Search Results (819)

Background: Cerebral amyloid angiopathy (CAA) is a common cause of spontaneous intracerebral hemorrhage (ICH) in elderly individuals, and it is characterized by the deposition of amyloid β protein (Aß) in the walls of small-caliber cortical and leptomeningeal vessels. The diagnostic criteria for CAA highlight its association with spontaneous lobar hemorrhage, convexity subarachnoid hemorrhage (SAH), and cortical superficial siderosis but not with subdural hematoma (SDH). This article presents a three-patient case series of CAA who experienced a lobar ICH associated with an SDH, underscoring a potentially under-recognized correlation between an acute ICH and coexistent SDH. Case presentation: We present a case series of three patients in a single university medical center who experienced acute-onset lobar ICH with a concurrent SDH, treated with evacuation. Histopathological examination established the diagnosis of CAA in all three cases. This case series underscores a potentially under-recognized association between an acute ICH and coexistent SDH in the context of CAA. Conclusions: Considering our findings, we emphasize the possibility that SDH may be a more frequent manifestation of CAA than previously recognized. Therefore, patients with CAA who initially present with acute SDH may be underdiagnosed, consequently leading to delayed identification and missed opportunities for proper risk assessment and management. Full article

Given the overlap of epidemiological and clinical presentations of both rickettsioses and malaria infections, diagnostic testing where malaria is confirmed or excluded, without subsequent rickettsial testing, specifically in the case of Plasmodium vivax or P. ovale infection, may mask the possibility of relapse. A lack of clinical suspicion of co-infections, absence of knowledge on the geographic distribution of diseases, and lack of availability of point-of-care diagnostic testing for other tropical diseases can often lead to missed diagnosis or misdiagnosis of common tropical infections, including rickettsioses. We herein describe a case of confirmed intercurrent rickettsial and P. vivax infection, with the former potentially triggering a relapse of the latter in a febrile traveler returning to Canada from South America, and review the literature on tropical coinfections in returning travelers. Full article

Background: Pancreatic cystic lesions (PCLs), including intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs), pose a diagnostic challenge due to their variable malignant potential. Current guidelines, such as Fukuoka and American Gastroenterological Association (AGA), have moderate predictive accuracy and may lead to overtreatment or missed malignancies. Artificial intelligence (AI), incorporating machine learning (ML) and deep learning (DL), offers the potential to improve risk stratification, diagnosis, and management of PCLs by integrating clinical, radiological, and molecular data. This is the first systematic review to evaluate the application, performance, and clinical utility of AI models in the diagnosis, classification, prognosis, and management of pancreatic cysts. Methods: A systematic review was conducted in accordance with PRISMA guidelines and registered on PROSPERO (CRD420251008593). Databases searched included PubMed, EMBASE, Scopus, and Cochrane Library up to March 2025. The inclusion criteria encompassed original studies employing AI, ML, or DL in human subjects with pancreatic cysts, evaluating diagnostic, classification, or prognostic outcomes. Data were extracted on the study design, imaging modality, model type, sample size, performance metrics (accuracy, sensitivity, specificity, and area under the curve (AUC)), and validation methods. Study quality and bias were assessed using the PROBAST and adherence to TRIPOD reporting guidelines. Results: From 847 records, 31 studies met the inclusion criteria. Most were retrospective observational (n = 27, 87%) and focused on preoperative diagnostic applications (n = 30, 97%), with only one addressing prognosis. Imaging modalities included Computed Tomography (CT) (48%), endoscopic ultrasound (EUS) (26%), and Magnetic Resonance Imaging (MRI) (9.7%). Neural networks, particularly convolutional neural networks (CNNs), were the most common AI models (n = 16), followed by logistic regression (n = 4) and support vector machines (n = 3). The median reported AUC across studies was 0.912, with 55% of models achieving AUC ≥ 0.80. The models outperformed clinicians or existing guidelines in 11 studies. IPMN stratification and subtype classification were common focuses, with CNN-based EUS models achieving accuracies of up to 99.6%. Only 10 studies (32%) performed external validation. The risk of bias was high in 93.5% of studies, and TRIPOD adherence averaged 48%. Conclusions: AI demonstrates strong potential in improving the diagnosis and risk stratification of pancreatic cysts, with several models outperforming current clinical guidelines and human readers. However, widespread clinical adoption is hindered by high risk of bias, lack of external validation, and limited interpretability of complex models. Future work should prioritise multicentre prospective studies, standardised model reporting, and development of interpretable, externally validated tools to support clinical integration. Full article

Introduction: Although lipomas are common benign tumors found in adults, lipomas of the bladder are extremely rare. Bladder lipomas are infrequently reported in the urologic literature, with only 19 cases published worldwide. These can present as a mass on cystoscopy and cause irritative voiding symptoms, depending on their location. Upon transurethral resection, seeing fat can be concerning for a perforation, as lipoma can be mistaken for extravesical fat. Hence, familiarity with this rare entity is of paramount importance for urologists to prevent unnecessary investigations and interventions that are needed in case of a true bladder perforation. Case presentation: This study presents a case of bladder lipoma in a 73-year-old male with end-stage renal disease who presented for pretransplant urologic evaluation due to microscopic hematuria and irritative lower urinary tract symptoms (LUTS). During cystoscopy, a bladder mass was seen, and a transurethral resection of the bladder tumor (TURBT) revealed bright yellow adipose tissue immediately underneath the bladder mucosa. Concerns about perforation were obviated when seeing intact detrusor muscle underneath, visually confirming the integrity of the bladder wall. The resection was completed, and the CT scan was re-read with the radiologist, which confirmed the presence of a lipoma that was missed pre-operatively due to patient’s oliguria and collapsed bladder. No catheter drainage or cystogram was performed based on these findings. Outcome: The patient healed without any complications. Histopathology confirmed the diagnosis of a mature lipoma. The patient was cleared for transplant from a urologic standpoint and had a successful renal transplantation without delay. Discussion: This case documents the anomalous occurrence of a lipoma within the bladder and supports maintaining a broad differential, including liposarcoma, angiomyolipoma, and other non-malignant fatty tumors during the evaluation of a bladder mass. Full article

Cancer metabolic reprogramming plays a critical role in tumor progression and therapeutic resistance, underscoring the need for advanced analytical strategies. Metabolomics, leveraging mass spectrometry and nuclear magnetic resonance (NMR) spectroscopy, offers a comprehensive and functional readout of tumor biochemistry. By enabling both targeted metabolite quantification and untargeted profiling, metabolomics captures the dynamic metabolic alterations associated with cancer. The integration of metabolomics with machine learning (ML) approaches further enhances the interpretation of these complex, high-dimensional datasets, providing powerful insights into cancer biology from biomarker discovery to therapeutic targeting. This review systematically examines the transformative role of ML in cancer metabolomics. We discuss how various ML methodologies—including supervised algorithms (e.g., Support Vector Machine, Random Forest), unsupervised techniques (e.g., Principal Component Analysis, t-SNE), and deep learning frameworks—are advancing cancer research. Specifically, we highlight three major applications of ML–metabolomics integration: (1) cancer subtyping, exemplified by the use of Similarity Network Fusion (SNF) and LASSO regression to classify triple-negative breast cancer into subtypes with distinct survival outcomes; (2) biomarker discovery, where Random Forest and Partial Least Squares Discriminant Analysis (PLS-DA) models have achieved >90% accuracy in detecting breast and colorectal cancers through biofluid metabolomics; and (3) prognostic modeling, demonstrated by the identification of race-specific metabolic signatures in breast cancer and the prediction of clinical outcomes in lung and ovarian cancers. Beyond these areas, we explore applications across prostate, thyroid, and pancreatic cancers, where ML-driven metabolomics is contributing to earlier detection, improved risk stratification, and personalized treatment planning. We also address critical challenges, including issues of data quality (e.g., batch effects, missing values), model interpretability, and barriers to clinical translation. Emerging solutions, such as explainable artificial intelligence (XAI) approaches and standardized multi-omics integration pipelines, are discussed as pathways to overcome these hurdles. By synthesizing recent advances, this review illustrates how ML-enhanced metabolomics bridges the gap between fundamental cancer metabolism research and clinical application, offering new avenues for precision oncology through improved diagnosis, prognosis, and tailored therapeutic strategies. Full article

Background: Lysosomal storage diseases (LSDs) are a genetically and clinically heterogeneous group of inborn errors of metabolism caused by variants in genes encoding lysosomal hydrolases, membrane proteins, activator proteins, or transporters. These disease-causing variants lead to enzymatic deficiencies and the progressive accumulation of undegraded substrates within lysosomes, disrupting cellular function across multiple organ systems. While classical phenotypes typically manifest in infancy or early childhood with severe multisystem involvement, a combination of advances in molecular diagnostics [particularly next-generation sequencing (NGS)] and improved understanding of disease heterogeneity have enabled the identification of attenuated forms characterised by residual enzyme activity and later-onset presentations. These milder phenotypes often evade early recognition due to nonspecific or isolated symptoms, resulting in significant diagnostic delays and missed therapeutic opportunities. Objectives/Methods: This study characterises the clinical, biochemical, and molecular profiles of 10 adult patients diagnosed with LSDs, all representing attenuated forms, and discusses them alongside a narrative review. Results: Enzyme activity, molecular data, and phenotypic assessments are described to explore genotype–phenotype correlations and identify diagnostic challenges. Conclusions: These findings highlight the variable expressivity and organ involvement of attenuated LSDs and reinforce the importance of maintaining clinical suspicion in adults presenting with unexplained cardiovascular, neurological, ophthalmological, or musculoskeletal findings. Enhanced recognition of atypical presentations is critical to facilitate earlier diagnosis, guide management, and enable cascade testing for at-risk family members. Full article

Ischemic stroke is a complex, multifactorial disorder with a significant heritable component. Recent developments in genome-wide association studies (GWASs) have identified several common variants associated with clinical outcomes, stroke subtypes, and overall risk. Key loci implicated in biological pathways related to vascular integrity, lipid metabolism, inflammation, and atherogenesis include 9p21 (ANRIL), HDAC9, SORT1, and PITX2. Although polygenic risk scores (PRSs) hold promise for early risk prediction and stratification, their clinical utility remains limited by Eurocentric bias and missing heritability. Integrating multiomics approaches, such as functional genomics, transcriptomics, and epigenomics, enhances our understanding of stroke pathophysiology and paves the way for precision medicine. This review summarizes the current genetic landscape of ischemic stroke, emphasizing how evolving methodologies are shaping its prevention, diagnosis, and treatment. Full article

Tuberculosis (TB) is a global public health issue requiring early and accurate diagnosis. The loop-mediated isothermal amplification (LAMP) assay is a promising alternative recommended by the WHO for the initial diagnosis of pulmonary TB, particularly in resource-limited settings. This study evaluated the sensitivity and specificity of a commercial LAMP assay for TB detection using 198 samples from different countries including Mexico. The LAMP assay results were compared to the results of standard tests: AFB smear microscopy, cell culture, and Xpert PCR. Across all samples, LAMP showed a sensitivity of 96.20% and a specificity of 84.61%. When compared specifically to “true positives” and “true negatives” (defined by the consistency across the standard tests), LAMP demonstrated 100% sensitivity and 92.30% specificity. For context, the sensitivity of AFB smear microscopy against the culture and Xpert tests was 79.04%. A significant finding was that the LAMP test detected a high percentage (92.5%) of samples found positive by the culture and Xpert tests but negative by the AFB smear, highlighting its ability to identify cases missed by traditional microscopy. This study concluded that the LAMP assay is a sensitive and specific tool for TB diagnosis with potential for rapid and accurate diagnosis, especially in resource-limited areas. Full article

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that impairs motor function, including the fine motor control required for handwriting. Traditional diagnostic methods often lack sensitivity and objectivity in the early stages, limiting opportunities for timely intervention. There is a growing need for non-invasive, accessible tools capable of capturing subtle motor changes that precede overt clinical symptoms. Among early PD manifestations, handwriting impairments such as micrographia have shown potential as digital biomarkers. However, conventional handwriting analysis remains subjective and limited in scope. Recent advances in artificial intelligence (AI) and machine learning (ML) enable automated analysis of handwriting dynamics, such as pressure, velocity, and fluency, collected via digital tablets and smartpens. These tools support the detection of early-stage PD, monitoring of disease progression, and assessment of therapeutic response. This paper highlights how AI-enhanced handwriting analysis provides a scalable, non-invasive method to support diagnosis, enable remote symptom tracking, and personalize treatment strategies in PD. This approach integrates clinical neurology with computer science and rehabilitation, offering practical applications in telemedicine, digital health, and personalized medicine. By capturing dynamic features often missed by traditional assessments, AI-based handwriting analysis contributes to a paradigm shift in the early detection and long-term management of PD, with broad relevance across neurology, digital diagnostics, and public health innovation. Full article

Skin cancer is one of the most prevalent forms of cancer worldwide, and early and accurate diagnosis critically impacts patient outcomes. Given the sensitive nature of medical data and its fragmented distribution across institutions (data silos), privacy-preserving collaborative learning is essential to enable knowledge-sharing without compromising patient confidentiality. While federated learning (FL) offers a promising solution, existing methods struggle with heterogeneous and missing modalities across institutions, which reduce the diagnostic accuracy. To address these challenges, we propose an effective and flexible Personalized Multimodal Federated Learning framework (PMM-FL), which enables efficient cross-client knowledge transfer while maintaining personalized performance under heterogeneous and incomplete modality conditions. Our study contains three key contributions: (1) A hierarchical aggregation strategy that decouples multi-module aggregation from local deployment via global modular-separated aggregation and local client fine-tuning. Unlike conventional FL (which synchronizes all parameters in each round), our method adopts a frequency-adaptive synchronization mechanism, updating parameters based on their stability and functional roles. (2) A multimodal fusion approach based on multitask learning, integrating learnable modality imputation and attention-based feature fusion to handle missing modalities. (3) A custom dataset combining multi-year International Skin Imaging Collaboration(ISIC) challenge data (2018–2024) to ensure comprehensive coverage of diverse skin cancer types. We evaluate PMM-FL through diverse experiment settings, demonstrating its effectiveness in heterogeneous and incomplete modality federated learning settings, achieving 92.32% diagnostic accuracy with only a 2% drop in accuracy under 30% modality missingness, with a 32.9% communication overhead decline compared with baseline FL methods. Full article

Background/Objectives: Splenic torsion is a well-known and reported clinical problem. Splenic torsions after abdominal trauma represent a small group of cases that involve surgical management. They manifest primarily as abdominal pain, and the diagnosis is made based on imaging studies—ultrasound, CT, and MRI. Methods: This work aimed to analyze traumatic splenic torsions in terms of their clinical course, symptoms, timing, involvement of imaging techniques in the diagnosis, histopathological examination, and overall outcome. We searched databases using the desk research method under the keywords “splenic torsion”, “torsion”, and “spleen”, as well as in combination with “traumatic”, finding a total of eight cases, which we included in our analysis. Results: The eight cases were analyzed, comprising four females and four males, with an average age of 16.25 years (range 5–29 years). Traffic accidents were the most frequent cause of injury (five cases), while the circumstances were unclear in the remaining three. Immediate abdominal symptoms appeared in six patients. Splenic torsion was preoperatively diagnosed in five out of seven confirmed cases. A total of seven patients underwent laparotomy with splenectomy. In one case, laparoscopy converted to laparotomy with splenopexy preserved the spleen. Histopathology, performed in only two cases, confirmed splenic infarction in one patient; infarction status could not be determined in the remaining five due to missing data. Conclusions: Post-traumatic splenic torsions are a group of atypical injuries as the primary and immediate consequence of the trauma suffered is not anatomical–structural damage to the organ, such as a rupture. Mostly affecting young people, the cases described in the professional literature involve the main spleen, which was considered to be “wandering”, suggesting that this is a key predisposing factor for splenic torsion following blunt trauma and requiring diagnostic imaging for diagnosis. Full article

Background/Objectives: Early diagnosis of stroke is crucial for effective treatment with tissue plasminogen activator (tPA) and endovascular thrombectomy. Emergency medical services (EMSs) screening and the early activation of emergency department (ED) stroke protocols reduce treatment times and improve patient outcomes. This study aims to validate ED stroke protocol activation by EMSs in a large stroke center. Methods: This retrospective cross-sectional study was conducted at Magen David Adom and Shamir Medical Center between 1 January 2019 and 31 December 2019. Data were categorized into patients suspected by EMSs of having a stroke and those not suspected by EMSs but diagnosed as having a stroke in the ED. The primary outcome was the accuracy of EMSs in activating ED stroke protocols. Results: In this study, there were 23,061 patients, of which 11,841 (51.9%) were females. The mean age was 61.4 (SD = 22.72) years old. EMSs suspected 743 (3.22%) patients were having a stroke. In 587 (79%), EMSs activated ED stroke protocols. There were 88 cases where strokes were diagnosed in the ED when EMSs did not suspect a stroke. The overall EMSs negative predictive value (NPV) was 100% while the positive predictive value (PPV) was 20%. Conclusions: While Israeli EMSs over-activate the ED stroke protocol, stroke patients are almost never missed, achieving the goal of prehospital stroke screening. To prevent resource waste, all involved teams should be notified, and the actual activation of the stroke protocol should be carried out by an ED physician upon patient arrival. Communication between all levels regarding stroke protocol should also be increased to decrease the time to treatment. Full article

Structural magnetic resonance imaging (sMRI) is a vital tool for diagnosing neurological brain diseases. However, sMRI scans often show significant structural changes only in limited brain regions due to localised atrophy, making the identification of discriminative features a key challenge. Importantly, the human brain exhibits inherent bilateral symmetry, and deviations from this symmetry—such as asymmetric atrophy—are strong indicators of early Alzheimer’s disease (AD). Patch-based methods help capture local brain changes for early AD diagnosis, but they often struggle with fixed-size limitations, potentially missing subtle asymmetries or broader contextual cues. To address these limitations, we propose a novel augmented reality (AR)-enhanced patch-level explainable deep learning (ARE-PaLED) system. It includes an adaptive multi-scale patch extraction network (AMPEN) to adjust patch sizes based on anatomical characteristics and spatial context, as well as an informative patch selection algorithm (IPSA) to identify discriminative patches, including those reflecting asymmetry patterns associated with AD; additionally, an AR module is proposed for future immersive explainability, complementing the patch-level interpretation framework. Evaluated on 1862 subjects from the ADNI and AIBL datasets, the framework achieved an accuracy of 92.5% (AD vs. NC) and 85.9% (AD vs. MCI). The proposed ARE-PaLED demonstrates potential as an interpretable and immersive diagnostic aid for sMRI-based AD diagnosis, supporting the interpretation of model predictions for AD diagnosis. Full article

Background: Artificial intelligence and machine learning are playing crucial roles in digital transformation, aiming to improve the efficiency, effectiveness, equity, and responsiveness of primary health systems and their services. Method: Using synthetic knowledge synthesis and bibliometric and thematic analysis triangulation, we identified the most productive and prolific countries, institutions, funding sponsors, source titles, publications productivity trends, and principal research categories and themes. Results: The United States and the United Kingdom were the most productive countries; Plos One and BJM Open were the most prolific journals; and the National Institutes of Health, USA, and the National Natural Science Foundation of China were the most productive funding sponsors. The publication productivity trend is positive and exponential. The main themes are related to natural language processing in clinical decision-making, primary health care optimization focusing on early diagnosis and screening, improving health-based social determinants, and using chatbots to optimize communications with patients and between health professionals. Conclusions: The use of machine learning in primary health care aims to address the significant global burden of so-called “missed diagnostic opportunities” while minimizing possible adverse effects on patients. Full article

Background/Objectives: In dental medicine, the integration of various types of X-ray images, such as periapical (PA), bitewing (BW), and panoramic (PANO) radiographs, is crucial for comprehensive oral health assessment. These complementary imaging modalities provide diverse diagnostic perspectives and support the early detection of oral diseases, thereby enhancing treatment outcomes. However, there is currently no existing system that integrates multiple types of dental X-rays for both adults and children to perform tooth localization and numbering. Methods: Therefore, this study aimed to propose a system based on YOLOv8 that integrates multiple dental X-ray images and automatically detects and numbers both permanent and deciduous teeth. Through image preprocessing, various types of dental X-ray images were standardized and enhanced to improve the recognition accuracy of individual teeth. Results: With the implementation of a novel image preprocessing method, the system achieved a detection precision of 98.16% for permanent and deciduous teeth, representing a 3% improvement over models without image enhancement. In addition, the system attained an average tooth numbering accuracy of 98.5% for permanent teeth and 96.3% for deciduous teeth, surpassing existing methods by 5.6%. Conclusions: These results might highlight the innovation of the proposed image processing method and show its practical value in assisting clinicians with accurate diagnosis of tooth loss and the identification of missing teeth, ultimately contributing to improved diagnosis and treatment in dental care. Full article