Search Results (203)

Background: Breast cancer remains a major public health challenge in Latin America, where access to personalized risk assessment tools is still limited. This study aimed to evaluate the implementation of a polygenic risk score (PRS)-based stratification model combined with remote genomic counseling in Colombian women with sporadic breast cancer and healthy women. Methods: In 2023, an embedded mixed-methods observational study was conducted in Medellín involving 1997 women aged 40–75 years who underwent clinical PRS testing. The intervention integrated PRS-based risk categorization with individualized risk factor assessment and lifestyle recommendations delivered through a remote counseling platform. Results: PRS analysis classified 9.7% of women as high risk and 46% as low risk. Healthier lifestyle patterns were significantly associated with lower PRS categories (p = 0.034). Physical activity showed a protective effect (OR = 0.60, 95% CI: 0.5–0.8), while prior smoking, elevated BMI, and sedentary behavior were associated with higher risk. The counseling model achieved high delivery (93%) and satisfaction (85%) rates. Qualitative insights revealed improved understanding of genomic risk and greater engagement in preventive behaviors. Only one new case of breast cancer was detected among intermediate-risk participants, with a diagnostic lead time of 12 months. Conclusions: These findings support the feasibility, acceptability, and potential impact of integrating PRS and genomic counseling in cancer prevention strategies in middle-income settings. Full article

Accurate diagnosis of gastrointestinal (GI) diseases typically requires invasive procedures or imaging studies that pose the risk of various post-procedural complications or involve radiation exposure. Bowel sounds (BSs), though typically described during a GI-focused physical exam, are highly inaccurate and variable, with low clinical value in diagnosis. Interpretation of the acoustic characteristics of BSs, i.e., using a phonoenterogram (PEG), may aid in diagnosing various GI conditions non-invasively. Use of artificial intelligence (AI) and improvements in computational analysis can enhance the use of PEGs in different GI diseases and lead to a non-invasive, cost-effective diagnostic modality that has not been explored before. The purpose of this work was to develop an automated AI model, You Only Listen Once (YOLO), to detect prominent bowel sounds that can enable real-time analysis for future GI disease detection and diagnosis. A total of 110 2-minute PEGs sampled at 44.1 kHz were recorded using the Eko DUO^® stethoscope from eight healthy volunteers at two locations, namely, left upper quadrant (LUQ) and right lower quadrant (RLQ) after IRB approval. The datasets were annotated by trained physicians, categorizing BSs as prominent or obscure using version 1.7 of Label Studio Software^®. Each BS recording was split up into 375 ms segments with 200 ms overlap for real-time BS detection. Each segment was binned based on whether it contained a prominent BS, resulting in a dataset of 36,149 non-prominent segments and 6435 prominent segments. Our dataset was divided into training, validation, and test sets (60/20/20% split). A 1D-CNN augmented transformer was trained to classify these segments via the input of Mel-frequency cepstral coefficients. The developed AI model achieved area under the receiver operating curve (ROC) of 0.92, accuracy of 86.6%, precision of 86.85%, and recall of 86.08%. This shows that the 1D-CNN augmented transformer with Mel-frequency cepstral coefficients achieved creditable performance metrics, signifying the YOLO model’s capability to classify prominent bowel sounds that can be further analyzed for various GI diseases. This proof-of-concept study in healthy volunteers demonstrates that automated BS detection can pave the way for developing more intuitive and efficient AI-PEG devices that can be trained and utilized to diagnose various GI conditions. To ensure the robustness and generalizability of these findings, further investigations encompassing a broader cohort, inclusive of both healthy and disease states are needed. Full article

Introduction: Breast density is a well-recognized factor in breast cancer risk assessment, with higher density linked to increased malignancy risk and reduced sensitivity of conventional mammography. Background parenchymal enhancement (BPE), observed in contrast-enhanced imaging, reflects physiological contrast uptake in non-pathologic breast tissue. While extensively characterized in breast MRI, the role of BPE in contrast-enhanced mammography (CEM) remains uncertain due to inconsistent findings regarding its correlation with breast density and cancer risk. Unlike breast density—standardized through the ACR BI-RADS lexicon—BPE lacks a uniform classification system in CEM, leading to variability in clinical interpretation and research outcomes. To address this gap, we introduce the BPE-CEM Standard Scale (BCSS), a structured four-tiered classification system specifically tailored to the two-dimensional characteristics of CEM, aiming to improve consistency and diagnostic alignment in BPE evaluation. Materials and Methods: In this retrospective single-center study, 213 patients who underwent mammography (MG), ultrasound (US), and contrast-enhanced mammography (CEM) between May 2022 and June 2023 at the “A. Perrino” Hospital in Brindisi were included. Breast density was classified according to ACR BI-RADS (categories A–D). BPE was categorized into four levels: Minimal (< 10% enhancement), Light (10–25%), Moderate (25–50%), and Marked (> 50%). Three radiologists independently assessed BPE in a subset of 50 randomly selected cases to evaluate inter-observer agreement using Cohen’s kappa. Correlations between BPE, breast density, and age were examined through regression analysis. Results: BPE was Minimal in 57% of patients, Light in 31%, Moderate in 10%, and Marked in 2%. A significant positive association was found between higher breast density (BI-RADS C–D) and increased BPE (p < 0.05), whereas lower-density breasts (A–B) were predominantly associated with minimal or light BPE. Regression analysis confirmed a modest but statistically significant association between breast density and BPE (R² = 0.144), while age showed no significant effect. Inter-observer agreement for BPE categorization using the BCSS was excellent (κ = 0.85; 95% CI: 0.78–0.92), supporting its reproducibility. Conclusions: Our findings indicate that breast density is a key determinant of BPE in CEM. The proposed BCSS offers a reproducible, four-level framework for standardized BPE assessment tailored to the imaging characteristics of CEM. By reducing variability in interpretation, the BCSS has the potential to improve diagnostic consistency and facilitate integration of BPE into personalized breast cancer risk models. Further prospective multicenter studies are needed to validate this classification and assess its clinical impact. Full article

Background and Objectives: this pilot study aimed to evaluate the diagnostic accuracy of ChatGPT-4o in analyzing oral mucosal lesions from clinical images. Materials and Methods: a total of 110 clinical images, including 100 pathological lesions and 10 healthy mucosal images, were retrieved from Google Images and analyzed by ChatGPT-4o using a standardized prompt. An expert panel of five clinicians established a reference diagnosis, categorizing lesions as benign or malignant. The AI-generated diagnoses were classified as correct or incorrect and further categorized as plausible or not plausible. The accuracy, sensitivity, specificity, and agreement with the expert panel were analyzed. The Artificial Intelligence Performance Instrument (AIPI) was used to assess the quality of AI-generated recommendations. Results: ChatGPT-4o correctly diagnosed 85% of cases. Among the 15 incorrect diagnoses, 10 were deemed plausible by the expert panel. The AI misclassified three malignant lesions as benign but did not categorize any benign lesions as malignant. Sensitivity and specificity were 91.7% and 100%, respectively. The AIPI score averaged 17.6 ± 1.73, indicating strong diagnostic reasoning. The McNemar test showed no significant differences between AI and expert diagnoses (p = 0.084). Conclusions: In this proof-of-concept pilot study, ChatGPT-4o demonstrated high diagnostic accuracy and strong descriptive capabilities in oral mucosal lesion analysis. A residual 8.3% false-negative rate for malignant lesions underscores the need for specialist oversight; however, the model shows promise as an AI-powered triage aid in settings with limited access to specialized care. Full article

Primary and secondary immunodeficiencies comprise a wide array of illnesses marked by immune system abnormalities, resulting in heightened vulnerability to infections, autoimmunity, and cancers. Notwithstanding progress in diagnostic instruments and an enhanced comprehension of the underlying pathophysiology, delayed diagnosis and underreporting persist as considerable obstacles. The implementation of artificial intelligence into clinical practice has surfaced as a viable method to enhance early detection, risk assessment, and management of immunodeficiencies. Recent advancements illustrate how artificial intelligence-driven models, such as predictive algorithms, electronic phenotyping, and automated flow cytometry analysis, might enable early diagnosis, minimize diagnostic delays, and enhance personalized treatment methods. Furthermore, artificial intelligence-driven immunopeptidomics and phenotypic categorization are enhancing vaccine development and biomarker identification. Successful implementation necessitates overcoming problems associated with data standardization, model validation, and ethical issues. Future advancements will necessitate a multidisciplinary partnership among physicians, data scientists, and governments to effectively use the revolutionary capabilities of artificial intelligence, therefore ushering in an age of precision medicine in immunodeficiencies. Full article

Hypocalcemia is common in dairy cows within the first 72 h post-calving, and can be either clinical or subclinical. Early detection is critical, but traditional laboratory tests are time-consuming and cow-side tests remain costly. A classic symptom of hypocalcemia is reduced ear skin temperature, which has been explored as a diagnostic tool in a previous study, but was not recommended at the end. Additionally, ambient temperature was found to strongly influence ear skin temperature, complicating diagnosis. The present study investigates infrared thermography of the ear as a potential non-invasive method for helping in the detection of hypocalcemia in Holstein cows. In order to differ from the previous study, with the goal of improving diagnosis accuracy, this research analyzed the entire ear temperature using infrared imaging software. Ambient temperature was factored in by categorizing samples into two groups based on air temperature: colder (−1.6 to 14.6 °C) and hotter (15.3 to 31.2 °C). Forty-two cows were monitored during the perinatal period, with blood samples and thermographic images taken twice a day until 48 h after calving. This study found that the median surface temperature of the ear correlated strongly with environmental temperature (r = 0.806, p < 0.001) and weakly with blood ionized calcium levels (r = 0.310, p < 0.01). In colder air temperatures, ear surface temperature was significantly different between healthy and hypocalcemic cows (p = 0.014). Logistic regression models were used to assess ionized calcium status based on different combinations of ear surface temperature, its difference from air temperature, and days in milk. In hotter air temperatures, only ear surface temperature, with no other covariates, was able to generate a valid model (p = 0.029). In colder air temperatures, multiple combinations of those variables generated valid models (p < 0.05), with the difference between ear and air temperature, together with days in milk, performing the best. Thus, this study concluded that ear surface temperature obtained through infrared thermography, while not promising for warmer environments, does show application potential for helping in the detection of hypocalcemia in colder environments. Full article

Respiratory diseases present significant global health challenges. Recent advances in respiratory sound analysis (RSA) have shown great potential for automated disease diagnosis and patient management. The International Conference on Biomedical and Health Informatics 2017 (ICBHI2017) database stands as one of the most authoritative open-access RSA datasets. This review systematically examines 135 technical publications utilizing the database, and a comprehensive and timely summary of RSA methodologies is offered for researchers and practitioners in this field. Specifically, this review covers signal processing techniques including data resampling, augmentation, normalization, and filtering; feature extraction approaches spanning time-domain, frequency-domain, joint time–frequency analysis, and deep feature representation from pre-trained models; and classification methods for adventitious sound (AS) categorization and pathological state (PS) recognition. Current achievements for AS and PS classification are summarized across studies using official and custom data splits. Despite promising technique advancements, several challenges remain unresolved. These include a severe class imbalance in the dataset, limited exploration of advanced data augmentation techniques and foundation models, a lack of model interpretability, and insufficient generalization studies across clinical settings. Future directions involve multi-modal data fusion, the development of standardized processing workflows, interpretable artificial intelligence, and integration with broader clinical data sources to enhance diagnostic performance and clinical applicability. Full article

The integration of artificial intelligence into healthcare has advanced rapidly in recent years, with multimodal approaches emerging as promising tools for improving diagnostic accuracy and clinical decision making. These approaches combine heterogeneous data sources such as medical images, electronic health records, physiological signals, and clinical notes to better capture the complexity of disease processes. Despite this progress, only a limited number of studies offer a unified view of multimodal AI applications in medicine. In this review, we provide a comprehensive and up-to-date analysis of machine learning and deep learning-based multimodal architectures, fusion strategies, and their performance across a range of diagnostic tasks. We begin by summarizing publicly available datasets and examining the preprocessing pipelines required for harmonizing heterogeneous medical data. We then categorize key fusion strategies used to integrate information from multiple modalities and overview representative model architectures, from hybrid designs and transformer-based vision-language models to optimization-driven and EHR-centric frameworks. Finally, we highlight the challenges present in existing works. Our analysis shows that multimodal approaches tend to outperform unimodal systems in diagnostic performance, robustness, and generalization. This review provides a unified view of the field and opens up future research directions aimed at building clinically usable, interpretable, and scalable multimodal diagnostic systems. Full article

Background/Objectives: Knee osteoarthritis (KOA) is a prevalent disorder affecting both older adults and younger individuals, leading to compromised joint function and mobility. Early and accurate detection is critical for effective intervention, as treatment options become increasingly limited as the disease progresses. Traditional diagnostic methods rely heavily on the expertise of physicians and are susceptible to errors. The demand for utilizing deep learning models in order to automate and improve the accuracy of KOA image classification has been increasing. In this research, a unique deep learning model is presented that employs autoencoders as the primary mechanism for feature extraction, providing a robust solution for KOA classification. Methods: The proposed model differentiates between KOA-positive and KOA-negative images and categorizes the disease into its primary severity levels. Levels of severity range from “healthy knees” (0) to “severe KOA” (4). Symptoms range from typical joint structures to significant joint damage, such as bone spur growth, joint space narrowing, and bone deformation. Two experiments were conducted using different datasets to validate the efficacy of the proposed model. Results: The first experiment used the autoencoder for feature extraction and classification, which reported an accuracy of 96.68%. Another experiment using autoencoders for feature extraction and Extreme Learning Machines for actual classification resulted in an even higher accuracy value of 98.6%. To test the generalizability of the Knee-DNS system, we utilized the Butterfly iQ+ IoT device for image acquisition and Google Colab’s cloud computing services for data processing. Conclusions: This work represents a pioneering application of autoencoder-based deep learning models in the domain of KOA classification, achieving remarkable accuracy and robustness. Full article

As Artificial Intelligence systems increasingly surpass or replace traditional human roles, institutions founded on beliefs in human cognitive superiority, moral authority, and procedural oversight encounter a more profound challenge than mere disruption: expiration. This paper posits that, instead of being outperformed, many legacy institutions are becoming epistemically misaligned with the realities of AI-driven environments. To clarify this change, the paper presents the Expiration Theory. This conceptual model interprets institutional collapse not as a market failure but as the erosion of fundamental assumptions amid technological shifts. In addition, the paper introduces the AI Pressure Clock, a diagnostic tool that categorizes institutions based on their vulnerability to AI disruption and their capacity to adapt to it. Through an analysis across various sectors, including law, healthcare, education, finance, and the creative industries, the paper illustrates how specific systems are nearing functional obsolescence while others are actively restructuring their foundational norms. As a conceptual study, the paper concludes by highlighting the theoretical, policy, and leadership ramifications, asserting that institutional survival in the age of AI relies not solely on digital capabilities but also on the capacity to redefine the core principles of legitimacy, authority, and decision-making. Full article

Background: Artificial Intelligence (AI) is revolutionizing healthcare by enhancing diagnostic precision and risk assessment. In dentistry, AI has been increasingly integrated into Cone Beam Computed Tomography (CBCT) to improve image interpretation and pre-surgical planning. The lingual foramen (LF), a vital anatomical structure that transmits neurovascular elements, requires accurate evaluation during implant procedures. Traditional CBCT studies describe LF variations but lack a standardized risk classification. This study introduces a novel AI-based model for stratifying the surgical risk associated with LF using machine learning. Objectives: This study aimed to (1) assess the prevalence and anatomical variations of the lingual foramen (LF) using CBCT, (2) develop an AI-driven risk classification model based on LF characteristics, and (3) compare the AI model’s performance with that of traditional statistical methods. Materials and Methods: A retrospective analysis of 166 CBCT scans was conducted. K-means clustering and decision tree algorithms classified foramina into Low, Moderate, and High-Risk groups based on count, size, and proximity to the alveolar crest. The model performance was evaluated using confusion matrix analysis, heatmap correlations, and the elbow method. Traditional analyses (chi-square and logistic regression) were also performed. Results: The AI model categorized foramina into low (60%), moderate (30%), and high (10%) risk groups. The decision tree achieved a classification accuracy of 92.6 %, with 89.4% agreement with expert manual classification, confirming the model’s reliability. Conclusions: This study presents a validated AI-driven model for the risk assessment of the lingual foramen. Integrating AI into CBCT workflows offers a structured, objective, and automated method for enhancing surgical safety and precision in dental implant planning. Full article

Headache represents one of the most prevalent and disabling conditions in the pediatric population, with significant repercussions on mental and psychological well-being, as well as on academic achievement and social functioning, ultimately leading to a marked reduction in quality of life. Currently, the diagnosis of headache is based on the clinical criteria of the third edition of the International Classification of Headache Disorders (ICHD-3). However, the characteristics of headache may differ between adults and children, as well as the ability of children to provide a complete description of the pain and associated symptoms. The immature narrative skills of children can represent a limitation in defining the clinical phenotype of headache, making the diagnosis more complex. This is even more challenging when extracting information about the characteristics of the headache in children whose verbal expression is poorly developed or completely absent. Given these limitations, clinical psychology has long used drawing as an effective diagnostic instrument to bypass verbal communication barriers. This tool provides unique access to children’s psychological and emotional states, as a direct window into their inner world and as an expressive medium that often generates more detailed, accurate, and clinically actionable information, compared to verbal reports alone. For these reasons, drawing has been recognized as a valuable diagnostic tool for decades, with multiple studies demonstrating specificity and accuracy rates comparable to standard clinical assessments. Particularly for young children, drawings may give access to fundamental information that might otherwise remain inaccessible, thereby allowing both accurate diagnosis and individualized treatment planning. Multiple studies have highlighted and confirmed the graphic differences between representations of various types of headaches and the undeniable utility of an “artistic diagnosis” alongside the clinical one. Furthermore, the literature suggests and encourages the use of drawing in clinical practice, both in the diagnostic process and during subsequent follow-up, as an effective, enjoyable, easy-to-use, and low-cost resource. Accordingly, we propose a narrative review accompanied by a curated collection of drawings that may help identify and categorize specific correlations between graphic representations and clinical phenotypes, such as pain location, quality, intensity, association with nausea and vomiting, photophobia and phonophobia, and types of migraine aura. Our goal is to create a visual reference that can aid clinicians in the accurate interpretation of children’s drawings. Additionally, we aim to promote the integration of this method into routine clinical practice to improve diagnostic precision and support a more child-centered model of care. We also hope to propose new iconographic models to further enrich the diagnostic framework. Full article

Background and Objectives: Fibrosis stage is the key histopathological determinant of liver-related outcomes in metabolic dysfunction-associated steatohepatitis (MASH); however, a reliable noninvasive method for predicting fibrosis stage remains an unmet need. This study aimed to develop an accurate, practical, and noninvasive tool for identifying “at-risk MASH patients”. Materials and Methods: Fifty-six patients with biopsy-confirmed MASH were prospectively enrolled and categorized into fibrosis stages using the NASH-CRN system. In addition to anthropometric and biochemical parameters, seven serum fibrosis biomarkers were evaluated across fibrosis stages. Binary logistic regression analysis was used to construct a scoring model for predicting ≥F2 fibrosis. The diagnostic performance of the proposed model was compared with established noninvasive tests (NITs) and magnetic resonance elastography (MRE) for detecting both ≥F2 and ≥F3 fibrosis. Results: The total metabolic syndrome score was the only variable that significantly distinguished between F1 and F2 stages (p = 0.039). Among the biomarkers, matrix metalloproteinase-1 (MMP-1) showed a significant difference across fibrosis groups (p = 0.009). The AST/ALT ratio was the most robust predictor for differentiating ≥F3 (p < 0.001). A scoring model integrating the total metabolic syndrome score, MMP-1, and AST/ALT ratio demonstrated superior diagnostic accuracy for identifying ≥F2 (AUROC 0.88, 95% CI 0.79–0.97) compared to other NITs and MRE, and strong performance for detecting ≥F3 (AUROC 0.95, 95% CI 0.90–1.00). Conclusions: Total metabolic syndrome score and MMP-1 are promising candidates for future approaches. Combining total metabolic syndrome score, MMP-1, and AST/ALT ratio might detect ≥F2 in MASH with higher diagnostic accuracy than other NITs and MRE. Full article

Swept-source optical coherence tomography (SS-OCT) is a widely used imaging technique, particularly in medical diagnostics, due to its ability to provide high-resolution cross-sectional images. However, one of the main challenges in SS-OCT systems is the nonlinearity in wavelength sweeping, which leads to degraded depth resolution after Fourier transform. Correcting for this nonlinearity typically requires complex re-sampling and chirp compensation methods. In this paper, we introduce the first ultrafast time-stretch optical coherence tomography (TS-OCT) system that utilizes reservoir computing (RC) to perform direct temporal signal analysis without relying on Fourier transform techniques. By focusing solely on the temporal characteristics of the interference signal, regardless of frequency chirp, we demonstrate a more efficient solution to address the nonlinear wavelength sweeping issue. By leveraging the dynamic temporal processing capabilities of RC, the proposed system effectively bypasses the challenges faced by Fourier analysis, maintaining high-resolution depth measurement without being affected by chirp-introduced spectral broadening. The system operates by categorizing the interference signals generated by variations in sample position. This classification-based approach simplifies the data processing pipeline. We developed an RC-based model to interpret the temporal patterns in the interferometric signals, achieving high classification accuracy. A proof-of-the-concept experiment demonstrated that this method allows for precise depth resolution, independent of system chirp. With an A-scan rate of 50 MHz, the classification model yielded 100% accuracy with a root mean square error (RMSE) of 0.2416. This approach offers a robust alternative to Fourier-based analysis, particularly in systems prone to nonlinearities during signal acquisition. Full article

Objective: Although prior research developed an artificial intelligence (AI)-based classification system predicting colorectal lesion histology, the heavy computational demands limited its practical application. Recent advancements in medical AI emphasize decentralized architectures using edge computing devices, enhancing accessibility and real-time performance. This study aims to construct and evaluate a deep learning-based colonoscopy image classification model for automatic histologic categorization for real-time use on edge computing hardware. Design: We retrospectively collected 2418 colonoscopic images, subsequently dividing them into training, validation, and internal test datasets at a ratio of 8:1:1. Primary evaluation metrics included (1) classification accuracy across four histologic categories (advanced colorectal cancer, early cancer/high-grade dysplasia, tubular adenoma, and nonneoplasm) and (2) binary classification accuracy differentiating neoplastic from nonneoplastic lesions. Additionally, an external test was conducted using an independent dataset of 269 colonoscopic images. Results: For the internal-test dataset, the model achieved an accuracy of 83.5% (95% confidence interval: 78.8–88.2%) for the four-category classification. In binary classification (neoplasm vs. nonneoplasm), accuracy improved significantly to 94.6% (91.8–97.4%). The external test demonstrated an accuracy of 82.9% (78.4–87.4%) in the four-category task and a notably higher accuracy of 95.5% (93.0–98.0%) for binary classification. The inference speed of lesion classification was notably rapid, ranging from 2–3 ms/frame in GPU mode to 5–6 ms/frame in CPU mode. During real-time colonoscopy examinations, expert endoscopists reported no noticeable latency or interference from AI model integration. Conclusions: This study successfully demonstrates the feasibility of a deep learning-powered colonoscopy image classification system designed for the rapid, real-time histologic categorization of colorectal lesions on edge computing platforms. This study highlights how nature-inspired frameworks can improve the diagnostic capacities of medical AI systems by aligning technological improvements with biomimetic concepts. Full article