning from medical image analysis to clinical support. Specifically, in medicine, where every decision may severely affect human lives, the issue of ensuring that AI systems operate ethically and produce results that align with ethical considerations is of great importance. In this work, we investigate the combination of several key parameters on the performance of artificial neural networks (ANNs) used for medical image analysis in the presence of data corruption or errors. For this purpose, we examined five different ANN architectures (AlexNet, LeNet 5, VGG16, ResNet-50, and Vision Transformers - ViT), and for each architecture, we checked its performance under varying combinations of training dataset sizes and percentages of images that are corrupted through mislabeling. The image mislabeling simulates deliberate or nondeliberate changes to the dataset, which may cause the AI system to produce unreliable results. We found that the five ANN architectures produce different results for the same task, both for cases with and without dataset modification, which implies that the selection of which ANN architecture to implement may have ethical aspects that need to be considered. We also found that label corruption resulted in a mixture of performance metrics tendencies, indicating that it is difficult to conclude whether label corruption has occurred. Our findings demonstrate the relation between ethics in AI and ANN architecture implementation and AI computational parameters used therefor, and raise awareness of the need to find appropriate ways to determine whether label corruption has occurred.

Gliomas exhibit a high recurrence rate, particularly in the peritumoural brain zone after surgery. This study aims to develop and validate a radiomics-based model using preoperative fluid-attenuated inversion recovery (FLAIR) and T1-weighted contrast-enhanced (T1-CE) magnetic resonance imaging (MRI) sequences to predict glioma recurrence within specific quadrants of the surgical margin. In this retrospective study, 149 patients with confirmed glioma recurrence were included. 23 cases of data from Guizhou Medical University were used as a test set, and the remaining data were randomly used as a training set (70%) and a validation set (30%). Two radiologists from the research group established a Cartesian coordinate system centred on the tumour, based on FLAIR and T1-CE MRI sequences, dividing the tumour into four quadrants. Recurrence in each quadrant after surgery was assessed, categorising preoperative tumour quadrants as recurrent and non-recurrent. Following the division of tumours into quadrants and the removal of outliers, These quadrants were assigned to a training set (105 non-recurrence quadrants and 226 recurrence quadrants), a verification set (45 non-recurrence quadrants and 97 recurrence quadrants) and a test set (16 non-recurrence quadrants and 68 recurrence quadrants). Imaging features were extracted from preoperative sequences, and feature selection was performed using least absolute shrinkage and selection operator. Machine learning models included support vector machine, random forest, extra trees, and XGBoost. Clinical efficacy was evaluated through model calibration and decision curve analysis. The fusion model, which combines features from FLAIR and T1-CE sequences, exhibited higher predictive accuracy than single-modality models. Among the models, the LightGBM model demonstrated the highest predictive accuracy, with an area under the curve of 0.906 in the training set, 0.832 in the validation set and 0.805 in the test set. The study highlights the potential of a multimodal radiomics approach for predicting glioma recurrence, with the fusion model serving as a robust tool for clinical decision-making.

Enhancing the accuracy of tumor response predictions enables the development of tailored therapeutic strategies for patients with breast cancer. In this study, we developed deep radiomic models to enhance the prediction of chemotherapy response after the first treatment cycle. 18F-Fludeoxyglucose PET/CT imaging data and clinical record from 60 breast cancer patients were retrospectively obtained from the Cancer Imaging Archive. PET/CT scans were conducted at three distinct stages of treatment; prior to the initiation of chemotherapy (T1), following the first cycle of chemotherapy (T2), and after the full chemotherapy regimen (T3). The patient's primary gross tumor volume (GTV) was delineated on PET images using a 40% threshold of the maximum standardized uptake value (SUVmax). Radiomic features were extracted from the GTV based on the PET/CT images. In addition, a squeeze-and-excitation network (SENet) deep learning model was employed to generate additional features from the PET/CT images for combined analysis. A XGBoost machine learning model was developed and compared with the conventional machine learning algorithm [random forest (RF), logistic regression (LR) and support vector machine (SVM)]. The performance of each model was assessed using receiver operating characteristics area under the curve (ROC AUC) analysis, and prediction accuracy in a validation cohort. Model performance was evaluated through fivefold cross-validation on the entire cohort, with data splits stratified by treatment response categories to ensure balanced representation. The AUC values for the machine learning models using only radiomic features were 0.85(XGBoost), 0.76 (RF), 0.80 (LR), and 0.59 (SVM), with XGBoost showing the best performance. After incorporating additional deep learning-derived features from SENet, the AUC values increased to 0.92, 0.88, 0.90, and 0.61, respectively, demonstrating significant improvements in predictive accuracy. Predictions were based on pre-treatment (T1) and post-first-cycle (T2) imaging data, enabling early assessment of chemotherapy response after the initial treatment cycle. Integrating deep learning-derived features significantly enhanced the performance of predictive models for chemotherapy response in breast cancer patients. This study demonstrated the superior predictive capability of the XGBoost model, emphasizing its potential to optimize personalized therapeutic strategies by accurately identifying patients unlikely to respond to chemotherapy after the first treatment cycle.

The volume of image data generated in the medical field is continuously increasing. Manual annotation is both costly and prone to human error. Additionally, deep learning-based medical image algorithms rely on large, accurately annotated training datasets, which are expensive to produce and often result in instability. This study introduces LR-COBRAS, an interactive computer-aided data annotation algorithm designed for medical experts. LR-COBRAS aims to assist healthcare professionals in achieving more precise annotation outcomes through interactive processes, thereby optimizing medical image annotation tasks. The algorithm enhances must-link and cannot-link constraints during interactions through a logic reasoning module. It automatically generates potential constraint relationships, reducing the frequency of user interactions and improving clustering accuracy. By utilizing rules such as symmetry, transitivity, and consistency, LR-COBRAS effectively balances automation with clinical relevance. Experimental results based on the MedMNIST＋ dataset and ChestX-ray8 dataset demonstrate that LR-COBRAS significantly outperforms existing methods in clustering accuracy, efficiency, and interactive burden, showcasing superior robustness and applicability. This algorithm provides a novel solution for intelligent medical image analysis. The source code for our implementation is available on https://github.com/cjw-bbxc/MILR-COBRAS.