Based on preoperative clinical text data and lumbar magnetic resonance imaging (MRI), we applied machine learning (ML) algorithms to construct a model that would predict early recurrence in lumbar-disc herniation (LDH) patients who underwent percutaneous endoscopic lumbar discectomy (PELD). We then explored the clinical performance of this prognostic prediction model via multimodal-data fusion. Clinical text data and radiological images of LDH patients who underwent PELD at the Intervertebral Disc Center of the Affiliated Hospital of Gansu University of Traditional Chinese Medicine (AHGUTCM; Lanzhou, China) were retrospectively collected. Two radiologists with clinical-image reading experience independently outlined regions of interest (ROI) on the MRI images and extracted radiomic features using 3D Slicer software. We then randomly separated the samples into a training set and a test set at a 7:3 ratio, used eight ML algorithms to construct predictive radiomic-feature models, evaluated model performance by the area under the curve (AUC), and selected the optimal model for screening radiomic features and calculating radiomic scores (Rad-scores). Finally, after using logistic regression to construct a nomogram for predicting the early-recurrence rate, we evaluated the nomogram's clinical applicability using a clinical-decision curve. We initially extracted 851 radiomic features. After constructing our models, we determined based on AUC values that the optimal ML algorithm was least absolute shrinkage and selection operator (LASSO) regression, which had an AUC of 0.76 and an accuracy rate of 91%. After screening features using the LASSO model, we predicted Rad-score for each sample of recurrent LDH using nine radiomic features. Next, we fused three of these clinical features -age, diabetes, and heavy manual labor-to construct a nomogram with an AUC of 0.86 (95% confidence interval [CI], 0.79-0.94). Analysis of the clinical-decision and impact curves showed that the prognostic prediction model with multimodal-data fusion had good clinical validity and applicability. We developed and analyzed a prognostic prediction model for LDH with multimodal-data fusion. Our model demonstrated good performance in predicting early postoperative recurrence in LDH patients; therefore, it has good prospects for clinical application and can provide clinicians with objective, accurate information to help them decide on presurgical treatment plans. However, external-validation studies are still needed to further validate the model's comprehensive performance and improve its generalization and extrapolation.

Despite its high negative predictive value (NPV) for clinically significant prostate cancer (csPCa), MRI suffers from a substantial number of false positives, especially for intermediate-risk cases. In this work, we determine whether a deep learning model trained with PI-RADS-guided representation learning can disambiguate the PI-RADS 3 classification, detect csPCa from bi-parametric prostate MR images, and avoid unnecessary benign biopsies. This study included 28,263 MR examinations and radiology reports from 21,938 men imaged for known or suspected prostate cancer between 2015 and 2023 at our institution (21 imaging locations with 34 readers), with 6352 subsequent biopsies. We trained a deep learning model, a representation learner (RL), to learn how radiologists interpret conventionally acquired T2-weighted and diffusion-weighted MR images, using exams in which the radiologists are confident in their risk assessments (PI-RADS 1 and 2 for the absence of csPCa vs. PI-RADS 4 and 5 for the presence of csPCa, n=21,465). We then trained biopsy-decision models to detect csPCa (Gleason score ≥7) using these learned image representations, and compared them to the performance of radiologists, and of models trained on other clinical variables (age, prostate volume, PSA, and PSA density) for treatment-naïve test cohorts consisting of only PI-RADS 3 (n=253, csPCa=103) and all PI-RADS (n=531, csPCa=300) cases. On the 2 test cohorts (PI-RADS-3-only, all-PI-RADS), RL-based biopsy-decision models consistently yielded higher AUCs in detecting csPCa (AUC=0.73 [0.66, 0.79], 0.88 [0.85, 0.91]) compared with radiologists (equivocal, AUC=0.79 [0.75, 0.83]) and the clinical model (AUCs=0.69 [0.62, 0.75], 0.78 [0.74, 0.82]). In the PIRADS-3-only cohort, all of whom would be biopsied using our institution's standard of care, the RL decision model avoided 41% (62/150) of benign biopsies compared with the clinical model (26%, P<0.001), and improved biopsy yield by 10% compared with the PI-RADS ≥3 decision strategy (0.50 vs. 0.40). Furthermore, on the all-PI-RADS cohort, RL decision model avoided 27% of additional benign biopsies (138/231) compared to radiologists (33%, P<0.001) with comparable sensitivity (93% vs. 92%), higher NPV (0.87 vs. 0.77), and biopsy yield (0.75 vs. 0.64). The combination of clinical and RL decision models further avoided benign biopsies (46% in PI-RADS-3-only and 62% in all-PI-RADS) while improving NPV (0.82, 0.88) and biopsy yields (0.52, 0.76) across the 2 test cohorts. Our PI-RADS-guided deep learning RL model learns summary representations from bi-parametric prostate MR images that can provide additional information to disambiguate intermediate-risk PI-RADS 3 assessments. The resulting RL-based biopsy decision models also outperformed radiologists in avoiding benign biopsies while maintaining comparable sensitivity to csPCa for the all-PI-RADS cohort. Such AI models can easily be integrated into clinical practice to supplement radiologists' reads in general and improve biopsy yield for any equivocal decisions.

Hepatocellular carcinoma (HCC) is the fifth most common cancer worldwide and the third most common cause of cancer-related death. Cirrhosis is a major contributing factor, accounting for over 90% of HCC cases. With the high mortality rate of HCC, earlier detection of HCC is critical. When added to magnetic resonance imaging (MRI), artificial intelligence (AI) has been shown to improve HCC detection. Nonetheless, to date no cost-effectiveness analyses have been conducted on an AI tool to enhance earlier HCC detection. This study reports on the cost-effectiveness of detection of liver lesions with AI improved MRI in the surveillance for HCC in patients with a cirrhotic liver compared to usual care (UC). The model structure included a decision tree followed by a state-transition Markov model from an Italian healthcare perspective. Lifetime costs and quality-adjusted life years (QALY) were simulated in cirrhotic patients at risk of HCC. One-way sensitivity analyses and two-way sensitivity analyses were performed. Results were presented as incremental cost-effectiveness ratios (ICER). For patients receiving UC, the average lifetime costs per 1,000 patients were €16,604,800 compared to €16,610,250 for patients receiving the AI approach. With a QALY gained of 0.55 and incremental costs of €5,000 for every 1,000 patients, the ICER was €9,888 per QALY gained, indicating cost-effectiveness with the willingness-to-pay threshold of €33,000/QALY gained. Main drivers of cost-effectiveness included the cost and performance (sensitivity and specificity) of the AI tool. This study suggests that an AI-based approach to earlier detect HCC in cirrhotic patients can be cost-effective. By incorporating cost-effective AI-based approaches in clinical practice, patient outcomes and healthcare efficiency are improved.

Once deployed, medical image analysis methods are often faced with unexpected image corruptions and noise perturbations. These unknown covariate shifts present significant challenges to deep learning based methods trained on "clean" images. This often results in unreliable predictions and poorly calibrated confidence, hence hindering clinical applicability. While recent methods have been developed to address specific issues such as confidence calibration or adversarial robustness, no single framework effectively tackles all these challenges simultaneously. To bridge this gap, we propose LaDiNE, a novel ensemble learning method combining the robustness of Vision Transformers with diffusion-based generative models for improved reliability in medical image classification. Specifically, transformer encoder blocks are used as hierarchical feature extractors that learn invariant features from images for each ensemble member, resulting in features that are robust to input perturbations. In addition, diffusion models are used as flexible density estimators to estimate member densities conditioned on the invariant features, leading to improved modeling of complex data distributions while retaining properly calibrated confidence. Extensive experiments on tuberculosis chest X-rays and melanoma skin cancer datasets demonstrate that LaDiNE achieves superior performance compared to a wide range of state-of-the-art methods by simultaneously improving prediction accuracy and confidence calibration under unseen noise, adversarial perturbations, and resolution degradation.

Breast cancer is the leading cause of death among women worldwide, and early detection through the standardized BI-RADS framework helps physicians assess the risk of malignancy and guide appropriate diagnostic and treatment decisions. In this study, an interpretable deep learning model (BIScreener) was proposed for predicting BI-RADS classifications from breast ultrasound images, aiding in the accurate assessment of breast cancer risk and improving diagnostic efficiency. BIScreener utilizes the stacked generalization of three pretrained convolutional neural networks to analyze ultrasound images obtained from two specific instruments (Mindray R5 and HITACHI) used at local hospitals. BIScreener achieved a classification total accuracy of 90.0% and ROC-AUC value of 0.982 in the external test set for five BI-RADS categories. The proposed method achieved 83.8% classification total accuracy and 0.967 ROC-AUC value for seven BI-RADS categories. In addition, the model improved the diagnostic accuracy of two radiologists by more than 8.1% for five BI-RADS categories and by more than 4.8% for seven BI-RADS categories and reduced the explanation time by more than 19.0%, demonstrating its potential to accelerate and improve the breast cancer diagnosis process.

One of the most promising auxiliaries for screening breast cancer (BC) is ultrasound (US) radio-frequency (RF) time series. It has the superiority of not requiring any supplementary equipment over other methods. This article sought to propound a machine learning (ML) method for the automated categorization of breast lesions-categorized as benign, probably benign, suspicious, or malignant-using features extracted from the accumulated US RF time series. In this research, 220 data points of the categories as mentioned earlier, recorded from 118 patients, were analyzed. The RFTSBU dataset was registered by a SuperSonic Imagine Aixplorer® medical/research system fitted with a linear transducer. The expert radiologist manually selected regions of interest (ROIs) in B-mode images before extracting 283 features from each ROI in the ML approach, utilizing textural features such as Gabor filter (GF), gray-level co-occurrence matrix (GLCM), gray-level run-length matrix (GLRLM), gray-level size zone matrix (GLSZM), and gray-level dependence matrix (GLDM). Subsequently, the particle swarm optimization (PSO) narrowed the features to 131 highly effective ones. Ultimately, the features underwent classification using an innovative multi-origin method classification (MOMC), marking a significant leap in BC diagnosis. Employing 5-fold cross-validation, the study achieved notable accuracy rates of 98.57 ± 1.09%, 91.53 ± 0.89%, and 83.71 ± 1.30% for 2-, 3-, and 4-class classifications, respectively, using MOMC-SVM and MOMC-ensemble classifiers. This research introduces an innovative ML-based approach to differentiate between diverse breast lesion types using in vivo US RF time series data. The findings underscore its efficacy in enhancing classification accuracy, promising significant strides in computer-aided diagnosis (CAD) for BC screening.

This study aimed to investigate the role of volumetric and dissemination parameters obtained from pretreatment 18-fluorodeoxyglucose PET/computed tomography (18F-FDG PET/CT) in predicting progression/relapse in patients with diffuse large B-cell lymphoma (DLBCL) with machine learning algorithms. Patients diagnosed with DLBCL histopathologically, treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone, and followed for at least 1 year were reviewed retrospectively. Quantitative parameters such as tumor volume [total metabolic tumor volume (tMTV)], tumor burden [total lesion glycolysis (tTLG)], and the longest distance between two tumor foci (Dmax) were obtained from PET images with a standard uptake value threshold of 4.0. The MTV obtained from the volume of interest with the highest volume was noted as metabolic bulk volume (MBV). By analyzing the patients' PET parameters and clinical information with machine learning algorithms, models that attempt to predict progression/recurrence over 1 year were obtained. Of the 90 patients included, 16 had progression within 1 year. Significant differences were found in tMTV, tTLG, MBV, and Dmax values between patients with and without progression. The area under curve (AUC) of the model obtained with clinical data was 0.701. While a model with an AUC of 0.871 was obtained with a random forest algorithm using PET parameters, the model obtained with the Naive Bayes algorithm including clinical data in PET parameters had an AUC of 0.838. Using quantitative parameters derived from staging PET with machine learning algorithms may enable us to detect early progression in patients with DLBCL and improve early risk stratification and guide treatment decisions in these patients.

Motion artifacts are common for knee MRI, which usually lead to rescanning. Effective removal of motion artifacts would be clinically useful. To construct an effective deep learning-based model to remove motion artifacts for knee MRI using real-world data. Retrospective. Model construction: 90 consecutive patients (1997 2D slices) who had knee MRI images with motion artifacts paired with immediately rescanned images without artifacts served as ground truth. Internal test dataset: 25 patients (795 slices) from another period; external test dataset: 39 patients (813 slices) from another hospital. 3-T/1.5-T knee MRI with T1-weighted imaging, T2-weighted imaging, and proton-weighted imaging. A deep learning-based supervised conditional diffusion model was constructed. Objective metrics (root mean square error [RMSE], peak signal-to-noise ratio [PSNR], structural similarity [SSIM]) and subjective ratings were used for image quality assessment, which were compared with three other algorithms (enhanced super-resolution [ESR], enhanced deep super-resolution, and ESR using a generative adversarial network). Diagnostic performance of the output images was compared with the rescanned images. The Kappa Test, Pearson chi-square test, Fredman's rank-sum test, and the marginal homogeneity test. A p value < 0.05 was considered statistically significant. Subjective ratings showed significant improvements in the output images compared to the input, with no significant difference from the ground truth. The constructed method demonstrated the smallest RMSE (11.44  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  5.47 in the validation cohort; 13.95  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  4.32 in the external test cohort), the largest PSNR (27.61  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  3.20 in the validation cohort; 25.64  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  2.67 in the external test cohort) and SSIM (0.97  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  0.04 in the validation cohort; 0.94  <math xmlns="http://www.w3.org/1998/Math/MathML"> <semantics><mrow><mo>±</mo></mrow> <annotation>$$ \pm $$</annotation></semantics> </math>  0.04 in the external test cohort) compared to the other three algorithms. The output images achieved comparable diagnostic capability as the ground truth for multiple anatomical structures. The constructed model exhibited feasibility and effectiveness, and outperformed multiple other algorithms for removing motion artifacts in knee MRI. Level 3. Stage 2.

Artificial intelligence (AI) has enabled accurate and fast plaque quantification from coronary computed tomography angiography (CCTA). However, AI detects any coronary plaque in up to 97% of patients. To avoid overdiagnosis, a plaque burden safety cut-off for future coronary events is needed. Percent atheroma volume (PAV) was quantified with AI-guided quantitative computed tomography in a blinded fashion. Safety cut-off derivation was performed in the Turku CCTA registry (Finland), and pre-defined as ≥90% sensitivity for acute coronary syndrome (ACS). External validation was performed in the Amsterdam CCTA registry (the Netherlands). In the derivation cohort, 100/2271 (4.4%) patients experienced ACS (median follow-up 6.9 years). A threshold of PAV ≥ 2.6% was derived with 90.0% sensitivity and negative predictive value (NPV) of 99.0%. In the validation cohort 27/568 (4.8%) experienced ACS (median follow-up 6.7 years) with PAV ≥ 2.6% showing 92.6% sensitivity and 99.0% NPV for ACS. In the derivation cohort, 45.2% of patients had PAV < 2.6 vs. 4.3% with PAV 0% (no plaque) (P < 0.001) (validation cohort: 34.3% PAV < 2.6 vs. 2.6% PAV 0%; P < 0.001). Patients with PAV ≥ 2.6% had higher adjusted ACS rates in the derivation [Hazard ratio (HR) 4.65, 95% confidence interval (CI) 2.33-9.28, P < 0.001] and validation cohort (HR 7.31, 95% CI 1.62-33.08, P = 0.010), respectively. This study suggests that PAV up to 2.6% quantified by AI is associated with low-ACS risk in two independent patient cohorts. This cut-off may be helpful for clinical application of AI-guided CCTA analysis, which detects any plaque in up to 96-97% of patients.

Thyroid ultrasound (US) is an essential tool for detecting and characterizing thyroid nodules. In this study, we propose an innovative approach to enhance thyroid nodule assessment by integrating Doppler US images with grayscale US images through weakly supervised data augmentation networks (WSDAN). Our method reduces background noise by replacing inefficient augmentation strategies, such as random cropping, with an advanced technique guided by bounding boxes derived from Doppler US images. This targeted augmentation significantly improves model performance in both classification and localization of thyroid nodules. The training dataset comprises 1288 paired grayscale and Doppler US images, with an additional 190 pairs used for three-fold cross-validation. To evaluate the model's efficacy, we tested it on a separate set of 190 grayscale US images. Compared to five state-of-the-art models and the original WSDAN, our Enhanced WSDAN model achieved superior performance. For classification, it reached an accuracy of 91%. For localization, it achieved Dice and Jaccard indices of 75% and 87%, respectively, demonstrating its potential as a valuable clinical tool.