This study aimed to develop a predictive model for lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC) patients by deep learning radiomic (DLRad) and clinical features.

This study included 271 thyroid lobes from 228 PTC patients who underwent preoperative neck non-contrast CT at Center 1 (May 2021-April 2024). LNM status was confirmed via postoperative pathology, with each thyroid lobe labeled accordingly. The cohort was divided into training (n = 189) and validation (n = 82) cohorts, with additional temporal (n = 59 lobes, Center 1, May-August 2024) and external (n = 66 lobes, Center 2) test cohorts. Thyroid lobes were manually segmented from the isthmus midline, ensuring interobserver consistency (ICC ≥ 0.8). Deep learning and radiomics features were selected using LASSO algorithms to compute DLRad scores. Logistic regression identified independent predictors, forming DLRad, clinical, and combined models. Model performance was evaluated using AUC, calibration, decision curves, and the DeLong test, compared against radiologists' assessments.

Independent predictors of LNM included age, gender, multiple nodules, tumor size group, and DLRad. The combined model demonstrated superior diagnostic performance with AUCs of 0.830 (training), 0.799 (validation), 0.819 (temporal test), and 0.756 (external test), outperforming the DLRad model (AUCs: 0.786, 0.730, 0.753, 0.642), clinical model (AUCs: 0.723, 0.745, 0.671, 0.660), and radiologist evaluations (AUCs: 0.529, 0.606, 0.620, 0.503). It also achieved the lowest Brier scores (0.167, 0.184, 0.175, 0.201) and the highest net benefit in decision-curve analysis at threshold probabilities > 20 %.

The combined model integrating DLRad and clinical features exhibits good performance in predicting LNM in PTC patients.

Medical image analysis, empowered by artificial intelligence (AI), plays a crucial role in modern healthcare diagnostics. However, the effectiveness of machine learning models hinges on their ability to generalize to diverse patient populations, presenting domain shift challenges. This study explores the domain shift problem in chest X-ray classification, focusing on cross-population variations, especially in underrepresented groups. We analyze the impact of domain shifts across three population datasets acting as sources using a Nigerian chest X-ray dataset acting as the target. Model performance is evaluated to assess disparities between source and target populations, revealing large discrepancies when the models trained on a source were applied to the target domain. To address with the evident domain shift among the populations, we propose a supervised adversarial domain adaptation (ADA) technique. The feature extractor is first trained on the source domain using a supervised loss function in ADA. The feature extractor is then frozen, and an adversarial domain discriminator is introduced to distinguish between the source and target domains. Adversarial training fine-tunes the feature extractor, making features from both domains indistinguishable, thereby creating domain-invariant features. The technique was evaluated on the Nigerian dataset, showing significant improvements in chest X-ray classification performance. The proposed model achieved a 90.08% accuracy and a 96% AUC score, outperforming existing approaches such as multi-task learning (MTL) and continual learning (CL). This research highlights the importance of developing domain-aware models in AI-driven healthcare, offering a solution to cross-population domain shift challenges in medical imaging.

Brain tumors are incredibly harmful and can drastically reduce life expectancy. Most researchers use magnetic resonance (MR) scans to detect tumors because they can provide detailed images of the affected area. Recently, AI-based deep learning methods have emerged to enhance diagnostic accuracy through efficient data processing. This study investigates the effectiveness of deep transfer learning techniques for accurate brain tumor diagnosis. A preprocessing pipeline is used to enhance the image quality. This pipeline includes morphological operations such as erosion and dilation for shape refinement, Gaussian blurring for noise reduction, and thresholding for image cropping. Principal Component Analysis (PCA) is applied for dimensionality reduction, and data augmentation enriches the dataset. The dataset is partitioned into training (80 %) and testing (20 %). Pretrained ResNet152 and GoogleNet extract meaningful features from the images. These extracted features are then classified using conventional machine learning classifiers: Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Classification and Regression Trees (CART), and Gaussian Naive Bayes (GNB). This study compares the performance of two pre-trained models for medical image analysis. Performance metrics such as accuracy, sensitivity, recall, and F1-Score evaluate the final classification results. ResNet152 outperforms GoogleNet, achieving a 98.53 % accuracy, an F1 score of 97.4 %, and a sensitivity of 96.52 %. This study highlights integrating deep learning and traditional machine-learning techniques in medical image analysis for effective brain tumor detection.

In recent years, X-ray low-dose computed tomography (LDCT) has garnered widespread attention due to its significant reduction in the risk of patient radiation exposure. However, LDCT images often contain a substantial amount of noises, adversely affecting diagnostic quality. To mitigate this, a plethora of LDCT denoising methods have been proposed. Among them, deep learning (DL) approaches have emerged as the most effective, due to their robust feature extraction capabilities. Yet, the prevalent use of supervised training paradigms is often impractical due to the challenges in acquiring low-dose and normal-dose CT pairs in clinical settings. Consequently, unsupervised and self-supervised deep learning methods have been introduced for LDCT denoising, showing considerable potential for clinical applications. These methods' efficacy hinges on training strategies. Notably, there appears to be no comprehensive reviews of these strategies. Our review aims to address this gap, offering insights and guidance for researchers and practitioners. Based on training strategies, we categorize the LDCT methods into six groups: (i) cycle consistency-based, (ii) score matching-based, (iii) statistical characteristics of noise-based, (iv) similarity-based, (v) LDCT synthesis model-based, and (vi) hybrid methods. For each category, we delve into the theoretical underpinnings, training strategies, strengths, and limitations. In addition, we also summarize the open source codes of the reviewed methods. Finally, the review concludes with a discussion on open issues and future research directions.

Generative artificial intelligence (AI) has emerged as a transformative force in various fields, including anatomic pathology, where it offers the potential to significantly enhance diagnostic accuracy, workflow efficiency, and research capabilities.

To explore the applications, benefits, and challenges of generative AI in anatomic pathology, with a focus on its impact on diagnostic processes, workflow efficiency, education, and research.

A comprehensive review of current literature and recent advancements in the application of generative AI within anatomic pathology, categorized into unimodal and multimodal applications, and evaluated for clinical utility, ethical considerations, and future potential.

Generative AI demonstrates significant promise in various domains of anatomic pathology, including diagnostic accuracy enhanced through AI-driven image analysis, virtual staining, and synthetic data generation; workflow efficiency, with potential for improvement by automating routine tasks, quality control, and reflex testing; education and research, facilitated by AI-generated educational content, synthetic histology images, and advanced data analysis methods; and clinical integration, with preliminary surveys indicating cautious optimism for nondiagnostic AI tasks and growing engagement in academic settings. Ethical and practical challenges require rigorous validation, prompt engineering, federated learning, and synthetic data generation to help ensure trustworthy, reliable, and unbiased AI applications. Generative AI can potentially revolutionize anatomic pathology, enhancing diagnostic accuracy, improving workflow efficiency, and advancing education and research. Successful integration into clinical practice will require continued interdisciplinary collaboration, careful validation, and adherence to ethical standards to ensure the benefits of AI are realized while maintaining the highest standards of patient care.

Increased monitoring of health-related data for ICU patients holds great potential for the early prediction of medical outcomes. Research on whether the use of clinical notes and concepts from knowledge bases can improve the performance of prediction models is limited. We investigated the effects of combining clinical variables, clinical notes, and clinical concepts. We focus on the early prediction of Acute Kidney Injury (AKI) in the intensive care unit (ICU). AKI is a sudden reduction in kidney function measured by increased serum creatinine (SCr) or decreased urine output. AKI may occur in up to 30% of ICU stays. We developed three models based on convolutional neural networks using data from the Medical Information Mart for Intensive Care (MIMIC) database. The models used clinical variables, free-text notes, and concepts from the Elsevier H-Graph. Our models achieved good predictive performance (AUROC 0.73-0.90). These models were assessed both when using Scr and urine output as predictors and when omitting them. When Scr and urine output were used as predictors, models that included clinical notes and concepts together with clinical variables performed on par with models that only used clinical variables. When excluding SCr and urine output, predictive performance improved by combining multiple modalities. The models that used only clinical variables were externally validated on the eICU dataset and transported fairly to the new population (AUROC 0.68-0.77). Our in-depth comparison of modalities and text representations may further guide researchers and practitioners in applying multimodal models for predicting AKI and inspire them to investigate multimodality and contextualized embeddings for other tasks. Our models can support clinicians to promptly recognize and treat deteriorating AKI patients and may improve patient outcomes in the ICU.

Accurate segmentation of the optic disk (OD) and optic cup (OC) regions of the optic nerve head is a critical step in glaucoma diagnosis. Existing architectures based on convolutional neural networks (CNNs) still suffer from insufficient global information and poor generalization ability to small sample datasets. Besides, advanced transformer-based models, although capable of capturing global image features, perform poorly in medical image segmentation due to numerous parameters and insufficient local spatial information. To address the above two problems, we propose an innovative W-shaped hybrid network framework, CC-TransXNet, which combines the advantages of CNN and transformer. Firstly, by employing TransXNet and improved ResNet as feature extraction modules, the network considers local and global features to enhance its generalization ability. Secondly, the convolutional block attention module (CBAM) is introduced in the residual structure to improve the ability to recognize the OD and OC by applying attention in both the channel and spatial dimensions. Thirdly, the Contextual Attention (CoT) self-attention mechanism is used in the skip connection to adaptively allocate attention to the contextual information, further enhancing the segmentation's accuracy. We conducted experiments on four publicly available datasets (REFUGE 2, RIM-ONE DL, GAMMA, and Drishti-GS). Compared with the traditional U-Net, CNN, and transformer-based networks, our proposed CC-TransXNet improves the segmentation accuracy and significantly enhances the generalization ability on small datasets. Moreover, CC-TransXNet effectively controls the number of parameters in the model through optimized design to avoid the risk of overfitting, proving its potential for efficient segmentation.

Lumbar foraminal stenosis (LFS) is a common spinal condition that requires accurate assessment. Current magnetic resonance imaging (MRI) reporting processes are often inefficient, and while deep learning has potential for improvement, challenges in generalization and interpretability limit its diagnostic effectiveness compared to physician expertise.

The present study aimed to leverage a multimodal large language model to improve the accuracy and efficiency of LFS image classification, thereby enabling rapid and precise automated diagnosis, reducing the dependence on manually annotated data, and enhancing diagnostic efficiency.

Retrospective study conducted from April 2017 to March 2023.

Sagittal T1-weighted MRI data for the lumbar spine were collected from 1,200 patients across three medical centers. A total of 810 patient cases were included in the final analysis, with data collected from seven different MRI devices.

Automated classification of LFS using the multi modal large language model. Accuracy, sensitivity, Specificity and Cohen's Kappa coefficient were calculated.

An advanced multimodal fusion framework GPT4LFS was developed with the primary objective of integrating imaging data and natural language descriptions to comprehensively capture the complex LFS features. The model employed a pre-trained ConvNeXt as the image processing module for extracting high-dimensional imaging features. Concurrently, medical descriptive texts generated by the multimodal large language model GPT-4o and encoded and feature-extracted using RoBERTa were utilized to optimize the model's contextual understanding capabilities. The Mamba architecture was implemented during the feature fusion stage, effectively integrating imaging and textual features and thereby enhancing the performance of the classification task. Finally, the stability of the model's detection results was validated by evaluating classification task metrics, such as the accuracy, sensitivity, specificity, and Kappa coefficients.

The training set comprised 6,299 images from 635 patients, the internal test set included 820 images from 82 patients, and the external test set was composed of 930 images from 93 patients. The GPT4LFS model demonstrated an overall accuracy of 93.7%, sensitivity of 95.8%, and specificity of 94.5% in the internal test set (Kappa = 0.89,95% confidence interval (CI): 0.84-0.96, p<.001). In the external test set, the overall accuracy was 92.2%, with a sensitivity of 92.2% and a specificity of 97.4% (Kappa = 0.88, 95% CI: 0.84-0.89, p<.001). Both the internal and external test sets showed excellent consistency in the model. After the article is published, we will make the full code publicly available on GitHub.

Using the GPT4LFS model for LFS image categorization demonstrated accuracy and the capacity for feature description at a level commensurate with that of professional clinicians.

Schizophrenia is a complicated mental condition marked by disruptions in thought processes, perceptions, and emotional responses, which can cause severe impairment in everyday functioning. sMRI is a non-invasive neuroimaging technology that visualizes the brain's structure while providing precise information on its anatomy and potential problems. This paper investigates the role of multidimensional Convolutional Neural Network (CNN) architectures: 1D-CNN, 2D-CNN and 3D-CNN, using the DWT subbands of sMRI data. 1D-CNN involves energy features extracted from the CD subband of sMRI data. The sum of gradient magnitudes of CD subband, known as energy feature, highlights diagonal high frequency elements associated with schizophrenia. 2D-CNN uses the CH subband decomposed by DWT that enables feature extraction from horizontal high frequency coefficients of sMRI data. In the case of 3D-CNNs, the CV subband is used which leads to volumetric feature extraction from vertical high frequency coefficients. Feature extraction in DWT domain explores textural changes, edges, coarse and fine details present in sMRI data from which the multidimensional feature extraction is carried out for classification.Through maximum voting technique, the proposed model optimizes schizophrenia classification from the multidimensional CNN models. The generalization of the proposed model for the two datasets proves convincing in improving the classification accuracy. The multidimensional CNN architectures achieve an average accuracy of 93.2%, 95.8%, and 98.0%, respectively, while the proposed model achieves an average accuracy of 98.9%.

Current clinical practice in imaging medical physics is concerned with quality assurance, image processing and analysis, radiation dosimetry, risk assessment and radiation protection, and in-house training and research. Physicist workloads are projected to increase as medical imaging technologies become more sophisticated. Artificial intelligence (AI) is a rising technology with potential to assist medical physicists in their work. Exploration of AI integration into imaging medical physicist workloads is limited. In this review paper, we provide an overview of AI techniques, outline their potential usage in imaging medical physics, and discuss the limitations and challenges to clinical adoption of AI technologies.

Hydrocephalus, particularly congenital hydrocephalus in infants, remains underexplored in deep learning research. While deep learning has been widely applied to medical image analysis, few studies have specifically addressed the automated classification of hydrocephalus. This study proposes a convolutional neural network (CNN) model based on the VGG16 architecture to detect hydrocephalus in infant head CT images. The model integrates an automated method for ventricular volume extraction, applying windowing, histogram equalization, and thresholding techniques to segment the ventricles from surrounding brain structures. Morphological operations refine the segmentation and contours are extracted for visualization and volume measurement. The dataset consists of 105 head CT scans, each with 60 slices covering the ventricular volume, resulting in 6300 slices. Manual segmentation by three trained radiologists served as the reference standard. The automated method showed a high correlation with manual measurements, with R2 values ranging from 0.94 to 0.99. The mean absolute percentage error (MAPE) ranged 3.99 to 11.13%, while the root mean square error (RRMSE) from 4.56 to 13.74%. To improve model robustness, the dataset was preprocessed, normalized, and augmented with rotation, shifting, zooming, and flipping. The VGG16-based CNN used pre-trained convolutional layers with additional fully connected layers for classification, predicting hydrocephalus or normal labels. Performance evaluation using a multi-split strategy (15 independent splits) achieved a mean accuracy of 90.4% ± 1.2%. This study presents an automated approach for ventricular volume extraction and hydrocephalus detection, offering a promising tool for clinical and research applications with high accuracy and reduced observer bias.

Traditional surgical education is based on observation and assistance in surgical practice. Recently introduced deep learning (DL) techniques enable the recognition of the surgical view and automatic identification of surgical landmarks. However, there was no previous studies have conducted to develop surgical guide for robotic breast surgery. To develop a DL model for guiding the dissection plane during robotic mastectomy for beginners and trainees.

Ten surgical videos of robotic mastectomy procedures were recorded. Video frames taken at 1-s intervals were converted to PNG format. The ground truth was manually delineated by two experienced surgeons using ImageJ software. The evaluation metrics were the Dice similarity coefficient (DSC) and Hausdorff distance (HD).

A total of 8,834 images were extracted from ten surgical videos of robotic mastectomies performed between 2016 and 2020. Skin flap dissection during the robotic mastectomy console time was recorded. The median age and body mass index of the patients was 47.5 (38-52) years and 22.00 (19.30-29.52) kg/m2, respectively, and the median console time was 32 (21-48) min. Among the 8,834 images, 428 were selected and divided into training, validation, and testing datasets at a ratio of 7:1:2. Two experts determined that the DSC of our model was 0.828[Formula: see text]5.28 and 0.818[Formula: see text]6.96, while the HDs were 9.80[Formula: see text]2.57 and 10.32[Formula: see text]1.09.

DL can serve as a surgical guide for beginners and trainees, and can be used as a training tool to enhance surgeons' surgical skills.

Accurate segmentation is essential for creating digital twins based on volumetric images for high fidelity composite material analysis. Conventional techniques typically require labor-intensive and time-consuming manual effort, restricting their practical use. This paper presents a deep learning model, MBL-TransUNet, to address challenges in accurate tow-tow boundary identification via a Boundary-guided Learning module. Fabrics exhibit periodic characteristics; therefore, a Multi-scale Feature Fusion module was integrated to capture both local details and global patterns, thereby enhancing feature fusion and facilitating the effective integration of information across multiple scales. Furthermore, BatchFormerV2 was used to improve generalization through cross-batch learning. Experimental results show that MBL-TransUNet outperforms TransUNet. MIoU improved by 2.38%. In the zero-shot experiment, MIoU increased by 4.23%. The model demonstrates higher accuracy and robustness compared to existing methods. Ablation studies confirm that integrating these modules achieves optimal segmentation performance.

This study addresses the challenges of confounding effects and interpretability in artificial-intelligence-based medical image analysis. Whereas existing literature often resolves confounding by removing confounder-related information from latent representations, this strategy risks affecting image reconstruction quality in generative models, thus limiting their applicability in feature visualization. To tackle this, we propose a different strategy that retains confounder-related information in latent representations while finding an alternative confounder-free representation of the image data. Our approach views the latent space of an autoencoder as a vector space, where imaging-related variables, such as the learning target (t) and confounder (c), have a vector capturing their variability. The confounding problem is addressed by searching a confounder-free vector which is orthogonal to the confounder-related vector but maximally collinear to the target-related vector. To achieve this, we introduce a novel correlation-based loss that not only performs vector searching in the latent space, but also encourages the encoder to generate latent representations linearly correlated with the variables. Subsequently, we interpret the confounder-free representation by sampling and reconstructing images along the confounder-free vector. The efficacy and flexibility of our proposed method are demonstrated across three applications, accommodating multiple confounders and utilizing diverse image modalities. Results affirm the method's effectiveness in reducing confounder influences, preventing wrong or misleading associations, and offering a unique visual interpretation for in-depth investigations by clinical and epidemiological researchers. The code is released in the following GitLab repository: https://gitlab.com/radiology/compopbio/ai_based_association_analysis.

Due to the inherent limitations of imaging sensors, acquiring medical images that simultaneously provide functional metabolic information and detailed structural organization remains a significant challenge. Multi-modal image fusion has emerged as a critical technology for clinical diagnosis and surgical navigation, as it enables the integration of complementary information from different imaging modalities. However, existing deep learning (DL)-based fusion methods often face difficulties in effectively combining high-frequency detail information with low-frequency contextual information, which frequently leads to the degradation of high-frequency details. Therefore, there is a pressing need for a method that addresses these challenges, preserving both high- and low-frequency information while maintaining clear structural contours. In response to this issue, a novel convolutional neural network (CNN), named the multi-scale pyramid residual weight network (LYWNet), is proposed. The objective of this approach is to improve the fusion process by effectively integrating high- and low-frequency information, thereby enhancing the quality and accuracy of multimodal image fusion. This method aims to overcome the limitations of current fusion techniques and ensure the preservation of both functional and structural details, ultimately contributing to more precise clinical diagnoses and better surgical navigation outcomes.

We propose a novel CNN, LYWNet, designed to address these challenges. LYWNet is composed of three modules: (I) data preprocessing module: utilizes three convolutional layers to extract both deep and shallow features from the input images. (II) Feature extraction module: incorporates three identical multi-scale pyramid residual weight (LYW) blocks in series, each featuring three interactive branches to preserve high-frequency detail information effectively. (III) Image reconstruction module: utilizes a fusion algorithm based on feature distillation to ensure the effective integration of functional and anatomical information. The proposed image fusion algorithm enhances the interaction of contextual cues and retains the metabolic details from functional images while preserving texture details from anatomical images.

The proposed LYWNet demonstrated its ability to retain high-frequency details during feature extraction, effectively combining them with low-frequency contextual information. The fusion results exhibited reduced differences between the fused image and the original images. The structural similarity (SSIM) and peak signal-to-noise ratio (PSNR) were 0.5592±0.0536 and 17.3594±1.0211, respectively, for single-photon emission computed tomography-magnetic resonance imaging (SPECT-MRI), 0.5195±0.0730 and 14.5324±1.7365 for PET-MRI; 0.5376±0.0442 and 13.9202±0.7265 for magnetic resonance imaging-computed tomography.

LYWNet excels at integrating high-frequency detail information and low-frequency contextual information, addressing the deficiencies of existing DL-based image fusion methods. This approach provides superior fused images that retain the functional metabolic information and anatomical texture, making it a valuable tool for clinical diagnosis and surgical navigation.

Current breast cancer diagnosis methods often face limitations such as high cost, time consumption, and inter-observer variability. To address these challenges, this research proposes a novel deep learning framework that leverages generative adversarial networks (GANs) for data augmentation and transfer learning to enhance breast cancer classification using convolutional neural networks (CNNs). The framework uses a two-stage augmentation approach. First, a conditional Wasserstein GAN (cWGAN) generates synthetic breast cancer images based on clinical data, enhancing training stability and enabling targeted feature incorporation. Second, traditional augmentation techniques (e.g., rotation, flipping, cropping) are applied to both original and synthetic images. A multi-scale transfer learning technique is also employed, integrating three pre-trained CNNs (DenseNet-201, NasNetMobile, ResNet-101) with a multi-scale feature enrichment scheme, allowing the model to capture features at various scales. The framework was evaluated on the BreakHis dataset, achieving an accuracy of 99.2% for binary classification and 98.5% for multi-class classification, significantly outperforming existing methods. This framework offers a more efficient, cost-effective, and accurate approach for breast cancer diagnosis. Future work will focus on generalizing the framework to clinical datasets and integrating it into diagnostic workflows.

Artificial intelligence (AI) demonstrates high potential when applied to radiomic analysis of magnetic resonance imaging (MRI) to discriminate sinonasal tumors. This can enhance diagnostic suspicion beyond visual assessment alone and prior to biopsy, leading to expedite the diagnostic timeline and the treatment planning. The aim of the present work is to evaluate the current advancements and accuracy of this technology in this domain.

A systematic literature review was conducted following PRISMA guidelines. Inclusion criteria comprised studies utilizing any machine learning approach applied to MRI of patients with sinonasal tumors. For each study, comprehensive data were gathered on the MRI protocols, feature extraction techniques, and classifiers employed to develop the AI model. The performance was assessed based on accuracy and area under the curve (AUC).

Fourteen studies, published between May 2017 and August 2024, were included. These studies were categorized into three groups: those examining both benign and malignant tumors, those investigating malignant tumor subpopulations, and those focusing on benign pathologies. All studies reported an AUC greater than 0.800, achieving AUC > 0.89 and accuracy > 0.81 when incorporating clinical-radiological variables. Notably, the best discrimination performance was observed in studies utilizing combined conventional MRI sequences, including T1-weighted, contrasted T1-weighted, and T2-weighted images.

The application of AI and radiomics in analyzing MRI scans presents significant promise for improving the discrimination of sinonasal tumors. Integrating clinical and radiological indicators enhances model performance, suggesting that future research should focus on larger patient cohorts and diverse AI methodologies to refine diagnostic accuracy and clinical utility.

Cancer is a major global health challenge, accounting for nearly one in six deaths worldwide. Early diagnosis significantly improves survival rates and patient outcomes, yet in resource-limited settings, the scarcity of medical resources often leads to late-stage diagnosis. Integrating artificial intelligence (AI) with smartphone-based imaging systems offers a promising solution by providing portable, cost-effective, and widely accessible tools for early cancer detection. This paper introduces advanced smartphone-based imaging systems that utilize various imaging modalities for in vivo detection of different cancer types and highlights the advancements of AI for in vivo cancer detection in smartphone-based imaging. However, these compact smartphone systems face challenges like low imaging quality and restricted computing power. The use of advanced AI algorithms to address the optical and computational limitations of smartphone-based imaging systems provides promising solutions. AI-based cancer detection also faces challenges. Transparency and reliability are critical factors in gaining the trust and acceptance of AI algorithms for clinical application, explainable and uncertainty-aware AI breaks the black box and will shape the future AI development in early cancer detection. The challenges and solutions for improving AI accuracy, transparency, and reliability are general issues in AI applications, the AI technologies, limitations, and potentials discussed in this paper are applicable to a wide range of biomedical imaging diagnostics beyond smartphones or cancer-specific applications. Smartphone-based multimodal imaging systems and deep learning algorithms for multimodal data analysis are also growing trends, as this approach can provide comprehensive information about the tissue being examined. Future opportunities and perspectives of AI-integrated smartphone imaging systems will be to make cutting-edge diagnostic tools more affordable and accessible, ultimately enabling early cancer detection for a broader population.

Choroid plexus (CP) is an important brain structure that produces cerebrospinal fluid (CSF). CP perfusion has been studied using multi-delay arterial spin labeling (MD-ASL) in adults but not in pediatric populations due to the challenge of small CP size in children. Here we present a high resolution (iso2 mm) MDASL protocol with 10-minute scan time and performed test-retest scans on 21 typically developing children aged 8 to 17 years. We further proposed a Transformer-based deep learning (DL) model with k-space weighted image average (KWIA) denoised images as reference for training the model. The performance of the model was evaluated by the SNR, bias and repeatability of the fitted perfusion parameters of the CP and gray matter. The proposed method was compared to several benchmark methods including KWIA, joint denoising and reconstruction with total generalized variation (TGV) regularization, as well as another self-supervised method termed Noise2Void. The results show that the proposed Transformer model with KWIA reference can effectively denoise multi-delay ASL images, not only improving the SNR for perfusion images of each delay, but also improving the SNR for the fitted perfusion maps for visualizing and quantifying CP perfusion in children. This may facilitate the use of MDASL in neurodevelopmental studies to characterize the development of CP and glymphatic system.

Detecting and measuring changes in longitudinal fundus imaging is key to monitoring disease progression in chronic ophthalmic diseases, such as glaucoma and macular degeneration. Clinicians assess changes in disease status by either independently reviewing or manually juxtaposing longitudinally acquired color fundus photos (CFPs). Distinguishing variations in image acquisition due to camera orientation, zoom, and exposure from true disease-related changes can be challenging. This makes manual image evaluation variable and subjective, potentially impacting clinical decision-making. We introduce our deep learning (DL) pipeline, "EyeLiner," for registering, or aligning, 2-dimensional CFPs. Improved alignment of longitudinal image pairs may compensate for differences that are due to camera orientation while preserving pathological changes.

EyeLiner registers a "moving" image to a "fixed" image using a DL-based keypoint matching algorithm.

We evaluate EyeLiner on 3 longitudinal data sets: Fundus Image REgistration (FIRE), sequential images for glaucoma forecast (SIGF), and our internal glaucoma data set from the Colorado Ophthalmology Research Information System (CORIS).

Anatomical keypoints along the retinal blood vessels were detected from the moving and fixed images using a convolutional neural network and subsequently matched using a transformer-based algorithm. Finally, transformation parameters were learned using the corresponding keypoints.

We computed the mean distance (MD) between manually annotated keypoints from the fixed and the registered moving image. For comparison to existing state-of-the-art retinal registration approaches, we used the mean area under the curve (AUC) metric introduced in the FIRE data set study.

EyeLiner effectively aligns longitudinal image pairs from FIRE, SIGF, and CORIS, as qualitatively evaluated through registration checkerboards and flicker animations. Quantitative results show that the MD decreased for this model after alignment from 321.32 to 3.74 pixels for FIRE, 9.86 to 2.03 pixels for CORIS, and 25.23 to 5.94 pixels for SIGF. We also obtained an AUC of 0.85, 0.94, and 0.84 on FIRE, CORIS, and SIGF, respectively, beating the current state-of-the-art SuperRetina (AUCFIRE = 0.76, AUCCORIS = 0.83, AUCSIGF = 0.74).

Our pipeline demonstrates improved alignment of image pairs in comparison to the current state-of-the-art methods on 3 separate data sets. We envision that this method will enable clinicians to align image pairs and better visualize changes in disease over time.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Understanding the BRAF alterations preoperatively could remarkably assist in predicting tumor behavior, which leads to a more precise prognostication and management strategy. Recent advances in artificial intelligence (AI) have resulted in effective predictive models. Therefore, for the first time, this study aimed to review the performance of machine learning and deep learning models in predicting the BRAF alterations in low-grade gliomas (LGGs)using imaging data.

PubMed/MEDLINE, Embase, and the Cochrane Library were systematically searched for studies published up to June 1, 2024, and evaluated the performance of AI models in predicting BRAF alterations in LGGs using imaging data. Pooled sensitivity, specificity, and area under the receiver operating curve (AUROC) were meta-analyzed.

A total of 6 studies with 951 patients were included in this systematic review. The pooled AUROC of internal validation cohorts for their best-performing model was 84.44%, with models detecting BRAF mutation, BRAF fusion, BRAF fusion from mutation, and BRAF wild type producing similar AUROCs of 90.75%, 84.59%, 82.33%, and 82%, respectively. The best-performing models had pooled sensitivities of 80.3%, 87.51%, and 74.14% and pooled specificities of 88.57%, 70.41%, and 83.98% for detection of BRAF fusion from mutation, BRAF fusion, and BRAF mutation, respectively.

AI models may perform relatively well in predicting BRAF alterations in LGG using imaging data and appear to be capable of high sensitivities and specificities. However, future studies with larger sample sizes implementing different machine learning or deep learning algorithms are required to reduce imprecision.

Recent advancements in cardiac imaging have been significantly enhanced by integrating deep learning models, offering transformative potential in early diagnosis and patient care. The research paper explores the application of hybrid deep learning methodologies, focusing on the roles of Autoencoders, Convolutional Neural Networks (CNNs), and Recurrent Neural Networks (RNNs) in enhancing cardiac image analysis. The study implements a comprehensive approach, combining traditional algorithms such as Sobel, Watershed, and Otsu's Thresholding with advanced deep learning models to achieve precise and accurate imaging outcomes. The Autoencoder model, developed for image enhancement and feature extraction, achieved a notable accuracy of 99.66% on the test data. Optimized for image recognition tasks, the CNN model demonstrated a high precision rate of 98.9%. The RNN model, utilized for sequential data analysis, showed a prediction accuracy of 98%, further underscoring the robustness of the hybrid framework. The research drew upon a diverse range of academic databases and pertinent publications within cardiac imaging and deep learning, focusing on peer-reviewed articles and studies published in the past five years. Models were implemented using the TensorFlow and Keras frameworks. The proposed methodology was evaluated in the clinical validation phase using advanced imaging protocols, including the QuickScan technique and balanced steady-state free precession (bSSFP) imaging. The validation metrics were promising: the Signal-to-Noise Ratio (SNR) was improved by 15%, the Contrast-to-Noise Ratio (CNR) saw an enhancement of 12%, and the ejection fraction (EF) analysis provided a 95% correlation with manually segmented data. These metrics confirm the efficacy of the models, showing significant improvements in image quality and diagnostic accuracy. The integration of adversarial defense strategies, such as adversarial training and model ensembling, have been analyzed to enhance model robustness against malicious inputs. The reliability and comparison of the model's ability have been investigated to maintain clinical integrity, even in adversarial attacks that could otherwise compromise segmentation outcomes. These findings indicate that integrating Autoencoders, CNNs, and RNNs within a hybrid deep-learning framework is promising for enhancing cardiac MRI segmentation and early diagnosis. The study contributes to the field by demonstrating the applicability of these advanced techniques in clinical settings, paving the way for improved patient outcomes through more accurate and timely diagnoses.

Machine learning models for the diagnosis of breast cancer can facilitate the prediction of cancer risk and subsequent patient management among other clinical tasks. For the models to impact clinical practice, they ought to follow standard workflows, help interpret mammography and ultrasound data, evaluate clinical contextual information, handle incomplete data and be validated in prospective settings. Here we report the development and testing of a multimodal model leveraging mammography and ultrasound modules for the stratification of breast cancer risk based on clinical metadata, mammography and trimodal ultrasound (19,360 images of 5,216 breasts) from 5,025 patients with surgically confirmed pathology across medical centres and scanner manufacturers. Compared with the performance of experienced radiologists, the model performed similarly at classifying tumours as benign or malignant and was superior at pathology-level differential diagnosis. With a prospectively collected dataset of 191 breasts from 187 patients, the overall accuracies of the multimodal model and of preliminary pathologist-level assessments of biopsied breast specimens were similar (90.1% vs 92.7%, respectively). Multimodal models may assist diagnosis in oncology.

Load-bearing structural degradation is crucial in knee osteoarthritis (KOA) progression, yet limited prediction models use load-bearing tissue radiomics for radiographic (structural) KOA incident.

We aim to develop and test a Load-Bearing Tissue plus Clinical variable Radiomic Model (LBTC-RM) to predict radiographic KOA incidents.

Risk prediction study.

The 700 knees without radiographic KOA at baseline were included from Osteoarthritis Initiative cohort. We selected 2164 knee MRIs during 4-year follow-up. LBTC-RM, which integrated MRI features of meniscus, femur, tibia, femorotibial cartilage, and clinical variables, was developed in total development cohort (n = 1082, 542 cases vs. 540 controls) using neural network algorithm. Final predictive model was tested in total test cohort (n = 1082, 534 cases vs. 548 controls), which integrated data from five visits: baseline (n = 353, 191 cases vs. 162 controls), 3 years prior KOA (n = 46, 19 cases vs. 27 controls), 2 years prior KOA (n = 143, 77 cases vs. 66 controls), 1 year prior KOA (n = 220, 105 cases vs. 115 controls), and at KOA incident (n = 320, 156 cases vs. 164 controls).

In total test cohort, LBTC-RM predicted KOA incident with AUC (95 % CI) of 0.85 (0.82-0.87); with LBTC-RM aid, performance of resident physicians for KOA prediction were improved, with specificity, sensitivity, and accuracy increasing from 50 %, 60 %, and 55 %-72 %, 73 %, and 72 %, respectively. The LBTC-RM output indicated an increased KOA risk (OR: 20.6, 95 % CI: 13.8-30.6, p < .001). Radiomic scores of load-bearing tissue raised KOA risk (ORs: 1.02-1.9) from 4-year prior KOA whereas 3-dimensional feature score of medial meniscus decreased the OR (0.99) of KOA incident at KOA confirmed. The 2-dimensional feature score of medial meniscus increased the ORs (1.1-1.2) of KOA symptom score from 2-year prior KOA.

We provided radiomic features of load-bearing tissue to improved KOA risk level assessment and incident prediction. The model has potential clinical applicability in predicting KOA incidents early, enabling physicians to identify high-risk patients before significant radiographic evidence appears. This can facilitate timely interventions and personalized management strategies, improving patient outcomes.

This study presents a novel approach integrating longitudinal MRI-based radiomics and clinical variables to predict knee osteoarthritis (KOA) incidence using machine learning. By leveraging deep learning for auto-segmentation and machine learning for predictive modeling, this research provides a more interpretable and clinically applicable method for early KOA detection. The introduction of a Radiomics Score System enhances the potential for radiomics as a virtual image-based biopsy tool, facilitating non-invasive, personalized risk assessment for KOA patients. The findings support the translation of advanced imaging and AI-driven predictive models into clinical practice, aiding early diagnosis, personalized treatment planning, and risk stratification for KOA progression. This model has the potential to be integrated into routine musculoskeletal imaging workflows, optimizing early intervention strategies and resource allocation for high-risk populations. Future validation across diverse cohorts will further enhance its clinical utility and generalizability.

Measurement noise often leads to inaccurate shear wave phase velocity estimation in ultrasound shear wave elastography. Filtering techniques are commonly used for denoising the shear wavefields. However, these filters are often not sufficient, especially in fatty tissues where the signal-to-noise ratio (SNR) can be very low.

The purpose of this study is to develop a deep learning approach for denoising shear wavefields in ultrasound shear wave elastography. This may lead to improved reconstruction of shear wave phase velocity image maps.

The study addresses noise by transforming particle velocity data into a time-frequency representation. A neural network with encoder and decoder convolutional blocks effectively decomposes the input and extracts the signal of interest, improving the SNR in high-noise scenarios. The network is trained on simulated phantoms with elasticity values ranging from 3  to 60 kPa. A total of 1 85 570 samples with 80%-20 % $\%$ split were used for training and validation. The approach is tested on experimental phantom and ex-vivo goat liver tissue data. Performance was compared with the traditional filtering methods such as bandpass, median, and wavelet filtering. Kruskal-Wallis one-way analysis of variance was performed to check statistical significance. Multiple comparisons were performed using the Mann-Whitney U test and Holm-Bonferroni adjustment ofp - values $p-{\rm values}$ .

The results are evaluated using SNR and the percentage of pixels that can be reconstructed in the phase velocity maps. The SNR levels in experimental data improved from -2 to 9.9 dB levels to 15.6 to 30.3 dB levels. Kruskal-Wallis one-way analysis showed statistical significance (p < 0.05 $p<0.05$ ). Multiple comparisons with p-value corrections also showed statistically significant improvement when compared to the bandpass and wavelet filtering scheme (p < 0.05 $p<0.05$ ). Smoother phase velocity maps were reconstructed after denoising. The coefficient of variation is less than5 % $5\%$ in CIRS phantom and less than18 % $18\%$ in ex-vivo goat liver tissue.

The proposed approach demonstrates improvement in shear wave phase velocity image map reconstruction and holds promise that deep learning methods can be effectively utilized to extract true shear wave signal from measured noisy data.

Meticulous clinical examination is essential for spinal disorders to utilize the diagnostic methods and technologies that strongly support physicians and enhance clinical practice. A significant change in the approach to diagnosing spinal disorders has occurred in the last three decades, which has enhanced a more nuanced understanding of spine pathology. Traditional radiographic methods such as conventional and functional X-rays and CT scans are still the first line in the diagnosis of spinal disorders due to their low cost and accessibility. As more advanced imaging technologies become increasingly available worldwide, there is a constantly increasing trend in MRI scans for detecting spinal pathologies and making treatment decisions. Not only do MRI scans have superior diagnostic capabilities, but they also assist surgeons in performing meticulous preoperative planning, making them currently the most widely used diagnostic tool for spinal disorders. Positron Emission Tomography (PET) can help detect inflammatory lesions, infections, and tumors. Other advanced diagnostic tools such as CT/MRI fusion image, Functional Magnetic Resonance Imaging (fMRI), Upright and Kinetic MRI, magnetic resonance spectroscopy (MRS), diffusion-weighted imaging (DWI), and diffusion tensor imaging (DTI) could play an important role when it comes to detecting more special pathologies. However, some technical difficulties in the daily praxis and their high costs act as obstacles to their further spread. Integrating artificial intelligence and advancements in data analytics and virtual reality promises to enhance spinal procedures' precision, safety, and efficacy. As these technologies continue to develop, they will play a critical role in transforming spinal surgery. This paradigm shift emphasizes the importance of continuous innovation and adaptability in improving the diagnosis and treatment of spinal disorders.

This paper proposes a Tangent Rat Swarm Optimization (TRSO)-DenseNet for the detection of skin cancer to reduce the severity rate of cancer. Initially, the input image is pre-processed by employing a linear smoothing filter. The pre-processed image is transferred to skin lesion segmentation, where Mask-RCNN is utilized for segmenting the skin lesion. Then, image augmentation is performed using techniques such as vertical shifting, horizontal shifting, random rotation, brightness adjustment, blurring, and cropping. The augmented image is then fed into the feature extraction phase to identify statistical features, Haralick texture features, Convolutional Neural Network (CNN) features, Local Ternary Pattern (LTP), Histogram of Oriented Gradients (HOG), and Local Vector Pattern (LVP). Finally, the extracted features are fed into the skin cancer detection phase, where DenseNet is used to detect skin cancer. Here, DenseNet is structurally optimized by TRSO, which has the combination of the Tangent Search Algorithm (TSA) and Rat Swarm Optimizer (RSO). The TRSO-DenseNet model is implemented using MATLAB tool and analayzsed using the Society for Imaging Informatics in Medicine-International Skin Imaging Collaboration's (SIIM-ISIC) Melanoma Classification dataset. The Proposed model for skin cancer detection attained superior performance with an accuracy of 94.63%, TPR of 91.51%, and TNR of 92.46%.

Axillary lymph node status, which was routinely assessed by axillary lymph node dissection (ALND) until the 1990s, is a crucial factor in determining the stage, prognosis, and therapeutic strategy used for breast cancer patients. Axillary surgery for breast cancer patients has evolved from ALND to minimally invasive approaches. Over the decades, the application of noninvasive imaging techniques, machine learning approaches and emerging clinical prediction models for the detection of axillary lymph node metastasis greatly improves clinical diagnostic efficacy and provides optimal surgical selection. In this work, we summarize the historical axillary surgery and updated perspectives of axillary management for breast cancer patients.

Many artificial intelligence (AI) machine learning (ML) papers focused on clinical outcomes use registry data inadequate for predictive modeling. In contrast, diagnostic imaging is an area where available information (pixels, etc.) can result in a reliable, clinically relevant, and accurate model. The use of deep learning for image analysis can reduce interobserver variability, and highlight subtle and meaningful features. AI augments, rather than replaces, clinical expertise, allowing faster, more consistent, and potentially more accurate diagnostic information. This is especially relevant when imaging data is abundant, as continuous model training can further refine diagnostic precision. An effective 3-step approach includes: 1) an efficient "detector" to determine where to look; 2) computational ability to focus on key features of the image and "blur out" background noise ("attention module"); and 3) interpreted key features ("explainability"). Next, the larger process of developing and employing a predictive model needs to be externally validated, to determine the extent to which these results will generalize outside of a single institution. Outside this setting, i.e., external validity, needs to be determined.

Osteoporosis is a disease characterized by low bone mineral density and an increased risk of fractures. In dentistry, mandibular bone morphology, assessed for example on panoramic images, has been employed to detect osteoporosis. Artificial intelligence (AI) can aid in diagnosing bone diseases from radiographs. We aimed to systematically review, synthesize and appraise the available evidence supporting AI in detecting osteoporosis on panoramic radiographs.

Studies that used AI to detect osteoporosis on dental panoramic images were included.

On April 8, 2023, a first comprehensive search of electronic databases was conducted, including PubMed, Scopus, Embase, IEEE, arXiv, and Google Scholar (grey literature). This search was subsequently updated on October 6, 2024.

The Quality Assessment and Diagnostic Accuracy Tool-2 was employed to determine the risk of bias in the studies. Quantitative analyses involved meta-analyses of diagnostic accuracy measures, including sensitivity and specificity, yielding Diagnostic Odds Ratios (DOR) and synthesized positive likelihood ratios (LR+). The certainty of evidence was assessed using the Grading of Recommendations Assessment, Development, and Evaluation system.

A total of 24 studies were included. Accuracy ranged from 50% to 99%, sensitivity from 50% to 100%, and specificity from 38% to 100%. A minority of studies (n=10) had a low risk of bias in all domains, while the majority (n=18) showed low risk of applicability concerns. Pooled sensitivity was 87.92% and specificity 81.93%. DOR was 32.99, and L+ 4.87. Meta-regression analysis indicated that sample size had only a marginal impact on heterogeneity (R² = 0.078, p = 0.052), suggesting other study-level factors may contribute to variability. Egger's test suggested potential small-study effects (p < 0.001), indicating a risk of publication bias.

AI, particularly deep learning, showed high diagnostic accuracy in detecting osteoporosis on panoramic radiographs. The results indicate a strong potential for AI to enhance osteoporosis screening in dental settings. However, significant heterogeneity across studies and potential small-study effects highlight the need for further validation, standardization, and larger, well-powered studies to improve model generalizability.

The application of AI in analyzing panoramic radiographs could transform osteoporosis screening in routine dental practice by providing early and accurate diagnosis. This has the potential to integrate osteoporosis detection seamlessly into dental workflows, improving patient outcomes and enabling timely referrals for medical intervention. Addressing issues of model validation and comparability is critical to translating these findings into widespread clinical use.

Magneto-Acousto-Electrical Tomography (MAET) is a hybrid imaging method that combines advantages of ultrasound imaging and electrical impedance tomography to image the electrical conductivity of biological tissues. In practical applications, different tissue or disease organization display various conductivity traits. However, the conductivity map consists of overlapping signals measured at multiple locations, the reconstruction results are affected by noise, which results in blurred reconstruction boundaries, low contrast, and irregular artifact distributions. To improve the image resolution and reduce noise of MAET, a dataset of conductivity maps reconstructed from MAET was established, dubbed MAET-IMAGE. Based on this dataset, we proposed a MAET tomography segmentation network based on the Segment Anything Model (SAM), termed as MAET-SAM. Specifically, we froze the encoder weights of SAM to extract rich feature information of image and design, an adaptive decoder with no prompts. In the end, an end-to-end segmentation model for specific MAET images with MAET-IMAGE was proposed. Qualitative and quantitative experiments demonstrated that MAET-SAM outperformed traditional segmentation methods and segmentation models with initial weights in terms of MAET image segmentation performance, bringing new breakthroughs and advancements to the field of medical imaging analysis and clinical diagnosis.

This meta-analysis aims to evaluate the diagnostic performance of CT-based artificial intelligence (AI) in diagnosing cervical lymph node metastasis (LNM) of papillary thyroid cancer (PTC).

A systematic search was conducted in PubMed, Embase, and Web of Science databases through December 2024, following PRISMA-DTA guidelines. Studies evaluating CT-based AI models for diagnosing cervical LNM in patients with pathologically confirmed PTC were included. The methodological quality was assessed using a modified QUADAS-2 tool. A bivariate random-effects model was used to calculate pooled sensitivity, specificity, and area under the curve (AUC). Heterogeneity was evaluated using I2 statistics, and meta-regression analyses were performed to explore potential sources of heterogeneity.

17 studies comprising 1778 patients in internal validation sets and 4072 patients in external validation sets were included. In internal validation sets, AI demonstrated a sensitivity of 0.80 (95% CI: 0.71-0.86), specificity of 0.79 (95% CI: 0.73-0.84), and AUC of 0.86 (95% CI: 0.83-0.89). Radiologists suggested comparable performance with sensitivity of 0.77 (95% CI: 0.64-0.87), specificity of 0.79 (95% CI: 0.72-0.85), and AUC of 0.85 (95% CI: 0.81-0.88). Subgroup analyses revealed that deep learning methods outperformed machine learning in sensitivity (0.86 vs 0.72, P<0.05). No significant publication bias was found in internal validation sets for AI diagnosis (P=0.78).

CT-based AI showed comparable diagnostic performance to radiologists for detecting cervical LNM in PTC patients, with deep learning models showing superior sensitivity. AI could potentially serve as a valuable diagnostic support tool, though further prospective validation is warranted. Limitations include high heterogeneity among studies and insufficient external validation in diverse populations.

Monitoring anemia recovery is crucial for clinical intervention. Morphological assessment of red blood cells (RBCs) with peripheral blood smears (PBSs) provides additional information beyond routine blood tests. However, the PBS test is labor-intensive, reliant on manual analysis, and susceptible to variability in expert interpretations. Here we introduce a deep semi-supervised learning method, RBCMatch, to classify RBCs during anemia recovery. Using an acute hemolytic anemic mouse model, PBS images at four different time points during anemia recovery were acquired and segmented into 10,091 single RBC images, with only 5% annotated and used in model training. By employing the semi-supervised strategy Fixmatch, RBCMatch achieved an impressive average classification accuracy of 91.2% on the validation dataset and 87.5% on a held-out dataset, demonstrating its superior accuracy and robustness compared to supervised learning methods, especially when labeled samples are scarce. To characterize the anemia recovery process, principal components (PCs) of RBC embeddings were extracted and visualized. Our results indicated that RBC embeddings quantified the state of anemia recovery, and the second PC had a strong correlation with RBC count and hemoglobin concentration, demonstrating the model's ability to accurately depict RBC morphological changes during anemia recovery. Thus, this study provides a valuable tool for the automatic classification of RBCs and offers novel insights into the assessment of anemia recovery, with the potential to aid in clinical decision-making and prognosis analysis in the future.

This review explores the significant role of artificial intelligence (AI) in managing metabolic disorders like diabetes, obesity, metabolic dysfunction-associated fatty liver disease (MAFLD), and thyroid dysfunction. AI applications in this context encompass early diagnosis, personalized treatment plans, risk assessment, prevention, and biomarker discovery for early and accurate disease management. This review also delves into techniques involving machine learning (ML), deep learning (DL), natural language processing (NLP), computer vision, and reinforcement learning associated with AI and their application in metabolic disorders. The following study also enlightens the challenges and ethical considerations associated with AI implementation, such as data privacy, model interpretability, and bias mitigation. We have reviewed various AI-based tools utilized for the diagnosis and management of metabolic disorders, such as Idx, Guardian Connect system, and DreaMed for diabetes. Further, the paper emphasizes the potential of AI to revolutionize the management of metabolic disorders through collaborations among clinicians and AI experts, the integration of AI into clinical practice, and the necessity for long-term validation studies. The references provided in the paper cover a range of studies related to AI, ML, personalized medicine, metabolic disorders, and diagnostic tools in healthcare, including research on disease diagnostics, personalized therapy, chronic disease management, and the application of AI in diabetes care and nutrition.

The interpretation of plain hip radiographs can vary widely among physicians. This study aimed to develop and validate a deep learning-based screening model for distinguishing normal hips from severe hip diseases on plain radiographs.

Electronic medical records and plain radiograph from 2004 to 2012 were used to construct two patient groups: the hip disease group (those who underwent total hip arthroplasty) and normal group. A total of 1,726 radiographs (500 normal hip radiographs and 1,226 radiographs with hip diseases, respectively) were included and were allocated for training (320 and 783), validation (80 and 196), and test (100 and 247) groups. Four different models were designed-raw image for both training and test set, preprocessed image for training but raw image for the test set, preprocessed images for both sets, and change of backbone algorithm from DenseNet to EfficientNet. The deep learning models were compared in terms of accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), F1-score, and area under the receiver operating characteristic curve (AUROC).

The mean age of the patients was 54.0 ± 14.8 years in the hip disease group and 49.8 ± 14.9 years in the normal group. The final model showed highest performance in both the internal test set (accuracy 0.96, sensitivity 0.96, specificity 0.97, PPV 0.99, NPV 0.99, F1-score 0.97, and AUROC 0.99) and the external validation set (accuracy 0.94, sensitivity 0.93, specificity 0.96, PPV 0.95, NPV 0.93, F1-score 0.94, and AUROC 0.98). In the gradcam image, while the first model depended on unrelated marks of radiograph, the second and third model mainly focused on the femur shaft and sciatic notch, respectively.

The deep learning-based model showed high accuracy and reliability in screening hip diseases on plain radiographs, potentially aiding physicians in more accurately diagnosing hip conditions.

Pulmonary granulomatous nodules (PGN) often exhibit similar CT morphological features to solid lung adenocarcinomas (SLA), making preoperative differentiation challenging. This study aims to address this diagnostic challenge by developing a novel deep learning model.

This study proposes MAEMC-NET, a model integrating generative (Masked AutoEncoder) and contrastive (Momentum Contrast) self-supervised learning to learn CT image representations of intra- and inter-solitary nodules. A generative self-supervised task of reconstructing masked axial CT patches containing lesions was designed to learn intra- and inter-slice image representations. Contrastive momentum is used to link the encoder in axial-CT-patch path with the momentum encoder in coronal-CT-patch path. A total of 494 patients from two centers were included.

MAEMC-NET achieved an area under curve (95% Confidence Interval) of 0.962 (0.934-0.973). These results not only significantly surpass the joint diagnosis by two experienced chest radiologists (77.3% accuracy) but also outperform the current state-of-the-art methods. The model performs best on medical images with a 50% mask ratio, showing a 1.4% increase in accuracy compared to the optimal 75% mask ratio on natural images.

The proposed MAEMC-NET effectively distinguishes between benign and malignant solitary pulmonary nodules and holds significant potential to assist radiologists in improving the diagnostic accuracy of PGN and SLA.

Accurate and precise identification of cholelithiasis is essential for saving the lives of millions of people worldwide. Although several computer-aided cholelithiasis diagnosis approaches have been introduced in the literature, their use is limited because Convolutional Neural Network (CNN) models are black box in nature. Therefore, a novel approach for cholelithiasis classification using custom CNN with post-hoc model explanation is proposed. This paper presents multiple contributions. First, a custom CNN architecture is proposed to classify and predict cholelithiasis from ultrasound image. Second, a modified deep convolutional generative adversarial network is proposed to produce synthetic ultrasound images for better model generalization. Third, a hybrid visual explanation method is proposed by combining gradient-weighted class activation with local interpretable model agnostic explanation to generate a visual explanation using a heatmap. Fourth, an exhaustive performance analysis of the proposed approach on ultrasound images collected from three different Indian hospitals is presented to showcase its efficacy for computer-aided cholelithiasis diagnosis. Fifth, a team of radiologists evaluates and validates the prediction and respective visual explanations made using the proposed approach. The results reveal that the proposed cholelithiasis classification approach beats the performance of state-of-the-art pre-trained CNN and Vision Transformer models. The heatmap generated through the proposed hybrid explanation method offers detailed visual explanations to enhance transparency and trustworthiness in the medical domain.

Photoacoustic tomography, a novel non-invasive imaging modality, combines the principles of optical and acoustic imaging for use in biomedical applications. In scenarios where photoacoustic signal acquisition is insufficient due to sparse-view sampling, conventional direct reconstruction methods significantly degrade image resolution and generate numerous artifacts. To mitigate these constraints, a novel sinogram-domain priors guided extremely sparse-view reconstruction method for photoacoustic tomography boosted by enhanced diffusion model is proposed. The model learns prior information from the data distribution of sinograms under full-ring, 512-projections. In iterative reconstruction, the prior information serves as a constraint in least-squares optimization, facilitating convergence towards more plausible solutions. The performance of the method is evaluated using blood vessel simulation, phantoms, and in vivo experimental data. Subsequently, the transformation of the reconstructed sinograms into the image domain is achieved through the delay-and-sum method, enabling a thorough assessment of the proposed method. The results show that the proposed method demonstrates superior performance compared to the U-Net method, yielding images of markedly higher quality. Notably, for in vivo data under 32 projections, the sinogram structural similarity improved by ∼21 % over U-Net, and the image structural similarity increased by ∼51 % and ∼84 % compared to U-Net and delay-and-sum methods, respectively. The reconstruction in the sinogram domain for photoacoustic tomography enhances sparse-view imaging capabilities, potentially expanding the applications of photoacoustic tomography.

Using Deep Learning in computer-aided diagnosis systems has been of great interest due to its impressive performance in the general domain and medical domain. However, a notable challenge is the lack of explainability of many advanced models, which poses risks in critical applications such as diagnosing findings in CXR. To address this problem, we propose ItpCtrl-AI, a novel end-to-end interpretable and controllable framework that mirrors the decision-making process of the radiologist. By emulating the eye gaze patterns of radiologists, our framework initially determines the focal areas and assesses the significance of each pixel within those regions. As a result, the model generates an attention heatmap representing radiologists' attention, which is then used to extract attended visual information to diagnose the findings. By allowing the directional input, our framework is controllable by the user. Furthermore, by displaying the eye gaze heatmap which guides the diagnostic conclusion, the underlying rationale behind the model's decision is revealed, thereby making it interpretable. In addition to developing an interpretable and controllable framework, our work includes the creation of a dataset, named Diagnosed-Gaze++, which aligns medical findings with eye gaze data. Our extensive experimentation validates the effectiveness of our approach in generating accurate attention heatmaps and diagnoses. The experimental results show that our model not only accurately identifies medical findings but also precisely produces the eye gaze attention of radiologists. The dataset, models, and source code will be made publicly available upon acceptance.

Laryngoscopy, essential for diagnosing laryngeal cancer (LCA), faces challenges due to high inter-observer variability and the reliance on endoscopist expertise. Distinguishing precancerous from early-stage cancerous lesions is particularly challenging, even for experienced practitioners, given their similar appearances. This study aims to enhance laryngoscopic image analysis to improve early screening/detection of cancer or precancerous conditions.

We propose MedFormer, a laryngeal cancer classification method based on the Vision Transformer (ViT). To address data scarcity, MedFormer employs a customized transfer learning approach that leverages the representational power of pre-trained transformers. This method enables robust out-of-domain generalization by fine-tuning a minimal set of additional parameters.

MedFormer exhibits sensitivity-specificity values of 98%-89% for identifying precancerous lesions (leukoplakia) and 89%-97% for detecting cancer, surpassing CNN counterparts significantly. Additionally, when compared to the two selected ViT-based models, MedFormer also demonstrates superior performance. It also outperforms physician visual evaluations (PVE) in certain scenarios and matches PVE performance in all cases. Visualizations using class activation maps (CAM) and deformable patches demonstrate MedFormer's interpretability, aiding clinicians in understanding the model's predictions.

We highlight the potential of visual transformers in clinical laryngoscopic assessments, presenting MedFormer as an effective method for the early detection of laryngeal cancer.

Prostate cancer (PCa) poses a significant threat to men's health, with early diagnosis being crucial for improving prognosis and reducing mortality rates. Transrectal ultrasound (TRUS) plays a vital role in the diagnosis and image-guided intervention of PCa. To facilitate physicians with more accurate and efficient computer-assisted diagnosis and interventions, many image processing algorithms in TRUS have been proposed and achieved state-of-the-art performance in several tasks, including prostate gland segmentation, prostate image registration, PCa classification and detection and interventional needle detection. The rapid development of these algorithms over the past 2 decades necessitates a comprehensive summary. As a consequence, this survey provides a narrative review of this field, outlining the evolution of image processing methods in the context of TRUS image analysis and meanwhile highlighting their relevant contributions. Furthermore, this survey discusses current challenges and suggests future research directions to possibly advance this field further.

The neglect of occult lymph nodes metastasis (OLNM) is one of the pivotal causes of early non-small cell lung cancer (NSCLC) recurrence after local treatments such as stereotactic body radiotherapy (SBRT) or surgery. This study aimed to develop and validate a computed tomography (CT)-based radiomics and deep learning (DL) fusion model for predicting non-invasive OLNM.

Patients with radiologically node-negative lung adenocarcinoma from two centers were retrospectively analyzed. We developed clinical, radiomics, and radiomics-clinical models using logistic regression. A DL model was established using a three-dimensional squeeze-and-excitation residual network-34 (3D SE-ResNet34) and a fusion model was created by integrating seleted clinical, radiomics features and DL features. Model performance was assessed using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, calibration curves, and decision curve analysis (DCA). Five predictive models were compared; SHapley Additive exPlanations (SHAP) and Gradient-weighted Class Activation Mapping (Grad-CAM) were employed for visualization and interpretation.

Overall, 358 patients were included: 186 in the training cohort, 48 in the internal validation cohort, and 124 in the external testing cohort. The DL fusion model incorporating 3D SE-Resnet34 achieved the highest AUC of 0.947 in the training dataset, with strong performance in internal and external cohorts (AUCs of 0.903 and 0.907, respectively), outperforming single-modal DL models, clinical models, radiomics models, and radiomics-clinical combined models (DeLong test: P<0.05). DCA confirmed its clinical utility, and calibration curves demonstrated excellent agreement between predicted and observed OLNM probabilities. Features interpretation highlighted the importance of textural characteristics and the surrounding tumor regions in stratifying OLNM risk.

The DL fusion model reliably and accurately predicts OLNM in early-stage lung adenocarcinoma, offering a non-invasive tool to refine staging and guide personalized treatment decisions. These results may aid clinicians in optimizing surgical and radiotherapy strategies.

Medical imaging is pivotal in early disease diagnosis, providing essential insights that enable timely and accurate detection of health anomalies. Traditional imaging techniques, such as Magnetic Resonance Imaging (MRI), Computer Tomography (CT), ultrasound, and Positron Emission Tomography (PET), offer vital insights into three-dimensional structures but frequently fall short of delivering a comprehensive and detailed anatomical analysis, capturing only amplitude details. Three-dimensional holography microscopic medical imaging provides a promising solution by capturing the amplitude (brightness) and phase (structural information) details of biological structures. In this study, we investigate the novel collaborative potential of Deep Learning (DL) and holography microscopic phase imaging for cancer diagnosis. The study comprehensively examines existing literature, analyzes advancements, identifies research gaps, and proposes future research directions in cancer diagnosis through the integrated Quantitative Phase Imaging (QPI) and DL methodology. This novel approach addresses a critical limitation of traditional imaging by capturing detailed structural information, paving the way for more accurate diagnostics. The proposed approach comprises tissue sample collection, holographic image scanning, preprocessing in case of imbalanced datasets, and training on annotated datasets using DL architectures like U-Net and Vision Transformer(ViT's). Furthermore, sophisticated concepts in DL, like the incorporation of Explainable AI (XAI) techniques, are suggested for comprehensive disease diagnosis and identification. The study thoroughly investigates the advantages of integrating holography imaging and DL for precise cancer diagnosis. Additionally, meticulous insights are presented by identifying the challenges associated with this integration methodology.

Cloud-based, surgical augmented intelligence (Cydar Medical, Cambridge, United Kingdom) can be used for surgical planning and intraoperative imaging guidance during complex endovascular aortic procedures. We aim to evaluate radiation exposure, operative safety metrics, and postoperative renal outcomes following implementation of Cydar imaging guidance using a manually matched cohort of aortic procedures.

We retrospectively reviewed our prospectively maintained database of endovascular aortic cases. Patients repaired using Cydar imaging were matched to patients who underwent a similar procedure without using Cydar. Matching was performed manually on a 1:1 basis using anatomy, device configuration, number of branches/fenestrations, and adjunctive procedures including in-situ laser fenestration. Radiation, contrast use, and other operative metrics were compared. Preoperative and postoperative maximum creatinine was compared to assess for acute kidney injury (AKI) based on risk, injury, failure, loss of kidney function, and end-stage kidney disease (RIFLE) criteria.

Hundred patients from 2012 to 2023 were identified: 50 cases (38 fenestrated endovascular aortic repairs, 2 thoracic endovascular aortic repairs, 3 octopus-type thoracoabdominal aortic aneurysm repair, 7 endovascular aneurysm repairs) where Cydar imaging was used, with suitable matches to 50 non-Cydar cases. Baseline characteristics including body mass index did not differ significantly between the 2 groups (27.8 ± 5.6 vs. 26.7 ± 6.1; P = 0.31). Radiation dose was significantly lower in the Cydar group (2529 ± 2256 vs. 3676 ± 2976 mGy; P < 0.03), despite there being no difference in fluoroscopy time (51 ± 29.4 vs. 58 ± 37.2 min; P = 0.37). Contrast volume (94 ± 37.4 vs. 93 ± 43.9 mL; P = 0.73), estimated blood loss (169 ± 223 vs. 193 ± 222 mL; P = 0.97), and procedure time (154 ± 78 vs. 165 ± 89.1 min) did not differ significantly. Additionally, Cydar versus non-Cydar patients did not show a significant difference between precreatinine and postcreatinine changes (0.13 ± 0.08 vs. 0.05 ± 0.07; P = 0.34). Only one patient in the non-Cydar group met RIFLE criteria for AKI postoperatively.

The use of cloud-based augmented intelligence imaging was associated with a significant reduction in radiation dose in a cohort of matched aortic procedures but did not appear to affect other parameters or renal function. Even with advanced imaging, surgeons should remain conscientious about radiation safety and administration of nephrotoxic contrast agents.