Foundation models have recently gained tremendous popularity in medical image analysis. State-of-the-art methods leverage either paired image-text data via vision-language pre-training or unpaired image data via self-supervised pre-training to learn foundation models with generalizable image features to boost downstream task performance. However, learning foundation models exclusively on either paired or unpaired image data limits their ability to learn richer and more comprehensive image features. In this paper, we investigate a novel task termed semi-supervised vision-language pre-training, aiming to fully harness the potential of both paired and unpaired image data for foundation model learning. To this end, we propose MaskedCLIP, a synergistic masked image modeling and contrastive language-image pre-training framework for semi-supervised vision-language pre-training. The key challenge in combining paired and unpaired image data for learning a foundation model lies in the incompatible feature spaces derived from these two types of data. To address this issue, we propose to connect the masked feature space with the CLIP feature space with a bridge transformer. In this way, the more semantic specific CLIP features can benefit from the more general masked features for semantic feature extraction. We further propose a masked knowledge distillation loss to distill semantic knowledge of original image features in CLIP feature space back to the predicted masked image features in masked feature space. With this mutually interactive design, our framework effectively leverages both paired and unpaired image data to learn more generalizable image features for downstream tasks. Extensive experiments on retinal image analysis demonstrate the effectiveness and data efficiency of our method.

To develop and assess the performance of an anatomically based multitask deep learning radiomics nomogram (AMDRN) system to predict implant failure risk before maxillary sinus floor elevation (MSFE) while incorporating automated segmentation of key anatomical structures. We retrospectively collected patients' preoperative cone beam computed tomography (CBCT) images and electronic medical records (EMRs). First, the nn-UNet v2 model was optimized to segment the maxillary sinus (MS), Schneiderian membrane (SM), and residual alveolar bone (RAB). Based on the segmentation mask, a deep learning model (3D-Attention-ResNet) and a radiomics model were developed to extract 3D features from CBCT scans, generating the DL Score, and Rad Score. Significant clinical features were also extracted from EMRs to build a clinical model. These components were then integrated using logistic regression (LR) to create the AMDRN model, which includes a visualization module to support clinical decision-making. Segmentation results for MS, RAB, and SM achieved high DICE coefficients on the test set, with values of 99.50% ± 0.84%, 92.53% ± 3.78%, and 91.58% ± 7.16%, respectively. On an independent test set, the Clinical model, Radiomics model, 3D-DL model, and AMDRN model achieved prediction accuracies of 60%, 76%, 82%, and 90%, respectively, with AMDRN achieving the highest AUC of 93%. The AMDRN system enables efficient preoperative prediction of implant failure risk in MSFE and accurate segmentation of critical anatomical structures, supporting personalized treatment planning and clinical risk management.

The study investigates the correlation between CD3 T-cell expression levels and cervical cancer (CC) while developing a magnetic resonance (MR) imaging-based radiomics model for preoperative prediction of CD3 T-cell expression levels. Prognostic correlations between CD3D, CD3E, and CD3G gene expressions and various cancers were analyzed using the Cancer Genome Atlas (TCGA) database. Protein-protein interaction (PPI) analysis via the STRING database identified associations between these genes and T lymphocyte activity. Gene Set Enrichment Analysis (GSEA) revealed immune pathway enrichment by categorizing genes based on CD3D expression levels. Correlations between immune checkpoint molecules and CD3 complex genes were also assessed. The study retrospectively included 202 patients with pathologically confirmed early-stage CC who underwent preoperative MRI, divided into training and test groups. Radiomic features were extracted from the whole-lesion tumor region of interest (ROI_tumor) and from peritumoral regions with 3 mm and 5 mm margins (ROI_3mm and ROI_5mm, respectively). Various machine learning algorithms, including Support Vector Machine (SVM), Logistic Regression, Random Forest, AdaBoost, and Decision Tree, were used to construct radiomics models based on different ROIs, and diagnostic performances were compared to identify the optimal approach. The best-performing algorithm was combined with intra- and peritumoral features and clinically relevant independent risk factors to develop a comprehensive predictive model. Analysis of the TCGA database demonstrated significant associations between CD3D, CD3E, and CD3G expressions and several cancers, including CC (p < 0.05). PPI analysis highlighted connections between these genes and T lymphocyte function, while GSEA indicated enrichment of immune-related pathways linked to CD3D. Immune checkpoint correlations showed positive associations with CD3 complex genes. Radiomics analysis selected 18 features from ROI_tumor and ROI_3mm across MRI sequences. The SVM algorithm achieved the highest predictive performance for CD3 T-cell expression status, with an area under the curve (AUC) of 0.93 in the training group and 0.92 in the test group. This MR-based radiomics model effectively predicts CD3 expression status in patients with early-stage CC, offering a non-invasive tool for preoperative assessment of CD3 expression, but its clinical utility needs further prospective validation.

Meningitis diagnosis requires a lumbar puncture (LP) to obtain cerebrospinal fluid (CSF) for a laboratory-based analysis. In high-income settings, LPs are part of the systematic approach to screen for meningitis, and most yield negative results. In low- and middle-income settings, LPs are seldom performed, and suspected cases are often treated empirically. The aim of this study was to validate a non-invasive transfontanellar white blood cell (WBC) counter in CSF to screen for meningitis. We conducted a prospective study across three Spanish hospitals, one Mozambican and one Moroccan hospital (2020-2023). We included patients under 24 months with suspected meningitis, an open fontanelle, and a LP performed within 24 h from recruitment. High-resolution-ultrasound (HRUS) images of the CSF were obtained using a customized probe. A deep-learning model was trained to classify CSF patterns based on LPs WBC counts, using a 30cells/mm³ threshold. The algorithm was applied to 3782 images from 76 patients. It correctly classified 17/18 CSFs with <math xmlns="http://www.w3.org/1998/Math/MathML"><mo>≥</mo></math> 30 WBC, and 55/58 controls (sensitivity 94.4%, specificity 94.8%). The only false negative was paired to a traumatic LP with 40 corrected WBC/mm³. This non-invasive device could be an accurate tool for screening meningitis in neonates and young infants, modulating LP indications. Our non-invasive, high-resolution ultrasound device achieved 94% accuracy in detecting elevated leukocyte counts in neonates and infants with suspected meningitis, compared to the gold standard (lumbar punctures and laboratory analysis). This first-in-class screening device introduces the first non-invasive method for neonatal and infant meningitis screening, potentially modulating lumbar puncture indications. This technology could substantially reduce lumbar punctures in low-suspicion cases and provides a viable alternative critically ill patients worldwide or in settings where lumbar punctures are unfeasible, especially in low-income countries).

The study assessed the feasibility of using synthetic data to fine-tune various open-source LLMs for free text to structured data conversation in radiology, comparing their performance with GPT models. A training set of 3000 synthetic thyroid nodule dictations was generated to train six open-source models (Starcoderbase-1B, Starcoderbase-3B, Mistral-7B, Llama-3-8B, Llama-2-13B, and Yi-34B). ACR TI-RADS template was the target model output. The model performance was tested on 50 thyroid nodule dictations from MIMIC-III patient dataset and compared against 0-shot, 1-shot, and 5-shot performance of GPT-3.5 and GPT-4. GPT-4 5-shot and Yi-34B showed the highest performance with no statistically significant difference between the models. Various open models outperformed GPT models with statistical significance. Overall, models trained with synthetic data showed performance comparable to GPT models in structured text conversion in our study. Given privacy preserving advantages, open LLMs can be utilized as a viable alternative to proprietary GPT models.

Noninvasive and precise tools for treatment response estimation in patients with high-risk hepatocellular carcinoma (HCC) who could benefit from hepatic arterial infusion chemotherapy (HAIC) plus lenvatinib and humanized programmed death receptor-1 inhibitors (PD-1) (HAIC-LEN-PD1) are lacking. This study aimed to evaluate the predictive potential of intratumoral and peritumoral radiomics for preoperative treatment response assessment to HAIC-LEN-PD1 in high-risk HCC cases. Totally 630 high-risk HCC cases administered HAIC-LEN-PD1 at three institutions were retrospectively identified and assigned to training, validation and external test sets. Totally 1834 radiomic features were, respectively, obtained from intratumoral and peritumoral regions and radiomics models were established using five classifiers. Based on the optimal model, a nomogram was developed and evaluated using areas under the curves (AUCs), calibration curves and decision curve analysis (DCA). Overall survival (OS) and progression-free survival (PFS) were assessed by Kaplan-Meier curves. The Intratumoral + Peritumoral 10 mm (Intra + Peri10) radiomics models were superior to the intratumor models and peritumor models, with AUCs of 0.919 (95%CI 0.889-0.949) in the training set, 0.874 (95%CI 0.812-0.936) in validation set and 0.893 (95%CI 0.839-0.948) in external test sets. The nomogram had good calibration ability and clinical value, with the AUCs of 0.936 (95%CI 0.907-0.965) in the training set, 0.878 (95%CI 0.916-0.940) in validation set and 0.902 (95%CI 0.848-0.957) in external test sets. The Kaplan-Meier analysis showed that high-score patients had significantly shorter OS and PFS than the low-score patients (median OS: 11.7 vs. 29.6 months, the whole set, p < 0.001; median PFS: 6.0 vs. 12.0 months, the whole set, p < 0.001). The Intra + Peri10 model can effectively predict the treatment response of high-risk HCC cases administered HAIC-LEN-PD1. The nomogram could provide an effective tool to evaluate the treatment response and risk stratification.

Pseudo-healthy image inpainting is an essential preprocessing step for analyzing pathological brain MRI scans. Most current inpainting methods favor slice-wise 2D models for their high in-plane fidelity, but their independence across slices produces discontinuities in the volume. Fully 3D models alleviate this issue, but their high model capacity demands extensive training data for reliable, high-fidelity synthesis -- often impractical in medical settings. We address these limitations with a hierarchical diffusion framework by replacing direct 3D modeling with two perpendicular coarse-to-fine 2D stages. An axial diffusion model first yields a coarse, globally consistent inpainting; a coronal diffusion model then refines anatomical details. By combining perpendicular spatial views with adaptive resampling, our method balances data efficiency and volumetric consistency. Our experiments show our approach outperforms state-of-the-art baselines in both realism and volumetric consistency, making it a promising solution for pseudo-healthy image inpainting. Code is available at https://github.com/dou0000/3dMRI-Consistent-Inpaint.

Recent advances in deep learning have led to robust automated tools for segmentation of abdominal computed tomography (CT). Meanwhile, segmentation of magnetic resonance imaging (MRI) is substantially more challenging due to the inherent signal variability and the increased effort required for annotating training datasets. Hence, existing approaches are trained on limited sets of MRI sequences, which might limit their generalizability. To characterize the landscape of MRI abdominal segmentation tools, we present here a comprehensive benchmarking of the three state-of-the-art and open-source models: MRSegmentator, MRISegmentator-Abdomen, and TotalSegmentator MRI. Since these models are trained using labor-intensive manual annotation cycles, we also introduce and evaluate ABDSynth, a SynthSeg-based model purely trained on widely available CT segmentations (no real images). More generally, we assess accuracy and generalizability by leveraging three public datasets (not seen by any of the evaluated methods during their training), which span all major manufacturers, five MRI sequences, as well as a variety of subject conditions, voxel resolutions, and fields-of-view. Our results reveal that MRSegmentator achieves the best performance and is most generalizable. In contrast, ABDSynth yields slightly less accurate results, but its relaxed requirements in training data make it an alternative when the annotation budget is limited. The evaluation code and datasets are given for future benchmarking at https://github.com/deepakri201/AbdoBench, along with inference code and weights for ABDSynth.

Detecting Pulmonary Embolism (PE) is critical for effective patient care, and Artificial Intelligence (AI) has shown promise in supporting radiologists in this task. Integrating AI into radiology workflows requires not only evaluation of its diagnostic accuracy but also assessment of its acceptance among clinical staff. This study aims to evaluate the performance of an AI algorithm in detecting pulmonary embolisms (PEs) on contrast-enhanced computed tomography pulmonary angiograms (CTPAs) and to assess the level of acceptance of the algorithm among radiology department staff. This retrospective study analyzed anonymized computed tomography pulmonary angiography (CTPA) data from a university clinic. Surveys were conducted at three and nine months after the implementation of a commercially available AI algorithm designed to flag CTPA scans with suspected PE. A thoracic radiologist and a cardiac radiologist served as the reference standard for evaluating the performance of the algorithm. The AI analyzed 59 CTPA cases during the initial evaluation and 46 cases in the follow-up assessment. In the first evaluation, the AI algorithm demonstrated a sensitivity of 84.6% and a specificity of 94.3%. By the second evaluation, its performance had improved, achieving a sensitivity of 90.9% and a specificity of 96.7%. Radiologists' acceptance of the AI tool increased over time. Nevertheless, despite this growing acceptance, many radiologists expressed a preference for hiring an additional physician over adopting the AI solution if the costs were comparable. Our study demonstrated high sensitivity and specificity of the AI algorithm, with improved performance over time and a reduced rate of unanalyzed scans. These improvements likely reflect both algorithmic refinement and better data integration. Departmental feedback indicated growing user confidence and trust in the tool. However, many radiologists continued to prefer the addition of a resident over reliance on the algorithm. Overall, the AI showed promise as a supportive "second-look" tool in emergency radiology settings. The AI algorithm demonstrated diagnostic performance comparable to that reported in similar studies for detecting PE on CTPA, with both sensitivity and specificity showing improvement over time. Radiologists' acceptance of the algorithm increased throughout the study period, underscoring its potential as a complementary tool to physician expertise in clinical practice.

Accurate preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) remains challenging. Current imaging biomarkers show limited predictive performance. To develop a deep learning model based on preoperative multiphase CT images of tumors and lesser omental adipose tissue (LOAT) for predicting MVI status and to analyze associated survival outcomes. This retrospective study included pathologically confirmed HCC patients from two medical centers between 2016 and 2023. A dual-branch feature fusion model based on ResNet18 was constructed, which extracted fused features from dual-phase CT images of both tumors and LOAT. The model's performance was evaluated on both internal and external test sets. Logistic regression was used to identify independent predictors of MVI. Based on MVI status, patients in the training, internal test, and external test cohorts were stratified into high- and low-risk groups, and overall survival differences were analyzed. The model incorporating LOAT features outperformed the tumor-only modality, achieving an AUC of 0.889 (95% CI: [0.882, 0.962], P=0.004) in the internal test set and 0.826 (95% CI: [0.793, 0.872], P=0.006) in the external test set. Both results surpassed the independent diagnoses of three radiologists (average AUC=0.772). Multivariate logistic regression confirmed that maximum tumor diameter and LOAT area were independent predictors of MVI. Further Cox regression analysis showed that MVI-positive patients had significantly increased mortality risks in both the internal test set (Hazard Ratio [HR]=2.246, 95% CI: [1.088, 4.637], P=0.029) and external test set (HR=3.797, 95% CI: [1.262, 11.422], P=0.018). This study is the first to use a deep learning framework integrating LOAT and tumor imaging features, improving preoperative MVI risk stratification accuracy. Independent prognostic value of LOAT has been validated in multicenter cohorts, highlighting its potential to guide personalized surgical planning.