Malignant tumors are a major global health crisis, causing 25% of deaths in China, with lung, liver, thyroid, breast, and colon cancers being the most common. Understanding the factors influencing hospitalization costs for these cancers is crucial for public health and economics. This study aimed to identify key cost factors and develop a neural network model for predicting hospitalization costs, thereby providing tools to ease the financial burden on patients and healthcare systems.

Data on hospitalization costs for 30,893 cancer patients from secondary or higher-level hospitals in Zhuhai, Guangdong Province, between 2017 and 2022, were analyzed. Neural network classification and feature importance analysis were used to determine the main factors influencing costs and to develop predictive models. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC), with a 95% confidence interval (CI) calculated for the AUROC value.

The key factors influencing hospitalization costs for lung cancer are metastasis and malignant solid tumor (MST), with correlation coefficients of 0.126 and 0.086, respectively, both showing statistical significance (p < 0.05). For colon cancer, the key factors influencing hospitalization costs are mortality and coronary disease (CD), with correlation coefficients of 0.092 and 0.090, respectively, both demonstrating statistical significance (p < 0.05). The AUROC value for the lung cancer model is 0.9078 (95% CI = 0.8975-0.9186), and the AUROC value for the colon cancer model is 0.9017 (95% CI = 0.8848-0.9196).

This study confirmed the strong clinical applicability of the neural network predictive model in analyzing hospitalization costs for lung and colon cancer and revealed the factors that influence hospitalization costs for these cancers.

Atopic dermatitis (AD) is a common and complex skin disease that significantly affects the quality of life of patients. The latest advances in artificial intelligence (AI) technology have introduced new methods for diagnosing, treating, and managing AD. AI has various innovative applications in the diagnosis and treatment of atopic dermatitis, with particular emphasis on its significant benefits in medical diagnosis, treatment monitoring, and patient care. AI algorithms, especially those that use deep learning techniques, demonstrate strong performance in recognizing skin images and effectively distinguishing different types of skin lesions, including common AD manifestations. In addition, artificial intelligence has also shown promise in creating personalized treatment plans, simplifying drug development processes, and managing clinical trials. Despite challenges in data privacy and model transparency, the potential of artificial intelligence in advancing AD care is enormous, bringing the future to precision medicine and improving patient outcomes. This manuscript provides a comprehensive review of the application of AI in the process of AD disease for the first time, aiming to play a key role in the advancement of AI in skin health care and further enhance the clinical diagnosis and treatment of AD.

Over 20 million colonoscopies are performed in China annually. An automatic clinical decision support system (CDSS) with accurate semantic recognition of colonoscopy reports and guideline-based is helpful to relieve the increasing medical burden and standardize the healthcare. In this study, the CDSS was built under a hierarchical-label interpretable classification framework, trained by a state-of-the-art transformer-based model, and validated in a multi-center style.

We conducted stratified sampling on a previously established dataset containing 302,965 electronic colonoscopy reports with pathology, identified 2041 records representative of overall features, and randomly divided into the training and testing sets (7:3). A total of 5 main labels and 22 sublabels were applied to annotate each record on a network platform, and the data were trained respectively by three pre-training models on Chinese corpus website, including BERT-base-Chinese (BC), the BERT-wwm-ext-Chinese (BWEC), and ernie-3.0-base-zh (E3BZ). The performance of trained models was subsequently compared with a randomly initialized model, and the preferred model was selected. Model fine-tuning was applied to further enhance the capacity. The system was validated in five other hospitals with 3177 consecutive colonoscopy cases.

The E3BZ pre-trained model exhibited the best performance, with a 90.18% accuracy and a 69.14% Macro-F1 score overall. The model achieved 100% accuracy in identifying cancer cases and 99.16% for normal cases. In external validation, the model exhibited favorable consistency and good performance among five hospitals.

The novel CDSS possesses high-level semantic recognition of colonoscopy reports, provides appropriate recommendations, and holds the potential to be a powerful tool for physicians and patients. The hierarchical multi-label strategy and pre-training method should be amendable to manage more medical text in the future.

The current era of high-throughput analysis-driven research offers invaluable insights into disease etiologies, accurate diagnostics, pathogenesis, and personalized therapy. In the field of movement disorders, investigators are facing an increasing growth in the volume of produced patient-derived datasets, providing substantial opportunities for precision medicine approaches based on extensive information accessibility and advanced annotation practices. Integrating data from multiple sources, including phenomics, genomics, and multi-omics, is crucial for comprehensively understanding different types of movement disorders. Here, we explore formats and analytics of big data generated for patients with movement disorders, including strategies to meaningfully share the data for optimized patient benefit. We review computational methods that are essential to accelerate the process of evaluating the increasing amounts of specialized data collected. Based on concrete examples, we highlight how bioinformatic approaches facilitate the translation of multidimensional biological information into clinically relevant knowledge. Moreover, we outline the feasibility of computer-aided therapeutic target evaluation, and we discuss the importance of expanding the focus of big data research to understudied phenotypes such as dystonia.

Nutrition screening is a fundamental step to ensure appropriate intervention in patients with malnutrition. An automatic tool of nutritional risk screening based on electronic health records will improve efficiency and elevate the malnutrition diagnosis rate. We aimed to develop an artificial intelligence (AI) malnutrition screening tool based on electronic medical records and compare it with the patient interview-based tool.

We conducted a cross-sectional study at a comprehensive tertiary hospital in China. Data of malnutrition information were extracted from electronic health records (EHR) and were used to train and test an AI tool for the malnutritional risk screening. We adopted the GLIM framework as a reference standard for assessing malnutrition. Six widely used machine learning algorithms for auxiliary diagnosis prediction, including Support Vector Machine, Random Forest, extreme gradient boosting (XGBoost), Logistic Regression, AdaBoost, and Gradient Boosting were compared and visualized using SHapley Additive exPlanations (SHAP). After feature screening, simplified algoritms were cross validated at an independent data set.

495 inpatients enrolled were randomly divided into training and validation groups for algorithm development. 10 features annotated manually from free texts and 32 features selected from structured EHRs entered the models. XGBoost had the highest area under the receiver operating characteristic curve (AUC) and the top six features were weight loss, decreased food intake, prealbmine, white cell, BMI group, and percent of neutrophils. In simplified models, Random Forest acquired the highest AUC of 0.97 based on first sources data from interviews and 0.87 based on EHR data.

Inpatients' EHR data could be integrated by AI to detect the risk of malnutrition. This AI-enabled tool may hold promise for timely and efficient nutrition screening in newly admitted inpatients.

Currently, antimicrobial resistance (AMR) poses a major public health challenge. The emergence of AMR, which significantly threatens public health, is primarily due to the overuse of antimicrobial agents. This study explored the possibility that the ethical dilemmas inherent in the context of AMR may hinder the adoption of diagnostic artificial intelligence (AI). We conducted a web survey across eight countries/areas to assess public preference between two hypothetical AI types: one prioritizing individual health and the other considering the global AMR threat. Our results revealed a societal preference for the utilization of both AI types, reflecting a conflict between recognizing the significance of AMR and the desire for individualized treatment. Interestingly, the survey indicated significant gender and age differences in AI preferences, and the majority of respondents opposed the idea of AI standardization in treatment. These findings highlight the challenges of incorporating AI into public health and the necessity of considering public sentiment in addressing global health issues such as AMR.

Artificial intelligence makes strides in specialized diagnostics but faces challenges in complex clinical scenarios, such as rare disease diagnosis and emergency condition identification. To address these limitations, we develop Meta General Practitioner (MetaGP), a 32-billion-parameter generative foundation model trained on extensive datasets, including over 8 million electronic health records, biomedical literature, and medical textbooks. MetaGP demonstrates robust diagnostic capabilities, achieving accuracy comparable to experienced clinicians. In rare disease cases, it achieves an average diagnostic score of 1.57, surpassing GPT-4's 0.93. For emergency conditions, it improves diagnostic accuracy for junior and mid-level clinicians by 53% and 46%, respectively. MetaGP also excels in generating medical imaging reports, producing high-quality outputs for chest X-rays and computed tomography, often rated comparable to or superior to physician-authored reports. These findings highlight MetaGP's potential to transform clinical decision-making across diverse medical contexts.

To examine the quality of information provided by artificial intelligence platforms ChatGPT-4 and Claude 2 surrounding the management of vestibular schwannomas.

Cross-sectional.

Skull base surgeons were involved from different centers and countries.

Thirty-six questions regarding vestibular schwannoma management were tested. Artificial intelligence responses were subsequently evaluated by 19 lateral skull base surgeons using the Quality Assessment of Medical Artificial Intelligence (QAMAI) questionnaire, assessing "Accuracy," "Clarity," "Relevance," "Completeness," "Sources," and "Usefulness."

The scores of the answers from both chatbots were collected and analyzed using the Student t test. Analysis of responses grouped by stakeholders was performed with McNemar test. Stuart-Maxwell test was used to compare reading level among chatbots. Intraclass correlation coefficient was calculated.

ChatGPT-4 demonstrated significantly improved quality over Claude 2 in 14 of 36 (38.9%) questions, whereas higher-quality scores for Claude 2 were only observed in 2 (5.6%) answers. Chatbots exhibited variation across the dimensions of "Accuracy," "Clarity," "Completeness," "Relevance," and "Usefulness," with ChatGPT-4 demonstrating a statistically significant superior performance. However, no statistically significant difference was found in the assessment of "Sources." Additionally, ChatGPT-4 provided information at a significant lower reading grade level.

Artificial intelligence platforms failed to consistently provide accurate information surrounding the management of vestibular schwannoma, although ChatGPT-4 achieved significantly higher scores in most analyzed parameters. These findings demonstrate the potential for significant misinformation for patients seeking information through these platforms.

To evaluate the potential of large language models (LLMs) in health education for patients with ankylosing spondylitis (AS)/spondyloarthritis (SpA), focusing on the accuracy of information transmission, patient acceptance and performance differences between different models.

Cross-sectional, single-blind study.

Multiple centres in China.

182 volunteers, including 4 rheumatologists and 178 patients with AS/SpA.

Scientificity, precision and accessibility of the content of the answers provided by LLMs; patient acceptance of the answers.

LLMs performed well in terms of scientificity, precision and accessibility, with ChatGPT-4o and Kimi models outperforming traditional guidelines. Most patients with AS/SpA showed a higher level of understanding and acceptance of the responses from LLMs.

LLMs have significant potential in medical knowledge transmission and patient education, making them promising tools for future medical practice.

Surgical planning can be highly complicated and personalized, where a surgeon needs to balance multiple decisional dimensions including surgical effectiveness, risk, cost, and patient's conditions and preferences. Turning to artificial intelligence is a great appeal. This study filled in this gap with Multi-Dimensional Recommendation (MUDI), an interpretable data-driven intelligent system that supported personalized surgical recommendations on both the patient's and the surgeon's side with joint consideration of multiple decisional dimensions. Applied to Pelvic Organ Prolapse, a common female disease with significant impacts on life quality, MUDI stood out from a crowd of competing methods and achieved excellent performance that was comparable to top urogynecologists, with a transparent process that made communications between surgeons and patients easier. Users showed a willingness to accept the recommendations and achieved higher accuracy with the aid of MUDI. Such a success indicated that MUDI had the potential to solve similar challenges in other situations.

The objective of this study was to develop machine learning (ML) algorithms utilizing natural language processing (NLP) techniques for the automated detection of cervical spondylotic myelopathy (CSM) through the analysis of positive symptoms in free-text admission notes. This approach enables the timely identification and management of CSM, leading to optimal outcomes.

The dataset consisted of 1,214 patients diagnosed with cervical diseases as their primary condition between June 2013 and June 2020. A random ratio of 7:3 was employed to partition the dataset into training and testing subsets. Two machine learning models, Extreme Gradient Boosting (XGBoost) and Bidirectional Long Short Term Memory Network (LSTM), were developed. The performance of these models was assessed using various metrics, including the Receiver Operating Characteristic (ROC) curve, Area Under the Curve (AUC), accuracy, precision, recall, and F1 score.

In the testing set, the LSTM achieved an AUC of 0.9025, an accuracy of 0.8740, a recall of 0.9560, an F1 score of 0.9122, and a precision of 0.8723. The LSTM model demonstrated superior clinical applicability compared to the XGBoost model, as evidenced by calibration curves and decision curve analysis.

The timely identification of suspected CSM allows for prompt confirmation of diagnosis and treatment. The utilization of NLP algorithm demonstrated excellent discriminatory capabilities in identifying CSM based on positive symptoms in free-text admission notes complaint data. This study showcases the potential of a pre-diagnosis system in the field of spine.

The delivery of accurate diagnoses is crucial in healthcare and represents the gateway to appropriate and timely treatment. Although recent large language models (LLMs) have demonstrated impressive capabilities in few-shot or zero-shot learning, their effectiveness in clinical diagnosis remains unproven. Here we present MedFound, a generalist medical language model with 176 billion parameters, pre-trained on a large-scale corpus derived from diverse medical text and real-world clinical records. We further fine-tuned MedFound to learn physicians' inferential diagnosis with a self-bootstrapping strategy-based chain-of-thought approach and introduced a unified preference alignment framework to align it with standard clinical practice. Extensive experiments demonstrate that our medical LLM outperforms other baseline LLMs and specialized models in in-distribution (common diseases), out-of-distribution (external validation) and long-tailed distribution (rare diseases) scenarios across eight specialties. Further ablation studies indicate the effectiveness of key components in our medical LLM training approach. We conducted a comprehensive evaluation of the clinical applicability of LLMs for diagnosis involving artificial intelligence (AI) versus physician comparison, AI-assistance study and human evaluation framework. Our proposed framework incorporates eight clinical evaluation metrics, covering capabilities such as medical record summarization, diagnostic reasoning and risk management. Our findings demonstrate the model's feasibility in assisting physicians with disease diagnosis as part of the clinical workflow.

The incidence of non-melanoma skin cancers, encompassing basal cell carcinoma (BCC) and cutaneous squamous cell carcinoma (cSCC), is on the rise globally and new methods to improve skin malignancy diagnosis are necessary. This study aims to assess the performance of a CE-certified medical device as a diagnosis support tool in a head & neck (H&N) outpatient clinic, specifically focusing on the classification of three key diagnostics: BCC, cSCC, and non-malignant lesions (such as Actinic Cheilitis, Actinic Keratosis, and Seborrheic Keratosis).

a prospective, longitudinal, non-randomized study was designed to evaluate the performance of a deep learning-based method as a diagnosis tool in a group of patients referred to the head & neck clinic for suspicious skin lesions.

135 patients were included, 92 (68.1%) were male and 43 (31.9%) were female. The median age was 71 years +/- 9 (Min: 56/Max: 91). Of those, 108 were malignant pathologies (54 basal cell carcinoma and 54 squamous cell carcinoma) and 27 benign pathologies (14 seborrheic keratoses, 2 actinic keratoses, and 11 actinic cheilitis). Of special significance is the remarkable performance of the medical device in identifying malignant lesions (basal cell carcinoma and squamous cell carcinoma) within the top-5 most likely diagnoses in above 90% of cases, underscoring its potential utility for early diagnosis and treatment.

In this study, the effectiveness of deep learning methods, with a particular focus on vision transformers, as a diagnostic aid for H&N cutaneous non-melanoma skin cancers was demonstrated, highlighting its potential value for early detection and treatment of non-melanoma skin cancers. In this vein, further research is needed in the future to elucidate the role of this technology, because of its potential in the primary care clinic, dermatology, and head & neck surgery clinic as well as in patients with suspicious lesions, as a self-exploration tool.

Early diagnosis of biliary atresia (BA) is crucial for improving patient outcomes, yet remains a significant global challenge. This challenge may be ameliorated through the application of artificial intelligence (AI). Despite the promise of AI in medical diagnostics, its application to multimodal BA data has not yet achieved substantial breakthroughs. This study aims to leverage diverse data sources and formats to develop an intelligent diagnostic system for BA.

We constructed the largest known multimodal BA dataset, comprising ultrasound images, clinical data, and laboratory results. Using this dataset, we developed a novel deep learning model and simplified it using easily obtainable data, eliminating the need for blood samples. The models were externally validated in a prospective study. We compared the performance of our model with human experts of varying expertise levels and evaluated the AI system's potential to enhance its diagnostic accuracy.

The retrospective study included 1579 participants. The multimodal model achieved an AUC of 0.9870 on the internal test set, outperforming human experts. The simplified model yielded an AUC of 0.9799. In the prospective study's external test set of 171 cases, the multimodal model achieved an AUC of 0.9740, comparable to that of a radiologist with over 10 years of experience (AUC = 0.9766). For less experienced radiologists, the AI-assisted diagnostic AUC improved from 0.6667 to 0.9006.

This AI-based screening application effectively facilitates early diagnosis of BA and serves as a valuable reference for addressing common challenges in rare diseases. The model's high accuracy and its ability to enhance the diagnostic performance of human experts underscore its potential for significant clinical impact.

Early diagnosis and timely initiation of treatment plans for diabetes are crucial for ensuring individuals' well-being. Emerging technologies like artificial intelligence (AI) and computer vision are highly regarded for their ability to enhance the accessibility of large datasets for dynamic training and deliver efficient real-time intelligent technologies and predictable models. The application of AI and computer vision techniques to enhance the analysis of clinical data is referred to as eHealth solutions that employ advanced approaches to aid medical applications. This study examines several advancements and applications of machine learning, deep learning, and machine vision in global perception, with a focus on sustainability. This article discusses the significance of utilizing artificial intelligence and computer vision to detect diabetes, as it has the potential to significantly mitigate harm to human life. This paper provides several comments addressing challenges and recommendations for the use of this technology in the field of diabetes. This study explores the potential of employing Industry 4.0 technologies, including machine learning, deep learning, and computer vision robotics, as effective tools for effectively dealing with diabetes related aspects.

Pneumonia, caused by microorganisms, is a severely contagious disease that damages one or both the lungs of the patients. Early detection and treatment are typically favored to recover infected patients since untreated pneumonia can lead to major complications in the elderly (>65 years) and children (<5 years). The objectives of this work are to develop several models to evaluate big X-ray images (XRIs) of the chest, to determine whether the images show/do not show signs of pneumonia, and to compare the models based on their accuracy, precision, recall, loss, and receiver operating characteristic area under the ROC curve scores. Enhanced convolutional neural network (CNN), VGG-19, ResNet-50, and ResNet-50 with fine-tuning are some of the deep learning (DL) algorithms employed in this study. By training the transfer learning model and enhanced CNN model using a big data set, these techniques are used to identify pneumonia. The data set for the study was obtained from Kaggle. It should be noted that the data set has been expanded to include further records. This data set included 5863 chest XRIs, which were categorized into 3 different folders (i.e., train, val, test). These data are produced every day from personnel records and Internet of Medical Things devices. According to the experimental findings, the ResNet-50 model showed the lowest accuracy, that is, 82.8%, while the enhanced CNN model showed the highest accuracy of 92.4%. Owing to its high accuracy, enhanced CNN was regarded as the best model in this study. The techniques developed in this study outperformed the popular ensemble techniques, and the models showed better results than those generated by cutting-edge methods. Our study implication is that a DL models can detect the progression of pneumonia, which improves the general diagnostic accuracy and gives patients new hope for speedy treatment. Since enhanced CNN and ResNet-50 showed the highest accuracy compared with other algorithms, it was concluded that these techniques could be effectively used to identify pneumonia after performing fine-tuning.

The aging population and increasing chronic diseases strain public health systems. Advancements in digital health promise to tackle these challenges and enhance public health outcomes. Digital health integrates digital health technology (DHT) across healthcare, including smart consumer devices. This article examines the application of DHT in public health and its significant impact on revolutionizing the field. Historically, DHT has not only enhanced the efficiency of disease prevention, diagnosis, and treatment but also facilitated the equitable distribution of global health resources. Looking ahead, DHT holds vast potential in areas such as personalized medicine, telemedicine, and intelligent health management. However, it also encounters challenges such as ethics, privacy, and data security. To further advance DHT, concerted efforts are essential, including policy support, investment in research and development, involvement of medical institutions, and improvement of public digital health literacy.

Functional Gastrointestinal Disorders (FGIDs) can pose a great burden on affected children, their families, and the healthcare system. Due to the lack of knowledge about the precise pathophysiology of FGIDs, a proper identification of children at risk to develop FGIDs has never been attempted. The research aims to identify early-life risk factors for FGIDs such as infantile colic, regurgitation, and functional constipation, within the first year of life.

This prospective observational cohort study enrolled both term and preterm infants from a tertiary care university hospital between January 1, 2020, and December 31, 2022. The study employed both traditional statistical methods and artificial intelligence (AI) techniques, specifically a random forest classification model, to identify key risk factors associated with the development of FGIDs. Based on these findings, an AI-based predictive model will be developed, along with a user-friendly, web-based interface designed for practical risk assessment.

6060 infants were enrolled. 8.1 % were born preterm. According to random forest classification model by AI, birth weight (BW), cord blood pH, and maternal age were the most relevant variables linked to development of FGIDs in the first year of life. Some discrepancies between potential risk factors identified through conventional statistics and AI were detected.

For the first time machine learning allowed to identify BW, cord blood pH and maternal age as important variable for risk prediction of FGIDs in the first year of life. This practical risk assessment tool would help clinicians to identify infants at risk of FGIDs who would benefit from a tailored preventive approach.

To construct a diagnostic model for mixed dentition using a multistage deep-learning network to predict potential ectopic eruption in permanent teeth by integrating dentition segmentation into the process of automatic classification of dental development stages.

A database was established by reviewing 1576 anonymous panoramic radiographs of children aged 6-12 years, collected at the Stomatology Hospital, Zhejiang University School of Medicine. These radiographs were categorised as normal or ectopic eruption, with expert diagnoses serving as a benchmark for training and evaluating artificial intelligence (AI) models. Furthermore, tooth boundaries and dental development stages were manually annotated by three pediatric dentistry experts. The dataset was split into training, validation, and test sets at an 8:1:1 ratio.

The diagnostic performance of the deep-learning model was rigorously evaluated. The model demonstrated accuracy in tooth segmentation, with Intersection over Union, precision, sensitivity, and F1 scores of 0.959, 0.993, 0.966, and 0.979, respectively. Furthermore, its ability to identify tooth ectopic eruptions on panoramic radiographs, when compared to evaluations by three dentists. Based on McNemar's test, the model's specificity and accuracy in identifying ectopic tooth eruptions on the test dataset surpassed that of Dentist 1 (P < 0.05), while no significant difference was observed compared to the other two dentists. Besides, the deep learning model also showed its potential in classifying dental development stages, as tested against three different standards.

The adaptability of the AI-enabled model in this study was demonstrated across multiple scenarios, with clinical validation highlighting its efficacy in diagnosing ectopic eruptions using a multistage deep-learning approach.

Our findings provide new insights and technical support for the prevention and treatment of abnormal tooth eruption, laying the groundwork for predictive models for other prevalent pediatric dentistry conditions.

 This study aimed to evaluate critical care provider perspectives about diagnostic practices for rare and atypical infections and the potential for using artificial intelligence (AI) as a decision support system (DSS).

 We conducted an anonymous web-based survey among critical care providers at Mayo Clinic Rochester between November 25, 2023, and January 15, 2024, to evaluate their experience with rare and atypical infection diagnostic processes and AI-based DSSs. We also assessed the perceived usefulness of AI-based DSSs, their potential impact on improving diagnostic practices for rare and atypical infections, and the perceived risks and benefits of their use.

 A total of 47/143 providers completed the survey. Thirty-eight out of 47 agreed that there was a delay in diagnosing rare and atypical infections. Among those who agreed, limited assessment of specific patient factors and failure to consider them were the most frequently cited important contributing factors (33/38). Thirty-eight out of 47 reported familiarity with the AI-based DSS applications available to critical care providers. Less than half (18/38) thought AI-based DSSs often provided valuable insights into patient care, but almost three-quarters (34/47) thought AI-based DDSs often provided valuable insight when specifically asked about their ability to improve the diagnosis of rare and atypical infections. All respondents rated reliability as important in enhancing the perceived utility of AI-based DSSs (47/47) and almost all rated interpretability and integration into the workflow as important (45/47). The primary concern about implementing an AI-based DSS in this context was alert fatigue (44/47).

 Most critical care providers perceive that there are delays in diagnosing rare infections, indicating inadequate assessment and consideration of the diagnosis as the major contributors. Reliability, interpretability, workflow integration, and alert fatigue emerged as key factors impacting the usability of AI-based DSS. These findings will inform the development and implementation of an AI-based diagnostic algorithm to aid in identifying rare and atypical infections.

Colorectal cancer (CRC) is characterized by high heterogeneity, aggressiveness, and high morbidity and mortality rates. With machine learning (ML) algorithms, patient, tumor, and treatment features can be used to develop and validate models for predicting survival. In addition, important variables can be screened and different applications can be provided that could serve as vital references when making clinical decisions and potentially improving patient outcomes in clinical settings.

To construct prognostic prediction models and screen important variables for patients with stage I to III CRC.

More than 1000 postoperative CRC patients were grouped according to survival time (with cutoff values of 3 years and 5 years) and assigned to training and testing cohorts (7:3). For each 3-category survival time, predictions were made by 4 ML algorithms (all-variable and important variable-only datasets), each of which was validated via 5-fold cross-validation and bootstrap validation. Important variables were screened with multivariable regression methods. Model performance was evaluated and compared before and after variable screening with the area under the curve (AUC). SHapley Additive exPlanations (SHAP) further demonstrated the impact of important variables on model decision-making. Nomograms were constructed for practical model application.

Our ML models performed well; the model performance before and after important parameter identification was consistent, and variable screening was effective. The highest pre- and postscreening model AUCs 95% confidence intervals in the testing set were 0.87 (0.81-0.92) and 0.89 (0.84-0.93) for overall survival, 0.75 (0.69-0.82) and 0.73 (0.64-0.81) for disease-free survival, 0.95 (0.88-1.00) and 0.88 (0.75-0.97) for recurrence-free survival, and 0.76 (0.47-0.95) and 0.80 (0.53-0.94) for distant metastasis-free survival. Repeated cross-validation and bootstrap validation were performed in both the training and testing datasets. The SHAP values of the important variables were consistent with the clinicopathological characteristics of patients with tumors. The nomograms were created.

We constructed a comprehensive, high-accuracy, important variable-based ML architecture for predicting the 3-category survival times. This architecture could serve as a vital reference for managing CRC patients.

Controlling colonoscopic quality is important in the detection of colon polyps during colonoscopy as it reduces the overall long-term colorectal cancer risk. Artificial intelligence has recently been introduced in various medical fields. In this study, we aimed to validate a previously developed artificial intelligence (AI) computer-aided detection (CADe) algorithm called ALPHAON® and compare outcomes with previous studies that showed that AI outperformed and assisted endoscopists of diverse levels of expertise in detecting colon polyps.

We used the retrospective data of 500 still images, including 100 polyp images and 400 healthy colon images. In addition, we validated the CADe algorithm and compared its diagnostic performance with that of two expert endoscopists and six trainees from Gachon University Gil Medical Center. After a washing-out period of over 2 weeks, endoscopists performed polyp detection on the same dataset with the assistance of ALPHAON®.

The CADe algorithm presented a high capability in detecting colon polyps, with an accuracy of 0.97 (95% CI: 0.96 to 0.99), sensitivity of 0.91 (95% CI: 0.85 to 0.97), specificity of 0.99 (95% CI: 0.97 to 0.99), and AUC of 0.967. When evaluating and comparing the polyp detection ability of ALPHAON® with that of endoscopists with different levels of expertise (regarding years of endoscopic experience), it was found that ALPHAON® outperformed the experts in accuracy (0.97, 95% CI: 0.96 to 0.99), sensitivity (0.91, 95% CI: 0.85 to 0.97), and specificity (0.99, 95% CI: 0.97 to 0.99). After a washing-out period of over 2 weeks, the overall capability significantly improved for both experts and trainees with the assistance of ALPHAON®.

The high performance of the CADe algorithm system in colon polyp detection during colonoscopy was verified. The sensitivity of ALPHAON® led to it outperforming the experts, and it demonstrated the ability to enhance the polyp detection ability of both experts and trainees, which suggests a significant possibility of ALPHAON® being able to provide endoscopic assistance.

The array of complex and evolving patient data has limited clinical decision making in the emergency department (ED). This study introduces an advanced deep learning algorithm designed to enhance real-time prediction accuracy for integration into a novel Clinical Decision Support System (CDSS). A retrospective study was conducted using data from a level 1 tertiary hospital. The algorithm's predictive performance was evaluated based on in-hospital cardiac arrest, inotropic circulatory support, advanced airway, and intensive care unit admission. We developed an artificial intelligence (AI) algorithm for CDSS that integrates multiple data modalities, including vitals, laboratory, and imaging results from electronic health records. The AI model was trained and tested on a dataset of 237,059 ED visits. The algorithm's predictions, based solely on triage information, significantly outperformed traditional logistic regression models, with notable improvements in the area under the precision-recall curve (AUPRC). Additionally, predictive accuracy improved with the inclusion of continuous data input at shorter intervals. This study suggests the feasibility of using AI algorithms in diverse clinical scenarios, particularly for earlier detection of clinical deterioration. Future work should focus on expanding the dataset and enhancing real-time data integration across multiple centers to further optimize its application within the novel CDSS.

The ageing global population puts heavy pressure on healthcare systems everywhere. Addressing ageing-related chronic conditions requires employment of novel innovative solutions. Telehealth technologies, including telepresence robots (TPRs), are being rapidly developed to provide healthcare services efficiently wherever needed. This article explores the role of TPRs in addressing the challenges of providing healthcare to an ageing population, emphasizing their potential advantages and drawbacks. Employing an exploratory research approach with qualitative data collection techniques, we tested three TPR usage scenarios in simulated healthcare settings: anamnesis, measurements, and falls and frailty. The study employed a non-random purposive sample comprising 25 participants, and was conducted at a medical facility in June 2023. The findings suggest that TPRs offer promising solutions for healthcare professionals and patients, especially in scenarios when physical presence is impossible or physical isolation is required to prevent contagion. However, the technology is not yet ready to substitute fully human medical workers, potentially causing patient reluctance and emphasizing the need for patient-centered approaches to technology adoption. In addition, more studies are needed to address ethical, privacy, and scalability concerns.

Gastric cancer ranks fifth for incidence and fourth in the leading causes of mortality worldwide. In this study, we aimed to validate previously developed artificial intelligence (AI) computer-aided detection (CADe) algorithm, called ALPHAON® in detecting gastric neoplasm.

We used the retrospective data of 500 still images, including 5 benign gastric ulcers, 95 with gastric cancer, and 400 normal images. Thereby we validated the CADe algorithm measuring accuracy, sensitivity, and specificity with the result of receiver operating characteristic curves (ROC) and area under curve (AUC) in addition to comparing the diagnostic performance status of four expert endoscopists, four trainees, and four beginners from two university-affiliated hospitals with CADe algorithm. After a washing-out period of over 2 weeks, endoscopists performed gastric detection on the same dataset of the 500 endoscopic images again marked by ALPHAON®.

The CADe algorithm presented high validity in detecting gastric neoplasm with accuracy (0.88, 95% CI: 0.85 to 0.91), sensitivity (0.93, 95% CI: 0.88 to 0.98), specificity (0.87, 95% CI: 0.84 to 0.90), and AUC (0.962). After a washing-out period of over 2 weeks, overall validity improved in the trainee and beginner groups with the assistance of ALPHAON®. Significant improvement was present, especially in the beginner group (accuracy 0.94 (0.93 to 0.96) p < 0.001, sensitivity 0.87 (0.82 to 0.92) p < 0.001, specificity 0.96 (0.95 to 0.97) p < 0.001).

The high validation performance state of the CADe algorithm system was verified. Also, ALPHAON® has demonstrated its potential to serve as an endoscopic educator for beginners improving and making progress in sensitivity and specificity.

Despite significant strides in big data technology, extracting information from unstructured clinical data remains a formidable challenge. This study investigated the utility of large language models (LLMs) for extracting clinical data from unstructured radiological reports without additional training. In this retrospective study, 1800 radiologic reports, 600 from each of the three university hospitals, were collected, with seven pulmonary outcomes defined. Three pulmonology-trained specialists discerned the presence or absence of diseases. Data extraction from the reports was executed using Google Gemini Pro 1.0, OpenAI's GPT-3.5, and GPT-4. The gold standard was predicated on agreement between at least two pulmonologists. This study evaluated the performance of the three LLMs in diagnosing seven pulmonary diseases (active tuberculosis, emphysema, interstitial lung disease, lung cancer, pleural effusion, pneumonia, and pulmonary edema) utilizing chest radiography and computed tomography scans. All models exhibited high accuracy (0.85-1.00) for most conditions. GPT-4 consistently outperformed its counterparts, demonstrating a sensitivity of 0.71-1.00; specificity of 0.89-1.00; and accuracy of 0.89 and 0.99 across both modalities, thus underscoring its superior capability in interpreting radiological reports. Notably, the accuracy of pleural effusion and emphysema on chest radiographs and pulmonary edema on chest computed tomography scans reached 0.99. The proficiency of LLMs, particularly GPT-4, in accurately classifying unstructured radiological data hints at their potential as alternatives to the traditional manual chart reviews conducted by clinicians.

Bronchopulmonary dysplasia (BPD) remains a significant complication of prematurity, impacting approximately 18,000 infants annually in the United States. Advances in neonatal care have not reduced BPD, and its management is challenged by the rising survival of extremely premature infants and the variability in clinical practices. Leveraging statistical and machine learning techniques, predictive analytics can enhance BPD management by utilizing large clinical datasets to predict individual patient outcomes. This review explores the foundations and applications of predictive analytics in the context of BPD, examining commonly used data sources, modeling techniques, and metrics for model evaluation. We also highlight bioinformatics' potential role in understanding BPD's molecular basis and discuss case studies demonstrating the use of machine learning models for risk prediction and prognosis in neonates. Challenges such as data bias, model complexity, and ethical considerations are outlined, along with strategies to address these issues. Future directions for advancing the integration of predictive analytics into clinical practice include improving model interpretability, expanding data sharing and interoperability, and aligning predictive models with precision medicine goals. By overcoming current challenges, predictive analytics holds promise for transforming neonatal care and providing personalized interventions for infants at risk of BPD.

Given the limited capacity to accurately determine the necessity for intubation in intensive care unit settings, this study aimed to develop and externally validate an interpretable machine learning model capable of predicting the need for intubation among ICU patients. Seven widely used machine learning (ML) algorithms were employed to construct the prediction models. Adult patients from the Medical Information Mart for Intensive Care IV database who stayed in the ICU for longer than 24 h were included in the development and internal validation. The model was subsequently externally validated using the eICU-CRD database. In addition, the SHapley Additive exPlanations method was employed to interpret the influence of individual parameters on the predictions made by the model. A total of 11,988 patients were included in the final cohort for this study. The CatBoost model demonstrated the best performance (AUC: 0.881). In the external validation set, the efficacy of our model was also confirmed (AUC: 0.750), which suggests robust generalization capabilities. The Glasgow Coma Scale (GCS), body mass index (BMI), arterial partial pressure of oxygen (PaO2), respiratory rate (RR) and length of stay (LOS) before ICU were the top 5 features of the CatBoost model with the greatest impact. We developed an externally validated CatBoost model that accurately predicts the need for intubation in ICU patients within 24 to 96 h of admission, facilitating clinical decision-making and has the potential to improve patient outcomes. The prediction model utilizes readily obtainable monitoring parameters and integrates the SHAP method to enhance interpretability, providing clinicians with clear insights into the factors influencing predictions.

Patients with rheumatoid arthritis (RA) have increased mortality and morbidity rates owing to cardiovascular diseases (CVD). Timely detection of CVD in RA can greatly improve patient prognosis; however, this technique remains challenging. We aimed to investigate the risk factors for CVD incidence in patients with RA.

This retrospective study included RA patients without CVD risk factors (n = 402), RA with CVD risk factors (n = 394), and RA with CVD (n = 201). Their data on routine examination indicators, vascular endothelial growth factor (VEGF), and immune cells were obtained from medical records. The characteristic variables between each group were screened using univariate analysis, least absolute shrinkage and selection operator (LASSO), random forest (RF), and logistic regression (LR) models, and individualized nomograms were further established to more conveniently observe the likelihood of CVD in RA.

Univariate analysis revealed significantly elevated levels of white blood cells (WBC), blood urea nitrogen (BUN), creatinine, creatine kinase (CK), lactate dehydrogenase (LDH), VEGF, serum total cholesterol (TC), triglycerides (TG), low-density lipoprotein (LDL), apolipoprotein B100 (ApoB100), and apolipoprotein E (ApoE) in RA patients with CVD, whereas apolipoprotein A1 (ApoA1) and high-density lipoprotein/cholesterol (HDL/TC) were decreased. Furthermore, the ratio of regulatory T (Treg) cells exhibiting excellent separation performance in RA patients with CVD was significantly lower than that in other groups, whereas the ratios of Th1/Th2/NK and Treg cells were significantly elevated. The LASSO, RF, and LR models were also used to identify the risk factors for CVD in patients with RA. Through the final selected indicators screened using the three machine learning models and univariate analysis, a convenient nomogram was established to observe the likelihood of CVD in patients with RA.

Serum lipids, lipoproteins, and reduction of Treg cells have been identified as risk factors for CVD in patients with RA. Three nomograms combining various risk factors were constructed to predict CVD occurring in patients with RA (RA with/without CVD risk factors).

Children with atrial septal defect (ASD) and ventricular septal defect (VSD) are frequently examined for respiratory symptoms, even when the underlying disease is not found. Chest radiographs often serve as the primary imaging modality. It is crucial to differentiate between ASD and VSD due to their distinct treatment.

To assess whether deep learning analysis of chest radiographs can more effectively differentiate between ASD and VSD in children.

In this retrospective study, chest radiographs and corresponding radiology reports from 1,194 patients were analyzed. The cases were categorized into a training set and a validation set, comprising 480 cases of ASD and 480 cases of VSD, and a test set with 115 cases of ASD and 119 cases of VSD. Four deep learning network models-ResNet-CBAM, InceptionV3, EfficientNet, and ViT-were developed for training, and a fivefold cross-validation method was employed to optimize the models. Receiver operating characteristic (ROC) curve analyses were conducted to assess the performance of each model. The most effective algorithm was compared with the interpretations provided by two radiologists on 234 images from the test group.

The average accuracy, sensitivity, and specificity of the four deep learning models in the differential diagnosis of VSD and ASD were higher than 70%. The AUC values of ResNet-CBAM, IncepetionV3, EfficientNet, and ViT were 0.87, 0.91, 0.90, and 0.66, respectively. Statistical analysis showed that the differential diagnosis efficiency of InceptionV3 was the highest, reaching 87% classification accuracy. The accuracy of InceptionV3 in the differential diagnosis of VSD and ASD was higher than that of the radiologists.

Deep learning methods such as IncepetionV3 based on chest radiographs in the study showed good performance for differential diagnosis of congenital VSD and ASD, which may be able to assist radiologists in diagnosis, education, and training, and reduce missed diagnosis and misdiagnosis.

A key best practice to prevent and treat pressure injuries (PIs) is to ensure at-risk individuals are repositioned regularly. Our team designed a non-contact position detection system that predicts an individual's position in bed using data from load cells under the bed legs. The system was originally designed to predict the individual's position as left-side lying, right-side lying, or supine. Our previous work suggested that a higher precision for detecting position (classifying more than three positions) may be needed to determine whether key bony prominences on the pelvis at high risk of PIs have been off-loaded. The objective of this study was to determine the impact of categorizing participant position with higher precision using the system prediction F1 score. Data from 18 participants was collected from four load cells placed under the bed legs and a pelvis-mounted inertial measurement unit while the participants assumed 21 positions. The data was used to train classifiers to predict the participants' transverse pelvic angle using three different position bin sizes (45°, ~30°, and 15°). A leave-one-participant-out cross validation approach was used to evaluate classifier performance for each bin size. Results indicated that our prediction F1 score dropped as the position category precision was increased.

Artificial intelligence (AI) technologies, particularly large language models (LLMs), have been widely employed by the medical community. In addressing the intricacies of urology, ChatGPT offers a novel possibility to aid in clinical decision-making. This study aimed to investigate the decision-making ability of LLMs in solving complex urology-related problems and assess their effectiveness in providing psychological support to patients with urological disorders.

This study evaluated the clinical and psychological support capabilities of ChatGPT 3.5 and 4.0 in the field of urology. A total of 69 clinical and 30 psychological questions were posed to the AI models, and both urologists and psychologists evaluated their response. As a control, clinicians from Chinese medical institutions responded to closed-book conditions. Statistical analyses were conducted separately for each subgroup.

In multiple-choice tests covering diverse urological topics, ChatGPT 4.0 was performed comparably to the physician group, with no significant overall score difference. Subgroup analyses revealed variable performance based on disease type and physician experience, with ChatGPT 4.0 generally outperforming ChatGPT 3.5 and exhibiting competitive results against physicians. When assessing the psychological support capabilities of AI, it is evident that ChatGPT 4.0 outperforms ChatGPT 3.5 across all urology-related psychological problems.

The performance of LLMs in dealing with standardized clinical problems and providing psychological support has certain advantages over clinicians. AI stands out as a promising tool for potential clinical aid.

Pneumonia is a common postoperative complication in patients with aneurysmal subarachnoid hemorrhage (aSAH), which is associated with poor prognosis and increased mortality. The aim of this study was to develop a predictive model for postoperative pneumonia (POP) in patients with aSAH. A retrospective analysis was conducted on 308 patients with aSAH who underwent surgery at the Neurosurgery Department of the First Affiliated Hospital of Soochow University. Univariate and multivariate logistic regression and lasso regression analysis were used to analyze the risk factors for POP. Receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) were used to evaluate the constructed model. Finally, the effectiveness of modeling these six variables in different machine learning methods was investigated. In our patient cohort, 23.4% (n = 72/308) of patients experienced POP. Univariate, multivariate logistic regression analysis and lasso regression analysis revealed age, Hunt-Hess grade, mechanical ventilation, leukocyte count, lymphocyte count, and platelet count as independent risk factors for POP. Subsequently, these six factors were used to build the final model. We found that age, Hunt-Hess grade, mechanical ventilation, leukocyte count, lymphocyte count, and platelet count were independent risk factors for POP in patients with aSAH. Through validation and comparison with other studies and machine learning models, our novel predictive model has demonstrated high efficacy in effectively predicting the likelihood of pneumonia during the hospitalization of aSAH patients.

Glaucoma (GLAU), Age-related Macular Degeneration (AMD), Retinal Vein Occlusion (RVO), and Diabetic Retinopathy (DR) are common blinding ophthalmic diseases worldwide.

This approach is expected to enhance the early detection and treatment of common blinding ophthalmic diseases, contributing to the reduction of individual and economic burdens associated with these conditions.

We propose an effective deep-learning pipeline that combine both segmentation model and classification model for diagnosis and grading of four common blinding ophthalmic diseases and normal retinal fundus.

In total, 102,786 fundus images of 75,682 individuals were used for training validation and external validation purposes. We test our model on internal validation data set, the micro Area Under the Receiver Operating Characteristic curve (AUROC) of which reached 0.995. Then, we fine-tuned the diagnosis model to classify each of the four disease into early and late stage, respectively, which achieved AUROCs of 0.597 (GL), 0.877 (AMD), 0.972 (RVO), and 0.961 (DR) respectively. To test the generalization of our model, we conducted two external validation experiments on Neimeng and Guangxi cohort, all of which maintained high accuracy.

Our algorithm demonstrates accurate artificial intelligence diagnosis pipeline for common blinding ophthalmic diseases based on Lesion-Focused fundus that overcomes the low-accuracy of the traditional classification method that based on raw retinal images, which has good generalization ability on diverse cases in different regions.

Within the ethical debate on Machine Learning-driven decision support systems (ML_CDSS), notions such as "human in the loop" or "meaningful human control" are often cited as being necessary for ethical legitimacy. In addition, ethical principles usually serve as the major point of reference in ethical guidance documents, stating that conflicts between principles need to be weighed and balanced against each other. Starting from a neo-Kantian viewpoint inspired by Onora O'Neill, this article makes a concrete suggestion of how to interpret the role of the "human in the loop" and to overcome the perspective of rivaling ethical principles in the evaluation of AI in health care. We argue that patients should be perceived as "fellow workers" and epistemic partners in the interpretation of ML_CDSS outputs. We further highlight that a meaningful process of integrating (rather than weighing and balancing) ethical principles is most appropriate in the evaluation of medical AI.

For medical diagnosis, clinicians typically begin with a patient's chief concerns, followed by questions about symptoms and medical history, physical examinations, and requests for necessary auxiliary examinations to gather comprehensive medical information. This complex medical investigation process has yet to be modeled by existing artificial intelligence (AI) methodologies.

The aim of this study was to develop an AI-driven medical inquiry assistant for clinical diagnosis that provides inquiry recommendations by simulating clinicians' medical investigating logic via reinforcement learning.

We compiled multicenter, deidentified outpatient electronic health records from 76 hospitals in Shenzhen, China, spanning the period from July to November 2021. These records consisted of both unstructured textual information and structured laboratory test results. We first performed feature extraction and standardization using natural language processing techniques and then used a reinforcement learning actor-critic framework to explore the rational and effective inquiry logic. To align the inquiry process with actual clinical practice, we segmented the inquiry into 4 stages: inquiring about symptoms and medical history, conducting physical examinations, requesting auxiliary examinations, and terminating the inquiry with a diagnosis. External validation was conducted to validate the inquiry logic of the AI model.

This study focused on 2 retrospective inquiry-and-diagnosis tasks in the emergency and pediatrics departments. The emergency departments provided records of 339,020 consultations including mainly children (median age 5.2, IQR 2.6-26.1 years) with various types of upper respiratory tract infections (250,638/339,020, 73.93%). The pediatrics department provided records of 561,659 consultations, mainly of children (median age 3.8, IQR 2.0-5.7 years) with various types of upper respiratory tract infections (498,408/561,659, 88.73%). When conducting its own inquiries in both scenarios, the AI model demonstrated high diagnostic performance, with areas under the receiver operating characteristic curve of 0.955 (95% CI 0.953-0.956) and 0.943 (95% CI 0.941-0.944), respectively. When the AI model was used in a simulated collaboration with physicians, it notably reduced the average number of physicians' inquiries to 46% (6.037/13.26; 95% CI 6.009-6.064) and 43% (6.245/14.364; 95% CI 6.225-6.269) while achieving areas under the receiver operating characteristic curve of 0.972 (95% CI 0.970-0.973) and 0.968 (95% CI 0.967-0.969) in the scenarios. External validation revealed a normalized Kendall τ distance of 0.323 (95% CI 0.301-0.346), indicating the inquiry consistency of the AI model with physicians.

This retrospective analysis of predominantly respiratory pediatric presentations in emergency and pediatrics departments demonstrated that an AI-driven diagnostic assistant had high diagnostic performance both in stand-alone use and in simulated collaboration with clinicians. Its investigation process was found to be consistent with the clinicians' medical investigation logic. These findings highlight the diagnostic assistant's promise in assisting the decision-making processes of health care professionals.