Machine Learning in Medicine

Publicly available datasets of breast histopathology H&E whole-slide images: A scoping review

Advancements in digital pathology and computing resources have made a significant impact in the field of computational pathology for breast cancer diagnosis and treatment. However, access to high-quality labeled histopathological images of breast cancer is a big challenge that limits the development of accurate and robust deep learning models. In this scoping review, we identified the publicly available datasets of breast H&E-stained whole-slide images (WSIs) that can be used to develop deep learning algorithms. We systematically searched 9 scientific literature databases and 9 research data repositories and found 17 publicly available datasets containing 10 385 H&E WSIs of breast cancer. Moreover, we reported image metadata and characteristics for each dataset to assist researchers in selecting proper datasets for specific tasks in breast cancer computational pathology. In addition, we compiled 2 lists of breast H&E patches and private datasets as supplementary resources for researchers. Notably, only 28% of the included articles utilized multiple datasets, and only 14% used an external validation set, suggesting that the performance of other developed models may be susceptible to overestimation. The TCGA-BRCA was used in 52% of the selected studies. This dataset has a considerable selection bias that can impact the robustness and generalizability of the trained algorithms. There is also a lack of consistent metadata reporting of breast WSI datasets that can be an issue in developing accurate deep learning models, indicating the necessity of establishing explicit guidelines for documenting breast WSI dataset characteristics and metadata.

Unraveling the genetic and molecular landscape of sepsis and acute kidney injury: A comprehensive GWAS and machine learning approach

OBJECTIVES

This study aimed to explore the underlying mechanisms of sepsis and acute kidney injury (AKI), including sepsis-associated AKI (SA-AKI), a frequent complication in critically ill sepsis patients.

METHODS

GWAS data was analyzed for genetic association between AKI and sepsis. Then, we systematically applied three distinct machine learning algorithms (LASSO, SVM-RFE, RF) to rigorously identify and validate signature genes of SA-AKI, assessing their diagnostic and prognostic value through ROC curves and survival analysis. The study also examined the functional and immunological aspects of these genes, potential drug targets, and ceRNA networks. A mouse model of sepsis was created to test the reliability of these signature genes.

RESULTS

LDSC confirmed a positive genetic correlation between AKI and sepsis, although no significant shared loci were found. Bidirectional MR analysis indicated mutual increased risks of AKI and sepsis. Then, 311 key genes common to sepsis and AKI were identified, with 42 significantly linked to sepsis prognosis. Six genes, selected through LASSO, SVM-RFE, and RF algorithms, showed excellent predictive performance for sepsis, AKI, and SA-AKI. The models demonstrated near-perfect AUCs in both training and testing datasets, and a perfect AUC in a sepsis mouse model. Significant differences in immune cells, immune-related pathways, HLA, and checkpoint genes were found between high- and low-risk groups. The study identified 62 potential drug treatments for sepsis and AKI and constructed a ceRNA network.

CONCLUSIONS

The identified signature genes hold potential clinical applications, including prognostic evaluation and targeted therapeutic strategies for sepsis and AKI. However, further research is needed to confirm these findings.

Can Chat-GPT read and understand guidelines? An example using the S2k guideline intrauterine growth restriction of the German Society for Gynecology and Obstetrics

PURPOSE

To investigate the capacity of chat-generative pre-trained transformer (Chat-GPT) to understand the S2k guideline of the German Society for Gynecology and Obstetrics on intrauterine growth restriction.

METHODS

The German-language free Chat-GPT version was used to test the ability of Chat-GPT to understand the definition of small for gestational age and intrauterine growth restriction, to indicate the correct time and place of delivery and to evaluate ist ability to recommend a spontaneous delivery versus a primary caesarean section in accordance with the guideline recommendations. In order to objectively evaluate the suggestions a simple three-color 'traffic light' evaluation system was employed.

RESULTS

Almost all Chat-GPT's suggestions in the context of definition of small for gestational age/intrauterine growth restriction as well as correct time of delivery were adequate, whereas more than half of the suggestions made in terms of correct delivery mode needed reformulation or even correction.

CONCLUSION

Chat-GPT appears to be a valuable form of artificial intelligence that could be integrated into everyday clinical practice.

Action tremor features discovery for essential tremor and Parkinson's disease with explainable multilayer BiLSTM

The tremors of Parkinson's disease (PD) and essential tremor (ET) are known to have overlapping characteristics that make it complicated for clinicians to distinguish them. While deep learning is robust in detecting features unnoticeable to humans, an opaque trained model is impractical in clinical scenarios as coincidental correlations in the training data may be used by the model to make classifications, which may result in misdiagnosis. This work aims to overcome the aforementioned challenge of deep learning models by introducing a multilayer BiLSTM network with explainable AI (XAI) that can better explain tremulous characteristics and quantify the respective discovered important regions in tremor differentiation. The proposed network classifies PD, ET, and normal tremors during drinking actions and derives the contribution from tremor characteristics, (i.e., time, frequency, amplitude, and actions) utilized in the classification task. The analysis shows that the XAI-BiLSTM marks the regions with high tremor amplitude as important in classification, which is verified by a high correlation between relevance distribution and tremor displacement amplitude. The XAI-BiLSTM discovered that the transition phases from arm resting to lifting (during the drinking cycle) is the most important action to classify tremors. Additionally, the XAI-BiLSTM reveals frequency ranges that only contribute to the classification of one tremor class, which may be the potential distinctive feature to overcome the overlapping frequencies problem. By revealing critical timing and frequency patterns unique to PD and ET tremors, this proposed XAI-BiLSTM model enables clinicians to make more informed classifications, potentially reducing misclassification rates and improving treatment outcomes.

Can we use machine learning to improve the interpretation and application of urodynamic data?: ICI-RS 2023

INTRODUCTION

A "Think Tank" at the International Consultation on Incontinence-Research Society meeting held in Bristol, United Kingdom in June 2023 considered the progress and promise of machine learning (ML) applied to urodynamic data.

METHODS

Examples of the use of ML applied to data from uroflowmetry, pressure flow studies and imaging were presented. The advantages and limitations of ML were considered. Recommendations made during the subsequent debate for research studies were recorded.

RESULTS

ML analysis holds great promise for the kind of data generated in urodynamic studies. To date, ML techniques have not yet achieved sufficient accuracy for routine diagnostic application. Potential approaches that can improve the use of ML were agreed and research questions were proposed.

CONCLUSIONS

ML is well suited to the analysis of urodynamic data, but results to date have not achieved clinical utility. It is considered likely that further research can improve the analysis of the large, multifactorial data sets generated by urodynamic clinics, and improve to some extent data pattern recognition that is currently subject to observer error and artefactual noise.

A novel hyperparameter search approach for accuracy and simplicity in disease prediction risk scoring

OBJECTIVE

Develop a novel technique to identify an optimal number of regression units corresponding to a single risk point, while creating risk scoring systems from logistic regression-based disease predictive models. The optimal value of this hyperparameter balances simplicity and accuracy, yielding risk scores of small scale and high accuracy for patient risk stratification.

MATERIALS AND METHODS

The proposed technique applies an adapted line search across all potential hyperparameter values. Additionally, DeLong test is integrated to ensure the selected value produces an accuracy insignificantly different from the best achievable risk score accuracy. We assessed the approach through two case studies predicting diabetic retinopathy (DR) within six months and hip fracture readmissions (HFR) within 30 days, involving cohorts of 90 400 diabetic patients and 18 065 hip fracture patients.

RESULTS

Our scores achieve accuracies insignificantly different from those obtained by existing approaches, reaching AUROCs of 0.803 and 0.645 for DR and HFR predictions, respectively. Regarding the scale, our scores ranged 0-53 for DR and 0-15 for HFR, while scores produced by existing methods frequently spanned hundreds or thousands.

DISCUSSION

According to the assessment, our risk scores offer simple and accurate predictions for diseases. Furthermore, our new DR score provides a competitive alternative to state-of-the-art risk scores for DR, while our HFR case study presents the first risk score for this condition.

CONCLUSION

Our technique offers a generalizable framework for crafting precise risk scores of compact scales, addressing the demand for user-friendly and effective risk stratification tool in healthcare.

The value of artificial intelligence for the treatment of mechanically ventilated intensive care unit patients: An early health technology assessment

PURPOSE

The health and economic consequences of artificial intelligence (AI) systems for mechanically ventilated intensive care unit patients often remain unstudied. Early health technology assessments (HTA) can examine the potential impact of AI systems by using available data and simulations. Therefore, we developed a generic health-economic model suitable for early HTA of AI systems for mechanically ventilated patients.

MATERIALS AND METHODS

Our generic health-economic model simulates mechanically ventilated patients from their hospitalisation until their death. The model simulates two scenarios, care as usual and care with the AI system, and compares these scenarios to estimate their cost-effectiveness.

RESULTS

The generic health-economic model we developed is suitable for estimating the cost-effectiveness of various AI systems. By varying input parameters and assumptions, the model can examine the cost-effectiveness of AI systems across a wide range of different clinical settings.

CONCLUSIONS

Using the proposed generic health-economic model, investors and innovators can easily assess whether implementing a certain AI system is likely to be cost-effective before an exact clinical impact is determined. The results of the early HTA can aid investors and innovators in deployment of AI systems by supporting development decisions, informing value-based pricing, clinical trial design, and selection of target patient groups.

Evolving role of the health system CIO: perspectives from 33 health systems

OBJECTIVES

This study explores the evolving position of the health system chief information officer (CIO) by identifying new core roles for success.

METHODS

An advisory board of industry executives and system leaders guided the study. Purposeful sampling was used to invite chief executive officer and CIOs from 65 not-for-profit US health systems to participate. Interviews were conducted with 51 executives from 33 different systems, using a comprehensive interview topic guide. Interview transcripts were analysed using NVivo software, focusing on themes related to the evolving role of the health system CIO.

RESULTS

Analyses revealed three main themes, with the CIO as (1) enabler of strategic change and transformation, (2) strategic developer of technology and leadership talent and (3) driver of organisational culture.

DISCUSSION

The role of CIO has undergone transformation from technology and information system management to strategic leadership within the broader health system context. It highlights the importance of comprehensive business knowledge for CIOs and the need for other C-suite executives to have a deeper understanding of information and technology.

CONCLUSION

As healthcare continues to evolve, the role of the CIO is expected to expand further, requiring a blend of technical and strategic business skills. This evolution presents opportunities for health systems to enhance their leadership development programmes, preparing leaders for the complexities of the contemporary health system sector.

A new model for determining risk of male infertility from serum hormone levels, without semen analysis

We investigated a screening method using only serum hormone levels and AI (artificial intelligence) predictive analysis. Among 3662 patients, numbers for NOA (non-obstructive azoospermia), OA (obstructive azoospermia), cryptozoospermia, oligozoospermia and/or asthenozoospermia, normal, and ejaculation disorder were 448, 210, 46, 1619, 1333, and 6, respectively. "Normal" was defined as semen findings normal according to the WHO (World Health Organization) Manual for Human Semen Testing of 2021. We extracted age, LH (luteinizing hormone), FSH (follicle stimulating hormone), PRL (prolactin), testosterone, E2 (estradiol), and T (testosterone)/E2 from medical records. A total motility sperm count of 9.408 × 106 (1.4 ml × 16 × 106/ml × 42%) was defined as the lower limit of normal. The Prediction One-based AI model had an AUC (area under the curve) of 74.42%. For the AutoML Tables-based model, AUC ROC (receiver operating characteristic) was 74.2% and AUC PR (precision-recall) 77.2%. In a ranking of feature importance from 1st to 3rd, FSH came a clear 1st. T/E2 and LH ranked 2nd and 3rd for both Prediction One and AutoML Tables. Using data from 2021 and 2022 to verify the Prediction One-based AI model, the predicted and actual results for NOA were 100% matched in both years.

Evaluation of ChatGPT as a diagnostic tool for medical learners and clinicians

BACKGROUND

ChatGPT is a large language model (LLM) trained on over 400 billion words from books, articles, and websites. Its extensive training draws from a large database of information, making it valuable as a diagnostic aid. Moreover, its capacity to comprehend and generate human language allows medical trainees to interact with it, enhancing its appeal as an educational resource. This study aims to investigate ChatGPT's diagnostic accuracy and utility in medical education.

METHODS

150 Medscape case challenges (September 2021 to January 2023) were inputted into ChatGPT. The primary outcome was the number (%) of cases for which the answer given was correct. Secondary outcomes included diagnostic accuracy, cognitive load, and quality of medical information. A qualitative content analysis was also conducted to assess its responses.

RESULTS

ChatGPT answered 49% (74/150) cases correctly. It had an overall accuracy of 74%, a precision of 48.67%, sensitivity of 48.67%, specificity of 82.89%, and an AUC of 0.66. Most answers were considered low cognitive load 51% (77/150) and most answers were complete and relevant 52% (78/150).

DISCUSSION

ChatGPT in its current form is not accurate as a diagnostic tool. ChatGPT does not necessarily give factual correctness, despite the vast amount of information it was trained on. Based on our qualitative analysis, ChatGPT struggles with the interpretation of laboratory values, imaging results, and may overlook key information relevant to the diagnosis. However, it still offers utility as an educational tool. ChatGPT was generally correct in ruling out a specific differential diagnosis and providing reasonable next diagnostic steps. Additionally, answers were easy to understand, showcasing a potential benefit in simplifying complex concepts for medical learners. Our results should guide future research into harnessing ChatGPT's potential educational benefits, such as simplifying medical concepts and offering guidance on differential diagnoses and next steps.

Artificial intelligence applications in dentistry: A bibliometric review with an emphasis on computational research trends within the field

BACKGROUND

The aim of this study was to understand the trends regarding the use of artificial intelligence in dentistry through a bibliometric review.

TYPES OF STUDIES REVIEWED

The authors performed a literature search on Web of Science. They collected the following data: articles-number and density of citations, year, key words, language, document type, study design, and theme (main objective, diagnostic method, and specialties); journals-impact factor; authors-country, continent, and institution. The authors used Visualization of Similarities Viewer software (Leiden University) to analyze the data and Spearman test for correlation analysis.

RESULTS

After selection, 1,478 articles were included. The number of citations ranged from 0 through 327. The articles were published from 1984 through 2024. Most articles were characterized as proof of concept (979). Definition and classification of structures and diseases was the most common theme (550 articles). There was an emphasis on radiology (333 articles) and radiographic-based diagnostic methods (715 articles). China was the country with the most articles (251), and Asia was the continent with the most articles (871). The Charité-University of Medicine Berlin was the institution with the most articles (42), and the author with the most articles was Schwendicke (53).

PRACTICAL IMPLICATIONS

Artificial intelligence is an important clinical tool to facilitate diagnosis and provide automation in various processes.

Assessment of Sports Concussion in Female Athletes: A Role for Neuroinformatics?

Over the past decade, the intricacies of sports-related concussions among female athletes have become readily apparent. Traditional clinical methods for diagnosing concussions suffer limitations when applied to female athletes, often failing to capture subtle changes in brain structure and function. Advanced neuroinformatics techniques and machine learning models have become invaluable assets in this endeavor. While these technologies have been extensively employed in understanding concussion in male athletes, there remains a significant gap in our comprehension of their effectiveness for female athletes. With its remarkable data analysis capacity, machine learning offers a promising avenue to bridge this deficit. By harnessing the power of machine learning, researchers can link observed phenotypic neuroimaging data to sex-specific biological mechanisms, unraveling the mysteries of concussions in female athletes. Furthermore, embedding methods within machine learning enable examining brain architecture and its alterations beyond the conventional anatomical reference frame. In turn, allows researchers to gain deeper insights into the dynamics of concussions, treatment responses, and recovery processes. This paper endeavors to address the crucial issue of sex differences in multimodal neuroimaging experimental design and machine learning approaches within female athlete populations, ultimately ensuring that they receive the tailored care they require when facing the challenges of concussions. Through better data integration, feature identification, knowledge representation, validation, etc., neuroinformaticists, are ideally suited to bring clarity, context, and explainabilty to the study of sports-related head injuries in males and in females, and helping to define recovery.

Identifying Predictors of Heart Failure Readmission in Patients From a Statutory Health Insurance Database: Retrospective Machine Learning Study

BACKGROUND

Patients with heart failure (HF) are the most commonly readmitted group of adult patients in Germany. Most patients with HF are readmitted for noncardiovascular reasons. Understanding the relevance of HF management outside the hospital setting is critical to understanding HF and factors that lead to readmission. Application of machine learning (ML) on data from statutory health insurance (SHI) allows the evaluation of large longitudinal data sets representative of the general population to support clinical decision-making.

OBJECTIVE

This study aims to evaluate the ability of ML methods to predict 1-year all-cause and HF-specific readmission after initial HF-related admission of patients with HF in outpatient SHI data and identify important predictors.

METHODS

We identified individuals with HF using outpatient data from 2012 to 2018 from the AOK Baden-Württemberg SHI in Germany. We then trained and applied regression and ML algorithms to predict the first all-cause and HF-specific readmission in the year after the first admission for HF. We fitted a random forest, an elastic net, a stepwise regression, and a logistic regression to predict readmission by using diagnosis codes, drug exposures, demographics (age, sex, nationality, and type of coverage within SHI), degree of rurality for residence, and participation in disease management programs for common chronic conditions (diabetes mellitus type 1 and 2, breast cancer, chronic obstructive pulmonary disease, and coronary heart disease). We then evaluated the predictors of HF readmission according to their importance and direction to predict readmission.

RESULTS

Our final data set consisted of 97,529 individuals with HF, and 78,044 (80%) were readmitted within the observation period. Of the tested modeling approaches, the random forest approach best predicted 1-year all-cause and HF-specific readmission with a C-statistic of 0.68 and 0.69, respectively. Important predictors for 1-year all-cause readmission included prescription of pantoprazole, chronic obstructive pulmonary disease, atherosclerosis, sex, rurality, and participation in disease management programs for type 2 diabetes mellitus and coronary heart disease. Relevant features for HF-specific readmission included a large number of canonical HF comorbidities.

CONCLUSIONS

While many of the predictors we identified were known to be relevant comorbidities for HF, we also uncovered several novel associations. Disease management programs have widely been shown to be effective at managing chronic disease; however, our results indicate that in the short term they may be useful for targeting patients with HF with comorbidity at increased risk of readmission. Our results also show that living in a more rural location increases the risk of readmission. Overall, factors beyond comorbid disease were relevant for risk of HF readmission. This finding may impact how outpatient physicians identify and monitor patients at risk of HF readmission.

Predicting Successful Chronic Total Occlusion Crossing With Primary Antegrade Wiring Using Machine Learning

BACKGROUND

There is limited data on predicting successful chronic total occlusion crossing using primary antegrade wiring (AW).

OBJECTIVES

The aim of this study was to develop and validate a machine learning (ML) prognostic model for successful chronic total occlusion crossing using primary AW.

METHODS

We used data from 12,136 primary AW cases performed between 2012 and 2023 at 48 centers in the PROGRESS CTO registry (Prospective Global Registry for the Study of Chronic Total Occlusion Intervention; NCT02061436) to develop 5 ML models. Hyperparameter tuning was performed for the model with the best performance, and the SHAP (SHapley Additive exPlanations) explainer was implemented to estimate feature importance.

RESULTS

Primary AW was successful in 6,965 cases (57.4%). Extreme gradient boosting was the best performing ML model with an average area under the receiver-operating characteristic curve of 0.775 (± 0.010). After hyperparameter tuning, the average area under the receiver-operating characteristic curve of the extreme gradient boosting model was 0.782 in the training set and 0.780 in the testing set. Among the factors examined, occlusion length had the most significant impact on predicting successful primary AW crossing followed by blunt/no stump, presence of interventional collaterals, vessel diameter, and proximal cap ambiguity. In contrast, aorto-ostial lesion location had the least impact on the outcome. A web-based application for predicting successful primary AW wiring crossing is available online (PROGRESS-CTO website) (https://www.progresscto.org/predict-aw-success).

CONCLUSIONS

We developed an ML model with 14 features and high predictive capacity for successful primary AW in chronic total occlusion percutaneous coronary intervention.

Application of artificial intelligence in cancer diagnosis and tumor nanomedicine

Cancer is a major health concern due to its high incidence and mortality rates. Advances in cancer research, particularly in artificial intelligence (AI) and deep learning, have shown significant progress. The swift evolution of AI in healthcare, especially in tools like computer-aided diagnosis, has the potential to revolutionize early cancer detection. This technology offers improved speed, accuracy, and sensitivity, bringing a transformative impact on cancer diagnosis, treatment, and management. This paper provides a concise overview of the application of artificial intelligence in the realms of medicine and nanomedicine, with a specific emphasis on the significance and challenges associated with cancer diagnosis. It explores the pivotal role of AI in cancer diagnosis, leveraging structured, unstructured, and multimodal fusion data. Additionally, the article delves into the applications of AI in nanomedicine sensors and nano-oncology drugs. The fundamentals of deep learning and convolutional neural networks are clarified, underscoring their relevance to AI-driven cancer diagnosis. A comparative analysis is presented, highlighting the accuracy and efficiency of traditional methods juxtaposed with AI-based approaches. The discussion not only assesses the current state of AI in cancer diagnosis but also delves into the challenges faced by AI in this context. Furthermore, the article envisions the future development direction and potential application of artificial intelligence in cancer diagnosis, offering a hopeful prospect for enhanced cancer detection and improved patient prognosis.

Study of the Berezinskii-Kosterlitz-Thouless transition: an unsupervised machine learning approach

The Berezinskii-Kosterlitz-Thouless (BKT) transition in magnetic systems is an intriguing phenomenon, and estimating the BKT transition temperature is a long-standing problem. In this work, we explore anisotropic classical Heisenberg XY and XXZ models with ferromagnetic exchange on a square lattice and antiferromagnetic exchange on a triangular lattice using an unsupervised machine learning approach called principal component analysis (PCA). The earlier PCA studies of the BKT transition temperature (TBKT) using the vorticities as input fail to give any conclusive results, whereas, in this work, we show that the proper analysis of the first principal component-temperature curve can estimateTBKTwhich is consistent with the existing literature. This analysis works well for the anisotropic classical Heisenberg with a ferromagnetic exchange on a square lattice and for frustrated antiferromagnetic exchange on a triangular lattice. The classical anisotropic Heisenberg antiferromagnetic model on the triangular lattice has two close transitions: theTBKTand Ising-like phase transition for chirality atTc, and it is difficult to separate these transition points. It is also noted that using the PCA method and manipulation of their first principal component not only makes the separation of transition points possible but also determines transition temperature.

Recommendations to promote fairness and inclusion in biomedical AI research and clinical use

OBJECTIVE

Understanding and quantifying biases when designing and implementing actionable approaches to increase fairness and inclusion is critical for artificial intelligence (AI) in biomedical applications.

METHODS

In this Special Communication, we discuss how bias is introduced at different stages of the development and use of AI applications in biomedical sciences and health care. We describe various AI applications and their implications for fairness and inclusion in sections on 1) Bias in Data Source Landscapes, 2) Algorithmic Fairness, 3) Uncertainty in AI Predictions, 4) Explainable AI for Fairness and Equity, and 5) Sociological/Ethnographic Issues in Data and Results Representation.

RESULTS

We provide recommendations to address biases when developing and using AI in clinical applications.

CONCLUSION

These recommendations can be applied to informatics research and practice to foster more equitable and inclusive health care systems and research discoveries.

Machine learning-based model to predict severe acute kidney injury after total aortic arch replacement for acute type A aortic dissection

Background

Severe acute kidney injury (AKI) after total aortic arch replacement (TAAR) is related to adverse outcomes in patients with acute type A aortic dissection (ATAAD). However, the early prediction of severe AKI remains a challenge. This study aimed to develop a novel model to predict severe AKI after TAAR in ATAAD patients using machine learning algorithms.

Methods

A total of 572 ATAAD patients undergoing TAAR were enrolled in this retrospective study, and randomly divided into a training set (70 %) and a validation set (30 %). Lasso regression, support vector machine-recursive feature elimination and random forest algorithms were used to screen indicators for severe AKI (defined as AKI stage III) in the training set, respectively. Then the intersection indicators were selected to construct models through artificial neural network (ANN) and logistic regression. The AUC-ROC curve was employed to ascertain the prediction efficacy of the ANN and logistic regression models.

Results

The incidence of severe AKI after TAAR was 22.9 % among ATAAD patients. The intersection predictors identified by different machine learning algorithms were baseline serum creatinine and ICU admission variables, including serum cystatin C, procalcitonin, aspartate transaminase, platelet, lactic dehydrogenase, urine N-acetyl-β-d-glucosidase and Acute Physiology and Chronic Health Evaluation II score. The ANN model showed a higher AUC-ROC than logistic regression (0.938 vs 0.908, p < 0.05). Furthermore, the ANN model could predict 89.1 % of severe AKI cases beforehand. In the validation set, the superior performance of the ANN model was further confirmed in terms of discrimination ability (AUC = 0.916), calibration curve analysis and decision curve analysis.

Conclusion

This study developed a novel and reliable clinical prediction model for severe AKI after TAAR in ATAAD patients using machine learning algorithms. Importantly, the ANN model showed a higher predictive ability for severe AKI than logistic regression.

Evaluating gender bias in ML-based clinical risk prediction models: A study on multiple use cases at different hospitals

BACKGROUND

An inherent difference exists between male and female bodies, the historical under-representation of females in clinical trials widened this gap in existing healthcare data. The fairness of clinical decision-support tools is at risk when developed based on biased data. This paper aims to quantitatively assess the gender bias in risk prediction models. We aim to generalize our findings by performing this investigation on multiple use cases at different hospitals.

METHODS

First, we conduct a thorough analysis of the source data to find gender-based disparities. Secondly, we assess the model performance on different gender groups at different hospitals and on different use cases. Performance evaluation is quantified using the area under the receiver-operating characteristic curve (AUROC). Lastly, we investigate the clinical implications of these biases by analyzing the underdiagnosis and overdiagnosis rate, and the decision curve analysis (DCA). We also investigate the influence of model calibration on mitigating gender-related disparities in decision-making processes.

RESULTS

Our data analysis reveals notable variations in incidence rates, AUROC, and over-diagnosis rates across different genders, hospitals and clinical use cases. However, it is also observed the underdiagnosis rate is consistently higher in the female population. In general, the female population exhibits lower incidence rates and the models perform worse when applied to this group. Furthermore, the decision curve analysis demonstrates there is no statistically significant difference between the model's clinical utility across gender groups within the interested range of thresholds.

CONCLUSION

The presence of gender bias within risk prediction models varies across different clinical use cases and healthcare institutions. Although inherent difference is observed between male and female populations at the data source level, this variance does not affect the parity of clinical utility. In conclusion, the evaluations conducted in this study highlight the significance of continuous monitoring of gender-based disparities in various perspectives for clinical risk prediction models.

The impact of machine learning on the prediction of diabetic foot ulcers - A systematic review

INTRODUCTION

Globally, diabetes mellitus poses a significant health challenge as well as the associated complications of diabetes, such as diabetic foot ulcers (DFUs). The early detection of DFUs is important in the healing process and machine learning may be able to help inform clinical staff during the treatment process.

METHODS

A PRISMA-informed search of the literature was completed via the Cochrane Library and MEDLINE (OVID), EMBASE, CINAHL Plus and Scopus databases for reports published in English and in the last ten years. The primary outcome of interest was the impact of machine learning on the prediction of DFUs. The secondary outcome was the statistical performance measures reported. Data were extracted using a predesigned data extraction tool. Quality appraisal was undertaken using the evidence-based librarianship critical appraisal tool.

RESULTS

A total of 18 reports met the inclusion criteria. Nine reports proposed models to identify two classes, either healthy skin or a DFU. Nine reports proposed models to predict the progress of DFUs, for example, classing infection versus non-infection, or using wound characteristics to predict healing. A variety of machine learning techniques were proposed. Where reported, sensitivity = 74.53-98 %, accuracy = 64.6-99.32 %, precision = 62.9-99 %, and the F-measure = 52.05-99.0 %.

CONCLUSIONS

A variety of machine learning models were suggested to successfully classify DFUs from healthy skin, or to inform the prediction of DFUs. The proposed machine learning models may have the potential to inform the clinical practice of managing DFUs and may help to improve outcomes for individuals with DFUs. Future research may benefit from the development of a standard device and algorithm that detects, diagnoses and predicts the progress of DFUs.

Development and validation of machine learning models to predict MDRO colonization or infection on ICU admission by using electronic health record data

BACKGROUND

Multidrug-resistant organisms (MDRO) pose a significant threat to public health. Intensive Care Units (ICU), characterized by the extensive use of antimicrobial agents and a high prevalence of bacterial resistance, are hotspots for MDRO proliferation. Timely identification of patients at high risk for MDRO can aid in curbing transmission, enhancing patient outcomes, and maintaining the cleanliness of the ICU environment. This study focused on developing a machine learning (ML) model to identify patients at risk of MDRO during the initial phase of their ICU stay.

METHODS

Utilizing patient data from the First Medical Center of the People's Liberation Army General Hospital (PLAGH-ICU) and the Medical Information Mart for Intensive Care (MIMIC-IV), the study analyzed variables within 24 h of ICU admission. Machine learning algorithms were applied to these datasets, emphasizing the early detection of MDRO colonization or infection. Model efficacy was evaluated by the area under the receiver operating characteristics curve (AUROC), alongside internal and external validation sets.

RESULTS

The study evaluated 3,536 patients in PLAGH-ICU and 34,923 in MIMIC-IV, revealing MDRO prevalence of 11.96% and 8.81%, respectively. Significant differences in ICU and hospital stays, along with mortality rates, were observed between MDRO positive and negative patients. In the temporal validation, the PLAGH-ICU model achieved an AUROC of 0.786 [0.748, 0.825], while the MIMIC-IV model reached 0.744 [0.723, 0.766]. External validation demonstrated reduced model performance across different datasets. Key predictors included biochemical markers and the duration of pre-ICU hospital stay.

CONCLUSIONS

The ML models developed in this study demonstrated their capability in early identification of MDRO risks in ICU patients. Continuous refinement and validation in varied clinical contexts remain essential for future applications.

Features of Importance in Nasal Endoscopy: Deriving a Meaningful Framework

OBJECTIVE

Critical components of the nasal endoscopic examination have not been definitively established for either the normal examination or for clinical disorders. This study aimed to identify concordance among rhinologists regarding the importance of examination findings for various nasal pathologies.

STUDY DESIGN

A consortium of 19 expert rhinologists across the United States was asked to rank the importance of findings on nasal endoscopy for 5 different sinonasal symptom presentations.

SETTING

An online questionnaire was distributed in July 2023.

METHODS

The questionnaire utilized JotForm® software and featured 5 cases with a set of 4 identical questions per case, each covering a common indication for nasal endoscopy. Rankings were synthesized into Normalized Attention Scores (NASs) and Weighted Normalized Attention Scores (W-NASs) to represent the perceived importance of each feature, scaled from 0 to 1.

RESULTS

General concordance was found for examination findings on nasal endoscopy within each case. The perceived features of importance differed between cases based on clinical presentation. For instance, in evaluating postnasal drip, the middle meatus was selected as the most important structure to examine (NAS, 0.73), with mucus selected as the most important abnormal finding (W-NAS, 0.66). The primary feature of interest for mucus was whether it was purulent or not (W-NAS, 0.67). Similar analyses were performed for features in each case.

CONCLUSION

The implicit framework existing among rhinologists may help standardize examinations and improve diagnostic accuracy, augment the instruction of trainees, and inform the development of artificially intelligent algorithms to enhance clinical decision-making during nasal endoscopy.

Development of a machine learning-based model to predict major adverse events after surgery for type A aortic dissection complicated by malnutrition

Objective

This study aims to develop a predictive model for the risk of major adverse events (MAEs) in type A aortic dissection (AAAD) patients with malnutrition after surgery, utilizing machine learning (ML) algorithms.

Methods

We retrospectively collected clinical data from AAAD patients with malnutrition who underwent surgical treatment at our center. Through least absolute shrinkage and selection operator (LASSO) regression analysis, we screened for preoperative and intraoperative characteristic variables. Based on the random forest (RF) algorithm, we constructed a ML predictive model, and further evaluated and interpreted this model.

Results

Through LASSO regression analysis and univariate analysis, we ultimately selected seven feature variables for modeling. After comparing six different ML models, we confirmed that the RF model demonstrated the best predictive performance in this dataset. Subsequently, we constructed a model using the RF algorithm to predict the risk of postoperative MAEs in AAAD patients with malnutrition. The test set results indicated that this model has excellent predictive efficacy and clinical applicability. Finally, we employed the Shapley additive explanations (SHAP) method to further interpret the predictions of this model.

Conclusion

We have successfully constructed a risk prediction model for postoperative MAEs in AAAD patients with malnutrition using the RF algorithm, and we have interpreted the model through the SHAP method. This model aids clinicians in early identification of high-risk patients for MAEs, thereby potentially mitigating adverse clinical outcomes associated with malnutrition.

Review of Machine Learning Techniques in Soft Tissue Biomechanics and Biomaterials

BACKGROUND AND OBJECTIVE

Advanced material models and material characterization of soft biological tissues play an essential role in pre-surgical planning for vascular surgeries and transcatheter interventions. Recent advances in heart valve engineering, medical device and patch design are built upon these models. Furthermore, understanding vascular growth and remodeling in native and tissue-engineered vascular biomaterials, as well as designing and testing drugs on soft tissue, are crucial aspects of predictive regenerative medicine. Traditional nonlinear optimization methods and finite element (FE) simulations have served as biomaterial characterization tools combined with soft tissue mechanics and tensile testing for decades. However, results obtained through nonlinear optimization methods are reliable only to a certain extent due to mathematical limitations, and FE simulations may require substantial computing time and resources, which might not be justified for patient-specific simulations. To a significant extent, machine learning (ML) techniques have gained increasing prominence in the field of soft tissue mechanics in recent years, offering notable advantages over conventional methods. This review article presents an in-depth examination of emerging ML algorithms utilized for estimating the mechanical characteristics of soft biological tissues and biomaterials. These algorithms are employed to analyze crucial properties such as stress-strain curves and pressure-volume loops. The focus of the review is on applications in cardiovascular engineering, and the fundamental mathematical basis of each approach is also discussed.

METHODS

The review effort employed two strategies. First, the recent studies of major research groups actively engaged in cardiovascular soft tissue mechanics are compiled, and research papers utilizing ML and deep learning (DL) techniques were included in our review. The second strategy involved a standard keyword search across major databases. This approach provided 11 relevant ML articles, meticulously selected from reputable sources including ScienceDirect, Springer, PubMed, and Google Scholar. The selection process involved using specific keywords such as "machine learning" or "deep learning" in conjunction with "soft biological tissues", "cardiovascular", "patient-specific," "strain energy", "vascular" or "biomaterials". Initially, a total of 25 articles were selected. However, 14 of these articles were excluded as they did not align with the criteria of focusing on biomaterials specifically employed for soft tissue repair and regeneration. As a result, the remaining 11 articles were categorized based on the ML techniques employed and the training data utilized.

RESULTS

ML techniques utilized for assessing the mechanical characteristics of soft biological tissues and biomaterials are broadly classified into two categories: standard ML algorithms and physics-informed ML algorithms. The standard ML models are then organized based on their tasks, being grouped into Regression and Classification subcategories. Within these categories, studies employ various supervised learning models, including support vector machines (SVMs), bagged decision trees (BDTs), artificial neural networks (ANNs) or deep neural networks (DNNs), and convolutional neural networks (CNNs). Additionally, the utilization of unsupervised learning approaches, such as autoencoders incorporating principal component analysis (PCA) and/or low-rank approximation (LRA), is based on the specific characteristics of the training data. The training data predominantly consists of three types: experimental mechanical data, including uniaxial or biaxial stress-strain data; synthetic mechanical data generated through non-linear fitting and/or FE simulations; and image data such as 3D second harmonic generation (SHG) images or computed tomography (CT) images. The evaluation of performance for physics-informed ML models primarily relies on the coefficient of determinationR 2 . In contrast, various metrics and error measures are utilized to assess the performance of standard ML models. Furthermore, our review includes an extensive examination of prevalent biomaterial models that can serve as physical laws for physics-informed ML models.

CONCLUSION

ML models offer an accurate, fast, and reliable approach for evaluating the mechanical characteristics of diseased soft tissue segments and selecting optimal biomaterials for time-critical soft tissue surgeries. Among the various ML models examined in this review, physics-informed neural network models exhibit the capability to forecast the mechanical response of soft biological tissues accurately, even with limited training samples. These models achieve highR 2 values ranging from 0.90 to 1.00. This is particularly significant considering the challenges associated with obtaining a large number of living tissue samples for experimental purposes, which can be time-consuming and impractical. Additionally, the review not only discusses the advantages identified in the current literature but also sheds light on the limitations and offers insights into future perspectives.

Bridging Health Disparities in the Data-Driven World of Artificial Intelligence: A Narrative Review

BACKGROUND

Artificial intelligence (AI) holds exciting potential to revolutionize healthcare delivery in the United States. However, there are concerns about its potential to perpetuate disparities among historically marginalized populations.

OBJECTIVE

Following the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses, we conducted a narrative review of current literature on AI and health disparities in the United States. We aimed to answer the question, Does AI have the potential to reduce or eliminate health disparities, or will its use further exacerbate these disparities?

METHODS

We searched the Ovid MEDLINE electronic database to identify and retrieve publications discussing AI and its impact on racial/ethnic health disparities. Articles were included if they discussed AI as a tool to mitigate racial health disparities with or without bias in developing and using AI.

RESULTS

This review included 65 articles. We identified six themes of limitations in AI that impact health equity: (1) biases in AI can perpetuate and exacerbate racial and ethnic inequities; (2) equity in algorithms should be a priority; (3) lack of diversity in the field of AI is concerning; (4) the need for regulation and testing algorithms for accuracy; (5) ethical standards for AI in health care are needed; and (6) the importance of promoting transparency and accountability.

CONCLUSIONS

While AI promises to enhance healthcare outcomes and address equity concerns, risks and challenges are associated with its implementation. To maximize the use of AI, it must be approached with an equity lens during all phases of development.

Development and evaluation of a deep learning framework for the diagnosis of malnutrition using a 3D facial points cloud: A cross-sectional study

BACKGROUND

The feasibility of diagnosing malnutrition using facial features has been validated. A tool to integrate all facial features associated with malnutrition for disease screening is still demanded. This work aims to develop and evaluate a deep learning (DL) framework to accurately determine malnutrition based on a 3D facial points cloud.

METHODS

A group of 482 patients were studied in this perspective work. The 3D facial data were obtained using a 3D camera and represented as a 3D facial points cloud. A DL model, PointNet++, was trained and evaluated using the points cloud as inputs and classified the malnutrition states. The performance was evaluated with the area under the receiver operating characteristic curve, accuracy, specificity, sensitivity, and F1 score.

RESULTS

Among the 482 patients, 150 patients (31.1%) were diagnosed as having moderate malnutrition and 54 patients (11.2%) as having severe malnutrition. The DL model achieved the performance with an area under the receiver operating characteristic curve of 0.7240 ± 0.0416.

CONCLUSION

The DL model achieved encouraging performance in accurately classifying nutrition states based on a points cloud of 3D facial information of patients with malnutrition.

Advancing equity in breast cancer care: natural language processing for analysing treatment outcomes in under-represented populations

OBJECTIVE

The study aimed to develop natural language processing (NLP) algorithms to automate extracting patient-centred breast cancer treatment outcomes from clinical notes in electronic health records (EHRs), particularly for women from under-represented populations.

METHODS

The study used clinical notes from 2010 to 2021 from a tertiary hospital in the USA. The notes were processed through various NLP techniques, including vectorisation methods (term frequency-inverse document frequency (TF-IDF), Word2Vec, Doc2Vec) and classification models (support vector classification, K-nearest neighbours (KNN), random forest (RF)). Feature selection and optimisation through random search and fivefold cross-validation were also conducted.

RESULTS

The study annotated 100 out of 1000 clinical notes, using 970 notes to build the text corpus. TF-IDF and Doc2Vec combined with RF showed the highest performance, while Word2Vec was less effective. RF classifier demonstrated the best performance, although with lower recall rates, suggesting more false negatives. KNN showed lower recall due to its sensitivity to data noise.

DISCUSSION

The study highlights the significance of using NLP in analysing clinical notes to understand breast cancer treatment outcomes in under-represented populations. The TF-IDF and Doc2Vec models were more effective in capturing relevant information than Word2Vec. The study observed lower recall rates in RF models, attributed to the dataset's imbalanced nature and the complexity of clinical notes.

CONCLUSION

The study developed high-performing NLP pipeline to capture treatment outcomes for breast cancer in under-represented populations, demonstrating the importance of document-level vectorisation and ensemble methods in clinical notes analysis. The findings provide insights for more equitable healthcare strategies and show the potential for broader NLP applications in clinical settings.

Using natural language processing in emergency medicine health service research: A systematic review and meta-analysis

OBJECTIVES

Natural language processing (NLP) represents one of the adjunct technologies within artificial intelligence and machine learning, creating structure out of unstructured data. This study aims to assess the performance of employing NLP to identify and categorize unstructured data within the emergency medicine (EM) setting.

METHODS

We systematically searched publications related to EM research and NLP across databases including MEDLINE, Embase, Scopus, CENTRAL, and ProQuest Dissertations & Theses Global. Independent reviewers screened, reviewed, and evaluated article quality and bias. NLP usage was categorized into syndromic surveillance, radiologic interpretation, and identification of specific diseases/events/syndromes, with respective sensitivity analysis reported. Performance metrics for NLP usage were calculated and the overall area under the summary of receiver operating characteristic curve (SROC) was determined.

RESULTS

A total of 27 studies underwent meta-analysis. Findings indicated an overall mean sensitivity (recall) of 82%-87%, specificity of 95%, with the area under the SROC at 0.96 (95% CI 0.94-0.98). Optimal performance using NLP was observed in radiologic interpretation, demonstrating an overall mean sensitivity of 93% and specificity of 96%.

CONCLUSIONS

Our analysis revealed a generally favorable performance accuracy in using NLP within EM research, particularly in the realm of radiologic interpretation. Consequently, we advocate for the adoption of NLP-based research to augment EM health care management.

Accuracy of machine learning in the diagnosis of odontogenic cysts and tumors: a systematic review and meta-analysis

BACKGROUND

The recent impact of artificial intelligence in diagnostic services has been enormous. Machine learning tools offer an innovative alternative to diagnose cysts and tumors radiographically that pose certain challenges due to the near similar presentation, anatomical variations, and superimposition. It is crucial that the performance of these models is evaluated for their clinical applicability in diagnosing cysts and tumors.

METHODS

A comprehensive literature search was carried out on eminent databases for published studies between January 2015 and December 2022. Studies utilizing machine learning models in the diagnosis of odontogenic cysts or tumors using Orthopantomograms (OPG) or Cone Beam Computed Tomographic images (CBCT) were included. QUADAS-2 tool was used for the assessment of the risk of bias and applicability concerns. Meta-analysis was performed for studies reporting sufficient performance metrics, separately for OPG and CBCT.

RESULTS

16 studies were included for qualitative synthesis including a total of 10,872 odontogenic cysts and tumors. The sensitivity and specificity of machine learning in diagnosing cysts and tumors through OPG were 0.83 (95% CI 0.81-0.85) and 0.82 (95% CI 0.81-0.83) respectively. Studies utilizing CBCT noted a sensitivity of 0.88 (95% CI 0.87-0.88) and specificity of 0.88 (95% CI 0.87-0.89). Highest classification accuracy was 100%, noted for Support Vector Machine classifier.

CONCLUSION

The results from the present review favoured machine learning models to be used as a clinical adjunct in the radiographic diagnosis of odontogenic cysts and tumors, provided they undergo robust training with a huge dataset. However, the arduous process, investment, and certain ethical concerns associated with the total dependence on technology must be taken into account. Standardized reporting of outcomes for diagnostic studies utilizing machine learning methods is recommended to ensure homogeneity in assessment criteria, facilitate comparison between different studies, and promote transparency in research findings.

When Do Humans Heed AI Agents' Advice? When Should They?

OBJECTIVE

We manipulate the presence, skill, and display of artificial intelligence (AI) recommendations in a strategy game to measure their effect on users' performance.

BACKGROUND

Many applications of AI require humans and AI agents to make decisions collaboratively. Success depends on how appropriately humans rely on the AI agent. We demonstrate an evaluation method for a platform that uses neural network agents of varying skill levels for the simple strategic game of Connect Four.

METHODS

We report results from a 2 × 3 between-subjects factorial experiment that varies the format of AI recommendations (categorical or probabilistic) and the AI agent's amount of training (low, medium, or high). On each round of 10 games, participants proposed a move, saw the AI agent's recommendations, and then moved.

RESULTS

Participants' performance improved with a highly skilled agent, but quickly plateaued, as they relied uncritically on the agent. Participants relied too little on lower skilled agents. The display format had no effect on users' skill or choices.

CONCLUSIONS

The value of these AI agents depended on their skill level and users' ability to extract lessons from their advice.

APPLICATION

Organizations employing AI decision support systems must consider behavioral aspects of the human-agent team. We demonstrate an approach to evaluating competing designs and assessing their performance.

Exome sequence analysis identifies rare coding variants associated with a machine learning-based marker for coronary artery disease

Coronary artery disease (CAD) exists on a spectrum of disease represented by a combination of risk factors and pathogenic processes. An in silico score for CAD built using machine learning and clinical data in electronic health records captures disease progression, severity and underdiagnosis on this spectrum and could enhance genetic discovery efforts for CAD. Here we tested associations of rare and ultrarare coding variants with the in silico score for CAD in the UK Biobank, All of Us Research Program and BioMe Biobank. We identified associations in 17 genes; of these, 14 show at least moderate levels of prior genetic, biological and/or clinical support for CAD. We also observed an excess of ultrarare coding variants in 321 aggregated CAD genes, suggesting more ultrarare variant associations await discovery. These results expand our understanding of the genetic etiology of CAD and illustrate how digital markers can enhance genetic association investigations for complex diseases.

Machine Learning: A Potential Therapeutic Tool to Facilitate Neonatal Therapeutic Decision Making

Bacterial infection is one of the major causes of neonatal morbidity and mortality worldwide. Finding rapid and reliable methods for early recognition and diagnosis of bacterial infections and early individualization of antibacterial drug administration are essential to eradicate these infections and prevent serious complications. However, this is often difficult to perform due to non-specific clinical presentations, low accuracy of current diagnostic methods, and limited knowledge of neonatal pharmacokinetics. Although neonatal medicine has been relatively late to embrace the benefits of machine learning (ML), there have been some initial applications of ML for the early prediction of neonatal sepsis and individualization of antibiotics. This article provides a brief introduction to ML and discusses the current state of the art in diagnosing and treating neonatal bacterial infections, gaps, potential uses of ML, and future directions to address the limitations of current studies. Neonatal bacterial infections involve a combination of physiologic development, disease expression, and treatment response outcomes. To address this complex relationship, future models could consider appropriate ML algorithms to capture time series features while integrating influences from the host, microbes, and drugs to optimize antimicrobial drug use in neonates. All models require prospective clinical trials to validate their clinical utility before clinical use.

Resilient Artificial Intelligence in Health: Synthesis and Research Agenda Toward Next-Generation Trustworthy Clinical Decision Support

Artificial intelligence (AI)-based clinical decision support systems are gaining momentum by relying on a greater volume and variety of secondary use data. However, the uncertainty, variability, and biases in real-world data environments still pose significant challenges to the development of health AI, its routine clinical use, and its regulatory frameworks. Health AI should be resilient against real-world environments throughout its lifecycle, including the training and prediction phases and maintenance during production, and health AI regulations should evolve accordingly. Data quality issues, variability over time or across sites, information uncertainty, human-computer interaction, and fundamental rights assurance are among the most relevant challenges. If health AI is not designed resiliently with regard to these real-world data effects, potentially biased data-driven medical decisions can risk the safety and fundamental rights of millions of people. In this viewpoint, we review the challenges, requirements, and methods for resilient AI in health and provide a research framework to improve the trustworthiness of next-generation AI-based clinical decision support.

The importance of data in Pulmonary Arterial Hypertension: from international registries to Machine Learning

Real-world registries have been critical to building the scientific knowledge of rare diseases, including Pulmonary Arterial Hypertension (PAH). In the past 4 decades, a considerable number of registries on this condition have allowed to improve the pathology and its subgroupś definition, to advance in the understanding of its pathophysiology, to elaborate prognostic scales and to check the transferability of the results from clinical trials to clinical practice. However, in a moment where a huge amount of data from multiple sources is available, they are not always taken into account by the registries. For that reason, Machine Learning (ML) offer a unique opportunity to manage all these data and, finally, to obtain tools that may help to get an earlier diagnose, to help to deduce the prognosis and, in the end, to advance in Personalized Medicine. Thus, we present a narrative revision with the aims of, in one hand, summing up the aspects in which data extraction is important in rare diseases -focusing on the knowledge gained from PAH real-world registries- and, on the other hand, describing some of the achievements and the potential use of the ML techniques on PAH.

From data to diagnosis: skin cancer image datasets for artificial intelligence

Artificial intelligence (AI) solutions for skin cancer diagnosis continue to gain momentum, edging closer towards broad clinical use. These AI models, particularly deep-learning architectures, require large digital image datasets for development. This review provides an overview of the datasets used to develop AI algorithms and highlights the importance of dataset transparency for the evaluation of algorithm generalizability across varying populations and settings. Current challenges for curation of clinically valuable datasets are detailed, which include dataset shifts arising from demographic variations and differences in data collection methodologies, along with inconsistencies in labelling. These shifts can lead to differential algorithm performance, compromise of clinical utility, and the propagation of discriminatory biases when developed algorithms are implemented in mismatched populations. Limited representation of rare skin cancers and minoritized groups in existing datasets are highlighted, which can further skew algorithm performance. Strategies to address these challenges are presented, which include improving transparency, representation and interoperability. Federated learning and generative methods, which may improve dataset size and diversity without compromising privacy, are also examined. Lastly, we discuss model-level techniques that may address biases entrained through the use of datasets derived from routine clinical care. As the role of AI in skin cancer diagnosis becomes more prominent, ensuring the robustness of underlying datasets is increasingly important.

Bilateral effect of acupuncture on cerebrum and cerebellum in ischaemic stroke patients with hemiparesis: a randomised clinical and neuroimaging trial

BACKGROUND

Acupuncture involving the limb region may be effective for stroke rehabilitation clinically, but the visualised and explanatory evidence is limited. Our objectives were to assess the specific effects of acupuncture for ischaemic stroke (IS) patients with hemiparesis and investigate its therapy-driven modification in functional connectivity.

METHODS

IS patients were randomly assigned (2:1) to receive 10 sessions of hand-foot 12 needles acupuncture (HA, n=30) or non-acupoint (NA) acupuncture (n=16), enrolling gender-matched and age-matched healthy controls (HCs, n=34). The clinical outcomes were the improved Fugl-Meyer Assessment scores including upper and lower extremity (ΔFM, ΔFM-UE, ΔFM-LE). The neuroimaging outcome was voxel-mirrored homotopic connectivity (VMHC). Static and dynamic functional connectivity (sFC, DFC) analyses were used to study the neuroplasticity reorganisation.

RESULTS

46 ISs (mean(SD) age, 59.37 (11.36) years) and 34 HCs (mean(SD) age, 52.88 (9.69) years) were included in the per-protocol analysis of clinical and neuroimaging. In clinical, ΔFM scores were 5.00 in HA group and 2.50 in NA group, with a dual correlation between ΔFM and ΔVMHC (angular: r=0.696, p=0.000; cerebellum: r=-0.716, p=0.000) fitting the linear regression model (R2=0.828). In neuroimaging, ISs demonstrated decreased VMHC in bilateral postcentral gyrus and cerebellum (Gaussian random field, GRF corrected, voxel p<0.001, cluster p<0.05), which fitted the logistic regression model (AUC=0.8413, accuracy=0.7500). Following acupuncture, VMHC in bilateral superior frontal gyrus orbital part was increased with cerebro-cerebellar changes, involving higher sFC between ipsilesional superior frontal gyrus orbital part and the contralesional orbitofrontal cortex as well as cerebellum (GRF corrected, voxel p<0.001, cluster p<0.05). The coefficient of variation of VMHC was decreased in bilateral posterior cingulate gyrus (PPC) locally (GRF corrected, voxel p<0.001, cluster p<0.05), with integration states transforming into segregation states overall (p<0.05). There was no acupuncture-related adverse event.

CONCLUSIONS

The randomised clinical and neuroimaging trial demonstrated acupuncture could promote the motor recovery and modified cerebro-cerebellar VMHC via bilateral static and dynamic reorganisations for IS patients with hemiparesis.

Alkaptonuria: From Molecular Insights to a Dedicated Digital Platform

Alkaptonuria (AKU) is a genetic disorder that affects connective tissues of several body compartments causing cartilage degeneration, tendon calcification, heart problems, and an invalidating, early-onset form of osteoarthritis. The molecular mechanisms underlying AKU involve homogentisic acid (HGA) accumulation in cells and tissues. HGA is highly reactive, able to modify several macromolecules, and activates different pathways, mostly involved in the onset and propagation of oxidative stress and inflammation, with consequences spreading from the microscopic to the macroscopic level leading to irreversible damage. Gaining a deeper understanding of AKU molecular mechanisms may provide novel possible therapeutical approaches to counteract disease progression. In this review, we first describe inflammation and oxidative stress in AKU and discuss similarities with other more common disorders. Then, we focus on HGA reactivity and AKU molecular mechanisms. We finally describe a multi-purpose digital platform, named ApreciseKUre, created to facilitate data collection, integration, and analysis of AKU-related data.

Hematoma expansion prediction based on SMOTE and XGBoost algorithm

Hematoma expansion (HE) is a high risky symptom with high rate of occurrence for patients who have undergone spontaneous intracerebral hemorrhage (ICH) after a major accident or illness. Correct prediction of the occurrence of HE in advance is critical to help the doctors to determine the next step medical treatment. Most existing studies focus only on the occurrence of HE within 6 h after the occurrence of ICH, while in reality a considerable number of patients have HE after the first 6 h but within 24 h. In this study, based on the medical doctors recommendation, we focus on prediction of the occurrence of HE within 24 h, as well as the occurrence of HE every 6 h within 24 h. Based on the demographics and computer tomography (CT) image extraction information, we used the XGBoost method to predict the occurrence of HE within 24 h. In this study, to solve the issue of highly imbalanced data set, which is a frequent case in medical data analysis, we used the SMOTE algorithm for data augmentation. To evaluate our method, we used a data set consisting of 582 patients records, and compared the results of proposed method as well as few machine learning methods. Our experiments show that XGBoost achieved the best prediction performance on the balanced dataset processed by the SMOTE algorithm with an accuracy of 0.82 and F1-score of 0.82. Moreover, our proposed method predicts the occurrence of HE within 6, 12, 18 and 24 h at the accuracy of 0.89, 0.82, 0.87 and 0.94, indicating that the HE occurrence within 24 h can be predicted accurately by the proposed method.

Prediction of Adolescents' Fluid Intelligence Scores based on Deep Learning with Reconstruction Regularization

Objective

The aim of this study was to develop a predictive model for uncorrected/actual fluid intelligence scores in 9-10 year old children using magnetic resonance T1-weighted imaging. Explore the predictive performance of an autoencoder model based on reconstruction regularization for fluid intelligence in adolescents.

Methods

We collected actual fluid intelligence scores and T1-weighted MRIs of 11,534 adolescents who completed baseline tasks from ABCD Data Release 3.0. A total of 148 ROIs were selected and 604 features were proposed by FreeSurfer segmentation. The training and testing sets were divided in a ratio of 7:3. To predict fluid intelligence scores, we used AE, MLP and classic machine learning models, and compared their performance on the test set. In addition, we explored their performance across gender subpopulations. Moreover, we evaluated the importance of features using the SHapley Additive Explain method. Results: The proposed model achieves optimal performance on the test set for predicting actual fluid intelligence scores (PCC = 0.209 ± 0.02, MSE = 105.212 ± 2.53). Results show that autoencoders with refactoring regularization are significantly more effective than MLPs and classical machine learning models. In addition, all models performed better on female adolescents than on male adolescents. Further analysis of relevant characteristics in different populations revealed that this may be related to gender differences in underlying fluid intelligence mechanisms.

Conclusions

We construct a weak but stable correlation between brain structural features and raw fluid intelligence using autoencoders. Future research may need to explore ensemble regression strategies utilizing multiple machine learning algorithms on multimodal data in order to improve the predictive performance of fluid intelligence based on neuroimaging features.

Antenatal screening for fetal structural anomalies - Routine or targeted practice?

Antenatal screening with ultrasound identifies fetal structural anomalies in 3-6% of pregnancies. Identification of anomalies during pregnancy provides an opportunity for counselling, targeted imaging, genetic testing, fetal intervention and delivery planning. Ultrasound is the primary modality for imaging the fetus in pregnancy, but magnetic resonance imaging (MRI) is evolving as an adjunctive tool providing additional structural and functional information. Screening should start from the first trimester when more than 50% of severe defects can be detected. The mid-trimester ultrasound balances the benefits of increased fetal growth and development to improve detection rates, whilst still providing timely management options. A routine third trimester ultrasound may detect acquired anomalies or those missed earlier in pregnancy but may not be available in all settings. Targeted imaging by fetal medicine experts improves detection in high-risk pregnancies or when an anomaly has been detected, allowing accurate phenotyping, access to advanced genetic testing and expert counselling.

Personalized prediction of survival rate with combination of penalized Cox models in patients with colorectal cancer

The investigation into individual survival rates within the patient population was typically conducted using the Cox proportional hazards model. This study was aimed to evaluate the performance of machine learning algorithm in predicting survival rates more than 5 years for individual patients with colorectal cancer. A total of 475 patients with colorectal cancer (CRC) and complete data who had underwent surgery for CRC were analyze to measure individual's survival rate more than 5 years using a machine learning based on penalized Cox regression. We conducted thorough calculations to measure the individual's survival rate more than 5 years for performance evaluation. The receiver operating characteristic curves for the LASSO penalized model, the SCAD penalized model, the unpenalized model, and the RSF model were analyzed. The least absolute shrinkage and selection operator penalized model displayed a mean AUC of 0.67 ± 0.06, the smoothly clipped absolute deviation penalized model exhibited a mean AUC of 0.65 ± 0.07, the unpenalized model showed a mean AUC of 0.64 ± 0.09. Notably, the random survival forests model outperformed the others, demonstrating the most favorable performance evaluation with a mean AUC of 0.71 ± 0.05. Compared to the conventional unpenalized Cox model, recent machine learning techniques (LASSO, SCAD, RSF) showed advantages for data interpretation.

Performance Drift in Machine Learning Models for Cardiac Surgery Risk Prediction: Retrospective Analysis

Background

The Society of Thoracic Surgeons and European System for Cardiac Operative Risk Evaluation (EuroSCORE) II risk scores are the most commonly used risk prediction models for in-hospital mortality after adult cardiac surgery. However, they are prone to miscalibration over time and poor generalization across data sets; thus, their use remains controversial. Despite increased interest, a gap in understanding the effect of data set drift on the performance of machine learning (ML) over time remains a barrier to its wider use in clinical practice. Data set drift occurs when an ML system underperforms because of a mismatch between the data it was developed from and the data on which it is deployed.

Objective

In this study, we analyzed the extent of performance drift using models built on a large UK cardiac surgery database. The objectives were to (1) rank and assess the extent of performance drift in cardiac surgery risk ML models over time and (2) investigate any potential influence of data set drift and variable importance drift on performance drift.

Methods

We conducted a retrospective analysis of prospectively, routinely gathered data on adult patients undergoing cardiac surgery in the United Kingdom between 2012 and 2019. We temporally split the data 70:30 into a training and validation set and a holdout set. Five novel ML mortality prediction models were developed and assessed, along with EuroSCORE II, for relationships between and within variable importance drift, performance drift, and actual data set drift. Performance was assessed using a consensus metric.

Results

A total of 227,087 adults underwent cardiac surgery during the study period, with a mortality rate of 2.76% (n=6258). There was strong evidence of a decrease in overall performance across all models (P<.0001). Extreme gradient boosting (clinical effectiveness metric [CEM] 0.728, 95% CI 0.728-0.729) and random forest (CEM 0.727, 95% CI 0.727-0.728) were the overall best-performing models, both temporally and nontemporally. EuroSCORE II performed the worst across all comparisons. Sharp changes in variable importance and data set drift from October to December 2017, from June to July 2018, and from December 2018 to February 2019 mirrored the effects of performance decrease across models.

Conclusions

All models show a decrease in at least 3 of the 5 individual metrics. CEM and variable importance drift detection demonstrate the limitation of logistic regression methods used for cardiac surgery risk prediction and the effects of data set drift. Future work will be required to determine the interplay between ML models and whether ensemble models could improve on their respective performance advantages.

Enhancing nasal endoscopy: Classification, detection, and segmentation of anatomic landmarks using a convolutional neural network

KEY POINTS

A convolutional neural network (CNN)-based model can accurately localize and segment turbinates in images obtained during nasal endoscopy (NE). This model represents a starting point for algorithms that comprehensively interpret NE findings.

Chinese Oncologists' Perspectives on Integrating AI into Clinical Practice: Cross-Sectional Survey Study

BACKGROUND

The rapid development of artificial intelligence (AI) has brought significant interest to its potential applications in oncology. Although AI-powered tools are already being implemented in some Chinese hospitals, their integration into clinical practice raises several concerns for Chinese oncologists.

OBJECTIVE

This study aims to explore the concerns of Chinese oncologists regarding the integration of AI into clinical practice and to identify the factors influencing these concerns.

METHODS

A total of 228 Chinese oncologists participated in a cross-sectional web-based survey from April to June in 2023 in mainland China. The survey gauged their worries about AI with multiple-choice questions. The survey evaluated their views on the statements of "The impact of AI on the doctor-patient relationship" and "AI will replace doctors." The data were analyzed using descriptive statistics, and variate analyses were used to find correlations between the oncologists' backgrounds and their concerns.

RESULTS

The study revealed that the most prominent concerns were the potential for AI to mislead diagnosis and treatment (163/228, 71.5%); an overreliance on AI (162/228, 71%); data and algorithm bias (123/228, 54%); issues with data security and patient privacy (123/228, 54%); and a lag in the adaptation of laws, regulations, and policies in keeping up with AI's development (115/228, 50.4%). Oncologists with a bachelor's degree expressed heightened concerns related to data and algorithm bias (34/49, 69%; P=.03) and the lagging nature of legal, regulatory, and policy issues (32/49, 65%; P=.046). Regarding AI's impact on doctor-patient relationships, 53.1% (121/228) saw a positive impact, whereas 35.5% (81/228) found it difficult to judge, 9.2% (21/228) feared increased disputes, and 2.2% (5/228) believed that there is no impact. Although sex differences were not significant (P=.08), perceptions varied-male oncologists tended to be more positive than female oncologists (74/135, 54.8% vs 47/93, 50%). Oncologists with a bachelor's degree (26/49, 53%; P=.03) and experienced clinicians (≥21 years; 28/56, 50%; P=.054). found it the hardest to judge. Those with IT experience were significantly more positive (25/35, 71%) than those without (96/193, 49.7%; P=.02). Opinions regarding the possibility of AI replacing doctors were diverse, with 23.2% (53/228) strongly disagreeing, 14% (32/228) disagreeing, 29.8% (68/228) being neutral, 16.2% (37/228) agreeing, and 16.7% (38/228) strongly agreeing. There were no significant correlations with demographic and professional factors (all P>.05).

CONCLUSIONS

Addressing oncologists' concerns about AI requires collaborative efforts from policy makers, developers, health care professionals, and legal experts. Emphasizing transparency, human-centered design, bias mitigation, and education about AI's potential and limitations is crucial. Through close collaboration and a multidisciplinary strategy, AI can be effectively integrated into oncology, balancing benefits with ethical considerations and enhancing patient care.

A multi-institutional machine learning algorithm for prognosticating facial nerve injury following microsurgical resection of vestibular schwannoma

Vestibular schwannomas (VS) are the most common tumor of the skull base with available treatment options that carry a risk of iatrogenic injury to the facial nerve, which can significantly impact patients' quality of life. As facial nerve outcomes remain challenging to prognosticate, we endeavored to utilize machine learning to decipher predictive factors relevant to facial nerve outcomes following microsurgical resection of VS. A database of patient-, tumor- and surgery-specific features was constructed via retrospective chart review of 242 consecutive patients who underwent microsurgical resection of VS over a 7-year study period. This database was then used to train non-linear supervised machine learning classifiers to predict facial nerve preservation, defined as House-Brackmann (HB) I vs. facial nerve injury, defined as HB II-VI, as determined at 6-month outpatient follow-up. A random forest algorithm demonstrated 90.5% accuracy, 90% sensitivity and 90% specificity in facial nerve injury prognostication. A random variable (rv) was generated by randomly sampling a Gaussian distribution and used as a benchmark to compare the predictiveness of other features. This analysis revealed age, body mass index (BMI), case length and the tumor dimension representing tumor growth towards the brainstem as prognosticators of facial nerve injury. When validated via prospective assessment of facial nerve injury risk, this model demonstrated 84% accuracy. Here, we describe the development of a machine learning algorithm to predict the likelihood of facial nerve injury following microsurgical resection of VS. In addition to serving as a clinically applicable tool, this highlights the potential of machine learning to reveal non-linear relationships between variables which may have clinical value in prognostication of outcomes for high-risk surgical procedures.

Integrating proteomics and explainable artificial intelligence: a comprehensive analysis of protein biomarkers for endometrial cancer diagnosis and prognosis

Endometrial cancer, which is the most common gynaecological cancer in women after breast, colorectal and lung cancer, can be diagnosed at an early stage. The first aim of this study is to classify age, tumor grade, myometrial invasion and tumor size, which play an important role in the diagnosis and prognosis of endometrial cancer, with machine learning methods combined with explainable artificial intelligence. 20 endometrial cancer patients proteomic data obtained from tumor biopsies taken from different regions of EC tissue were used. The data obtained were then classified according to age, tumor size, tumor grade and myometrial invasion. Then, by using three different machine learning methods, explainable artificial intelligence was applied to the model that best classifies these groups and possible protein biomarkers that can be used in endometrial prognosis were evaluated. The optimal model for age classification was XGBoost with AUC (98.8%), for tumor grade classification was XGBoost with AUC (98.6%), for myometrial invasion classification was LightGBM with AUC (95.1%), and finally for tumor size classification was XGBoost with AUC (94.8%). By combining the optimal models and the SHAP approach, possible protein biomarkers and their expressions were obtained for classification. Finally, EWRS1 protein was found to be common in three groups (age, myometrial invasion, tumor size). This article's findings indicate that models have been developed that can accurately classify factors including age, tumor grade, and myometrial invasion all of which are critical for determining the prognosis of endometrial cancer as well as potential protein biomarkers associated with these factors. Furthermore, we were able to provide an analysis of how the quantities of the proteins suggested as biomarkers varied throughout the classes by combining the SHAP values with these ideal models.