Artificial intelligence in healthcare

How do medical professionals make sense (or not) of AI? A social-media-based computational grounded theory study and an online survey

To investigate opinions and attitudes of medical professionals towards adopting AI-enabled healthcare technologies in their daily business, we used a mixed-methods approach. Study 1 employed a qualitative computational grounded theory approach analyzing 181 Reddit threads in the several subreddits of r/medicine. By utilizing an unsupervised machine learning clustering method, we identified three key themes: (1) consequences of AI, (2) physician-AI relationship, and (3) a proposed way forward. In particular Reddit posts related to the first two themes indicated that the medical professionals' fear of being replaced by AI and skepticism toward AI played a major role in the argumentations. Moreover, the results suggest that this fear is driven by little or moderate knowledge about AI. Posts related to the third theme focused on factual discussions about how AI and medicine have to be designed to become broadly adopted in health care. Study 2 quantitatively examined the relationship between the fear of AI, knowledge about AI, and medical professionals' intention to use AI-enabled technologies in more detail. Results based on a sample of 223 medical professionals who participated in the online survey revealed that the intention to use AI technologies increases with increasing knowledge about AI and that this effect is moderated by the fear of being replaced by AI.

Identification of MORN3 and LLGL2 as novel diagnostic biomarkers for latent tuberculosis infection using machine learning strategies and experimental verification

BACKGROUND

Current diagnostic methods cannot effectively distinguish between latent tuberculosis infection (LTBI) and active tuberculosis (ATB). This study aims to explore novel non-invasive diagnostic biomarkers for LTBI and to elucidate possible molecular mechanisms of LTBI pathogenesis.

METHODS

Three GEO datasets (GSE19439, GSE19444, and GSE62525) were utilized to analyze the differentially expressed genes (DEGs). Functional enrichment studies were then performed on these DEGs. To ascertain potential diagnostic biomarkers, we utilized two different machine learning techniques: LASSO and RF. ROC curves were constructed in both the training and validation datasets to assess the diagnostic efficacy. The expression of identified biomarkers was verified by RT-qPCR in our own Chinese cohort. Using CIBERSORT, we estimated the abundances of 22 immune cell types in LTBI group, and subsequently analyzed the relationship between biomarker expression and immune cell infiltration.

RESULTS

166 DEGs were identified between ATB and LTBI groups, which are primarily associated with immune responses, inflammatory signaling pathways, and infection factors. Following that, 22 candidate diagnostic biomarkers for LTBI were selected in the machine learning process. Three up-regulated genes, MORN3, LLGL2, and IFT140, whose expression levels were not previously reported in TB, were validated using the training and validation cohort datasets. In our own Chinese cohort, we also found that MORN3 and LLGL2 showed good diagnostic effect using RT-qPCR method. Finally, we revealed the specific infiltration features of immune cells in LTBI and observed a notable correlation between potential marker expression and immune cells.

CONCLUSIONS

MORN3 and LLGL2 emerged as candidate diagnostic biomarkers for LTBI, following the elucidation of the key immune cell types involved. Our findings will contribute to providing a potential target for early noninvasive diagnosis of LTBI patients.

Seeing the random forest through the decision trees. Supporting learning health systems from histopathology with machine learning models: Challenges and opportunities

This paper discusses some overlooked challenges faced when working with machine learning models for histopathology and presents a novel opportunity to support "Learning Health Systems" with them. Initially, the authors elaborate on these challenges after separating them according to their mitigation strategies: those that need innovative approaches, time, or future technological capabilities and those that require a conceptual reappraisal from a critical perspective. Then, a novel opportunity to support "Learning Health Systems" by integrating hidden information extracted by ML models from digitalized histopathology slides with other healthcare big data is presented.

Recent advances and applications of artificial intelligence in 3D bioprinting

3D bioprinting techniques enable the precise deposition of living cells, biomaterials, and biomolecules, emerging as a promising approach for engineering functional tissues and organs. Meanwhile, recent advances in 3D bioprinting enable researchers to build in vitro models with finely controlled and complex micro-architecture for drug screening and disease modeling. Recently, artificial intelligence (AI) has been applied to different stages of 3D bioprinting, including medical image reconstruction, bioink selection, and printing process, with both classical AI and machine learning approaches. The ability of AI to handle complex datasets, make complex computations, learn from past experiences, and optimize processes dynamically makes it an invaluable tool in advancing 3D bioprinting. The review highlights the current integration of AI in 3D bioprinting and discusses future approaches to harness the synergistic capabilities of 3D bioprinting and AI for developing personalized tissues and organs.

Implementing AI in Hospitals to Achieve a Learning Health System: Systematic Review of Current Enablers and Barriers

BACKGROUND

Efforts are underway to capitalize on the computational power of the data collected in electronic medical records (EMRs) to achieve a learning health system (LHS). Artificial intelligence (AI) in health care has promised to improve clinical outcomes, and many researchers are developing AI algorithms on retrospective data sets. Integrating these algorithms with real-time EMR data is rare. There is a poor understanding of the current enablers and barriers to empower this shift from data set-based use to real-time implementation of AI in health systems. Exploring these factors holds promise for uncovering actionable insights toward the successful integration of AI into clinical workflows.

OBJECTIVE

The first objective was to conduct a systematic literature review to identify the evidence of enablers and barriers regarding the real-world implementation of AI in hospital settings. The second objective was to map the identified enablers and barriers to a 3-horizon framework to enable the successful digital health transformation of hospitals to achieve an LHS.

METHODS

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were adhered to. PubMed, Scopus, Web of Science, and IEEE Xplore were searched for studies published between January 2010 and January 2022. Articles with case studies and guidelines on the implementation of AI analytics in hospital settings using EMR data were included. We excluded studies conducted in primary and community care settings. Quality assessment of the identified papers was conducted using the Mixed Methods Appraisal Tool and ADAPTE frameworks. We coded evidence from the included studies that related to enablers of and barriers to AI implementation. The findings were mapped to the 3-horizon framework to provide a road map for hospitals to integrate AI analytics.

RESULTS

Of the 1247 studies screened, 26 (2.09%) met the inclusion criteria. In total, 65% (17/26) of the studies implemented AI analytics for enhancing the care of hospitalized patients, whereas the remaining 35% (9/26) provided implementation guidelines. Of the final 26 papers, the quality of 21 (81%) was assessed as poor. A total of 28 enablers was identified; 8 (29%) were new in this study. A total of 18 barriers was identified; 5 (28%) were newly found. Most of these newly identified factors were related to information and technology. Actionable recommendations for the implementation of AI toward achieving an LHS were provided by mapping the findings to a 3-horizon framework.

CONCLUSIONS

Significant issues exist in implementing AI in health care. Shifting from validating data sets to working with live data is challenging. This review incorporated the identified enablers and barriers into a 3-horizon framework, offering actionable recommendations for implementing AI analytics to achieve an LHS. The findings of this study can assist hospitals in steering their strategic planning toward successful adoption of AI.

The diagnostic and triage accuracy of the GPT-3 artificial intelligence model: an observational study

BACKGROUND

Artificial intelligence (AI) applications in health care have been effective in many areas of medicine, but they are often trained for a single task using labelled data, making deployment and generalisability challenging. How well a general-purpose AI language model performs diagnosis and triage relative to physicians and laypeople is not well understood.

METHODS

We compared the predictive accuracy of Generative Pre-trained Transformer 3 (GPT-3)'s diagnostic and triage ability for 48 validated synthetic case vignettes (<50 words; sixth-grade reading level or below) of both common (eg, viral illness) and severe (eg, heart attack) conditions to a nationally representative sample of 5000 lay people from the USA who could use the internet to find the correct options and 21 practising physicians at Harvard Medical School. There were 12 vignettes for each of four triage categories: emergent, within one day, within 1 week, and self-care. The correct diagnosis and triage category (ie, ground truth) for each vignette was determined by two general internists at Harvard Medical School. For each vignette, human respondents and GPT-3 were prompted to list diagnoses in order of likelihood, and the vignette was marked as correct if the ground-truth diagnosis was in the top three of the listed diagnoses. For triage accuracy, we examined whether the human respondents' and GPT-3's selected triage was exactly correct according to the four triage categories, or matched a dichotomised triage variable (emergent or within 1 day vs within 1 week or self-care). We estimated GPT-3's diagnostic and triage confidence on a given vignette using a modified bootstrap resampling procedure, and examined how well calibrated GPT-3's confidence was by computing calibration curves and Brier scores. We also performed subgroup analysis by case acuity, and an error analysis for triage advice to characterise how its advice might affect patients using this tool to decide if they should seek medical care immediately.

FINDINGS

Among all cases, GPT-3 replied with the correct diagnosis in its top three for 88% (42/48, 95% CI 75-94) of cases, compared with 54% (2700/5000, 53-55) for lay individuals (p<0.0001) and 96% (637/666, 94-97) for physicians (p=0·012). GPT-3 triaged 70% correct (34/48, 57-82) versus 74% (3706/5000, 73-75; p=0.60) for lay individuals and 91% (608/666, 89-93%; p<0.0001) for physicians. As measured by the Brier score, GPT-3 confidence in its top prediction was reasonably well calibrated for diagnosis (Brier score=0·18) and triage (Brier score=0·22). We observed an inverse relationship between case acuity and GPT-3 accuracy (p<0·0001) with a fitted trend line of -8·33% decrease in accuracy for every level of increase in case acuity. For triage error analysis, GPT-3 deprioritised truly emergent cases in seven instances.

INTERPRETATION

A general-purpose AI language model without any content-specific training could perform diagnosis at levels close to, but below, physicians and better than lay individuals. We found that GPT-3's performance was inferior to physicians for triage, sometimes by a large margin, and its performance was closer to that of lay individuals. Although the diagnostic performance of GPT-3 was comparable to physicians, it was significantly better than a typical person using a search engine.

FUNDING

The National Heart, Lung, and Blood Institute.

Interpretation of acid-base metabolism on arterial blood gas samples via machine learning algorithms

BACKGROUND

Arterial blood gas evaluation is crucial for critically ill patients, as it provides essential information about acid-base metabolism and respiratory balance, but evaluation can be complex and time-consuming. Artificial intelligence can perform tasks that require human intelligence, and it is revolutionizing healthcare through technological advancements.

AIM

This study aims to assess arterial blood gas evaluation using artificial intelligence algorithms.

METHODS

The study included 21.541 retrospective arterial blood gas samples, categorized into 15 different classes by experts for evaluating acid-base metabolism status. Six machine learning algorithms were utilized; accuracy, balanced accuracy, sensitivity, specificity, precision, and F1 values of the models were determined; and ROC curves were drawn to assess areas under the curve for each class. Evaluation of which sample was estimated in which class was conducted using the confusion matrices of the models.

RESULTS

The bagging classifier (BC) model achieved the highest balanced accuracy with 99.24%, whereas the XGBoost model reached the highest accuracy with 99.66%. The BC model shows 100% sensitivity for nine classes and 100% specificity for 10 classes, and the model correctly predicted 6438 of 6463 test samples and achieved an accuracy of 99.61%, with an area under the curve > 0.9 in all classes on a class basis.

CONCLUSION

The machine learning models developed exhibited remarkable accuracy, sensitivity, and specificity in predicting the status of acid-base metabolism. However, implementing these models can aid clinicians, freeing up their time for more intricate tasks.

Artificial intelligence in clinical nutrition and dietetics: A brief overview of current evidence

The rapid surge in artificial intelligence (AI) has dominated technological innovation in today's society. As experts begin to understand the potential, a spectrum of opportunities could yield a remarkable revolution. The upsurge in healthcare could transform clinical interventions and outcomes, but it risks dehumanization and increased unethical practices. The field of clinical nutrition and dietetics is no exception. This article finds a multitude of developments underway, which include the use of AI for malnutrition screening; predicting clinical outcomes, such as disease onset, and clinical risks, such as drug interactions; aiding interventions, such as estimating nutrient intake; applying precision nutrition, such as measuring postprandial glycemic response; and supporting workflow through chatbots trained on natural language models. Although the opportunity and scalability of AI is incalculably attractive, especially in the face of poor healthcare resources, the threat cannot be ignored. The risk of malpractice and lack of accountability are some of the main concerns. As such, the healthcare professional's responsibility remains paramount. The data used to train AI models could be biased, which could risk the quality of care to vulnerable or minority patient groups. Standardized AI-development protocols, benchmarked to care recommendations, with rigorous large-scale validation are required to maximize application among different settings. AI could overturn the healthcare landscape, and this article skims the surface of its potential in clinical nutrition and dietetics.

Deciphering the Feature Representation of Deep Neural Networks for High-Performance AI

AI driven by deep learning is transforming many aspects of science and technology. The enormous success of deep learning stems from its unique capability of extracting essential features from Big Data for decision-making. However, the feature extraction and hidden representations in deep neural networks (DNNs) remain inexplicable, primarily because of lack of technical tools to comprehend and interrogate the feature space data. The main hurdle here is that the feature data are often noisy in nature, complex in structure, and huge in size and dimensionality, making it intractable for existing techniques to analyze the data reliably. In this work, we develop a computational framework named contrastive feature analysis (CFA) to facilitate the exploration of the DNN feature space and improve the performance of AI. By utilizing the interaction relations among the features and incorporating a novel data-driven kernel formation strategy into the feature analysis pipeline, CFA mitigates the limitations of traditional approaches and provides an urgently needed solution for the analysis of feature space data. The technique allows feature data exploration in unsupervised, semi-supervised and supervised formats to address different needs of downstream applications. The potential of CFA and its applications for pruning of neural network architectures are demonstrated using several state-of-the-art networks and well-annotated datasets across different disciplines.

Optimising antibiotic exposure by customising the duration of treatment for respiratory tract infections based on patient needs in primary care

Primary care antimicrobial stewardship programs have limited success in reducing antibiotic use, prompting the search for new strategies. Convincing general practitioners to resist antibiotic prescription amid uncertainty or patient demands usually poses a significant challenge. Despite common practice, standard durations for common infections lack support from clinical studies. Contrary to common belief, extending antibiotic treatment beyond the resolution of symptoms does not seem to prevent or reduce antimicrobial resistance. Shortening the duration of antibiotic therapy has shown to be effective in mitigating the spread of resistance, particularly in cases of pneumonia. Recent hospital randomised trials suggest that ending antibiotic courses by day three for most lower respiratory tract infections is effective and safe. While community studies are scarce, it is likely that these shorter, tailored courses to meet patients' needs would also be effective and safe in primary care. Therefore, primary care studies should investigate the outcomes of advising patients to discontinue antibiotic treatment upon symptom resolution. Implementing patient-centred, customised treatment durations, rather than fixed courses, is crucial for meeting individual patient needs.

Performance Metrics, Algorithms, and Applications of Artificial Intelligence in Vascular and Interventional Neurology: A Review of Basic Elements

Artificial intelligence (AI) is currently being used as a routine tool for day-to-day activity. Medicine is not an exception to the growing usage of AI in various scientific fields. Vascular and interventional neurology deal with diseases that require early diagnosis and appropriate intervention, which are crucial to saving patients' lives. In these settings, AI can be an extra pair of hands for physicians or in conditions where there is a shortage of clinical experts. In this article, the authors have reviewed the common metrics used in interpreting the performance of models and common algorithms used in this field.

Development of machine learning models for fractional flow reserve prediction in angiographically intermediate coronary lesions

BACKGROUND

Fractional flow reserve (FFR) represents the gold standard in guiding the decision to proceed or not with coronary revascularization of angiographically intermediate coronary lesion (AICL). Optical coherence tomography (OCT) allows to carefully characterize coronary plaque morphology and lumen dimensions.

OBJECTIVES

We sought to develop machine learning (ML) models based on clinical, angiographic and OCT variables for predicting FFR.

METHODS

Data from a multicenter, international, pooled analysis of individual patient's level data from published studies assessing FFR and OCT on the same target AICL were collected through a dedicated database to train (n = 351) and validate (n = 151) six two-class supervised ML models employing 25 clinical, angiographic and OCT variables.

RESULTS

A total of 502 coronary lesions in 489 patients were included. The AUC of the six ML models ranged from 0.71 to 0.78, whereas the measured F1 score was from 0.70 to 0.75. The ML algorithms showed moderate sensitivity (range: 0.68-0.77) and specificity (range: 0.59-0.69) in detecting patients with a positive or negative FFR. In the sensitivity analysis, using 0.75 as FFR cut-off, we found a higher AUC (0.78-0.86) and a similar F1 score (range: 0.63-0.76). Specifically, the six ML models showed a higher specificity (0.71-0.84), with a similar sensitivity (0.58-0.80) with respect to 0.80 cut-off.

CONCLUSIONS

ML algorithms derived from clinical, angiographic, and OCT parameters can identify patients with a positive or negative FFR.

Use of artificial intelligence embryo selection based on static images to predict first-trimester pregnancy loss

RESEARCH QUESTION

Can an artificial intelligence embryo selection assistant predict the incidence of first-trimester spontaneous abortion using static images of IVF embryos?

DESIGN

In a blind, retrospective study, a cohort of 172 blastocysts from IVF cases with single embryo transfer and a positive biochemical pregnancy test was ranked retrospectively by the artificial intelligence morphometric algorithm ERICA. Making use of static embryo images from a light microscope, each blastocyst was assigned to one of four possible groups (optimal, good, fair or poor), and linear regression was used to correlate the results with the presence or absence of a normal fetal heart beat as an indicator of ongoing pregnancy or spontaneous abortion, respectively. Additional analyses included modelling for recipient age and chromosomal status established by preimplantation genetic testing for aneuploidy (PGT-A).

RESULTS

Embryos classified as optimal/good had a lower incidence of spontaneous abortion (16.1%) compared with embryos classified as fair/poor (25%; OR = 0.46, P = 0.005). The incidence of spontaneous abortion in chromosomally normal embryos (determined by PGT-A) was 13.3% for optimal/good embryos and 20.0% for fair/poor embryos, although the difference was not significant (P = 0.531). There was a significant association between embryo rank and recipient age (P = 0.018), in that the incidence of spontaneous abortion was unexpectedly lower in older recipients (21.3% for age ≤35 years, 17.9% for age 36-38 years, 16.4% for age ≥39 years; OR = 0.354, P = 0.0181). Overall, these results support correlation between risk of spontaneous abortion and embryo rank as determined by artificial intelligence; classification accuracy was calculated to be 67.4%.

CONCLUSIONS

This preliminary study suggests that artificial intelligence (ERICA), which was designed as a ranking system to assist with embryo transfer decisions and ploidy prediction, may also be useful to provide information for couples on the risk of spontaneous abortion. Future work will include a larger sample size and karyotyping of miscarried pregnancy tissue.

Accuracy of Artificial Intelligence-Based Virtual Assistants in Responding to Frequently Asked Questions Related to Orthognathic Surgery

BACKGROUND

Despite increasing interest in how conversational agents might improve health care delivery and information dissemination, there is limited research assessing the quality of health information provided by these technologies, especially in orthognathic surgery (OGS).

PURPOSE

This study aimed to measure and compare the quality of four virtual assistants (VAs) in addressing the frequently asked questions about OGS.

STUDY DESIGN, SETTING, AND SAMPLE

This in-silico cross-sectional study assessed the responses of a sample of four VAs through a standardized set of 10 questionnaires related to OGS.

INDEPENDENT VARIABLE

The independent variables were the four VAs. The four VAs tested were VA1: Alexa (Seattle, Washington), VA2: Google Assistant (Google Mountain View, California), VA3: Siri (Cupertino, California), and VA4: Bing (San Diego, California).

MAIN OUTCOME VARIABLE(S)

The primary outcome variable was the quality of the answers generated by the four VAs. Four investigators (two orthodontists and two oral surgeons) assessed the quality of response of the four VAs through a standardized set of 10 questionnaires using a five-point modified Likert scale, with the lowest score (1) signifying the highest quality. The main outcome variables measured were the combined mean scores of the responses from each VA, and the secondary outcome assessed was the variability in responses among the different investigators.

COVARIATES

None.

ANALYSES

One-way analysis of variance was done to compare the average scores per question. One-way analysis of variance followed by Tukey's post hoc analyses was done to compare the combined mean scores among the VAs, and the combined mean scores of all questions were evaluated to determine variability if any among different VA's responses to the investigators.

RESULTS

Among the four VAs, VA4 (1.32 ± 0.57) had significantly the lowest (best) score, followed by VA2 (1.55 ± 0.78), VA1 (2.67 ± 1.49), and VA3 (3.52 ± 0.50) (P value <.001). There were no significant differences in how the VAs: VA3 (P value = .46), VA4 (P value = .45), and VA2 (P value = .44) responded to each of the investigators except VA1 (P value = .003).

CONCLUSION AND RELEVANCE

The VAs responded to the queries related to OGS, with VA4 displaying the best quality response, followed by VA2, VA1, and VA3. Technology companies and clinical organizations should partner for an intelligent VA with evidence-based responses specifically curated to educate patients.

Machine learning-enabled prediction of prolonged length of stay in hospital after surgery for tuberculosis spondylitis patients with unbalanced data: a novel approach using explainable artificial intelligence (XAI)

BACKGROUND

Tuberculosis spondylitis (TS), commonly known as Pott's disease, is a severe type of skeletal tuberculosis that typically requires surgical treatment. However, this treatment option has led to an increase in healthcare costs due to prolonged hospital stays (PLOS). Therefore, identifying risk factors associated with extended PLOS is necessary. In this research, we intended to develop an interpretable machine learning model that could predict extended PLOS, which can provide valuable insights for treatments and a web-based application was implemented.

METHODS

We obtained patient data from the spine surgery department at our hospital. Extended postoperative length of stay (PLOS) refers to a hospitalization duration equal to or exceeding the 75th percentile following spine surgery. To identify relevant variables, we employed several approaches, such as the least absolute shrinkage and selection operator (LASSO), recursive feature elimination (RFE) based on support vector machine classification (SVC), correlation analysis, and permutation importance value. Several models using implemented and some of them are ensembled using soft voting techniques. Models were constructed using grid search with nested cross-validation. The performance of each algorithm was assessed through various metrics, including the AUC value (area under the curve of receiver operating characteristics) and the Brier Score. Model interpretation involved utilizing methods such as Shapley additive explanations (SHAP), the Gini Impurity Index, permutation importance, and local interpretable model-agnostic explanations (LIME). Furthermore, to facilitate the practical application of the model, a web-based interface was developed and deployed.

RESULTS

The study included a cohort of 580 patients and 11 features include (CRP, transfusions, infusion volume, blood loss, X-ray bone bridge, X-ray osteophyte, CT-vertebral destruction, CT-paravertebral abscess, MRI-paravertebral abscess, MRI-epidural abscess, postoperative drainage) were selected. Most of the classifiers showed better performance, where the XGBoost model has a higher AUC value (0.86) and lower Brier Score (0.126). The XGBoost model was chosen as the optimal model. The results obtained from the calibration and decision curve analysis (DCA) plots demonstrate that XGBoost has achieved promising performance. After conducting tenfold cross-validation, the XGBoost model demonstrated a mean AUC of 0.85 ± 0.09. SHAP and LIME were used to display the variables' contributions to the predicted value. The stacked bar plots indicated that infusion volume was the primary contributor, as determined by Gini, permutation importance (PFI), and the LIME algorithm.

CONCLUSIONS

Our methods not only effectively predicted extended PLOS but also identified risk factors that can be utilized for future treatments. The XGBoost model developed in this study is easily accessible through the deployed web application and can aid in clinical research.

Understanding Natural Language: Potential Application of Large Language Models to Ophthalmology

Large language models (LLMs), a natural language processing technology based on deep learning, are currently in the spotlight. These models closely mimic natural language comprehension and generation. Their evolution has undergone several waves of innovation similar to convolutional neural networks. The transformer architecture advancement in generative artificial intelligence marks a monumental leap beyond early-stage pattern recognition via supervised learning. With the expansion of parameters and training data (terabytes), LLMs unveil remarkable human interactivity, encompassing capabilities such as memory retention and comprehension. These advances make LLMs particularly well-suited for roles in healthcare communication between medical practitioners and patients. In this comprehensive review we discuss the trajectory of LLMs and potential implications for clinicians and patients. For clinicians, LLMs can be used for automated medical documentation, and given better inputs and extensive validation, LLMs may be able to autonomously diagnose and treat in the future. For patient care, LLMs can be used for triage suggestions, summarization of medical documents, explanation of a patient's condition, and customizing patient education materials tailored to their comprehension level. The limitations of LLM and possible solutions for real world use are also presented. Given the rapid advancements in this area, this review attempts to briefly cover many roles that LLMs may play in the ophthalmic space, with a focus on improving the quality of healthcare delivery.

Advances in artificial intelligence for meibomian gland evaluation: A comprehensive review

Meibomian gland dysfunction (MGD) is increasingly recognized as a critical contributor to evaporative dry eye, significantly impacting visual quality. With a global prevalence estimated at 35.8 %, it presents substantial challenges for clinicians. Conventional manual evaluation techniques for MGD face limitations characterized by inefficiencies, high subjectivity, limited big data processing capabilities, and a dearth of quantitative analytical tools. With rapidly advancing artificial intelligence (AI) techniques revolutionizing ophthalmology, studies are now leveraging sophisticated AI methodologies--including computer vision, unsupervised learning, and supervised learning--to facilitate comprehensive analyses of meibomian gland (MG) evaluations. These evaluations employ various techniques, including slit lamp examination, infrared imaging, confocal microscopy, and optical coherence tomography. This paradigm shift promises enhanced accuracy and consistency in disease evaluation and severity classification. While AI has achieved preliminary strides in meibomian gland evaluation, ongoing advancements in system development and clinical validation are imperative. We review the evolution of MG evaluation, juxtapose AI-driven methods with traditional approaches, elucidate the specific roles of diverse AI technologies, and explore their practical applications using various evaluation techniques. Moreover, we delve into critical considerations for the clinical deployment of AI technologies and envisages future prospects, providing novel insights into MG evaluation and fostering technological and clinical progress in this arena.

Second opinion machine learning for fast-track pathway assignment in hip and knee replacement surgery: the use of patient-reported outcome measures

BACKGROUND

The frequency of hip and knee arthroplasty surgeries has been rising steadily in recent decades. This trend is attributed to an aging population, leading to increased demands on healthcare systems. Fast Track (FT) surgical protocols, perioperative procedures designed to expedite patient recovery and early mobilization, have demonstrated efficacy in reducing hospital stays, convalescence periods, and associated costs. However, the criteria for selecting patients for FT procedures have not fully capitalized on the available patient data, including patient-reported outcome measures (PROMs).

METHODS

Our study focused on developing machine learning (ML) models to support decision making in assigning patients to FT procedures, utilizing data from patients' self-reported health status. These models are specifically designed to predict the potential health status improvement in patients initially selected for FT. Our approach focused on techniques inspired by the concept of controllable AI. This includes eXplainable AI (XAI), which aims to make the model's recommendations comprehensible to clinicians, and cautious prediction, a method used to alert clinicians about potential control losses, thereby enhancing the models' trustworthiness and reliability.

RESULTS

Our models were trained and tested using a dataset comprising 899 records from individual patients admitted to the FT program at IRCCS Ospedale Galeazzi-Sant'Ambrogio. After training and selecting hyper-parameters, the models were assessed using a separate internal test set. The interpretable models demonstrated performance on par or even better than the most effective 'black-box' model (Random Forest). These models achieved sensitivity, specificity, and positive predictive value (PPV) exceeding 70%, with an area under the curve (AUC) greater than 80%. The cautious prediction models exhibited enhanced performance while maintaining satisfactory coverage (over 50%). Further, when externally validated on a separate cohort from the same hospital-comprising patients from a subsequent time period-the models showed no pragmatically notable decline in performance.

CONCLUSIONS

Our results demonstrate the effectiveness of utilizing PROMs as basis to develop ML models for planning assignments to FT procedures. Notably, the application of controllable AI techniques, particularly those based on XAI and cautious prediction, emerges as a promising approach. These techniques provide reliable and interpretable support, essential for informed decision-making in clinical processes.

Interpretation of Artificial Intelligence Models in Healthcare: A Pictorial Guide for Clinicians

Artificial intelligence (AI) models can play a more effective role in managing patients with the explosion of digital health records available in the healthcare industry. Machine-learning (ML) and deep-learning (DL) techniques are two methods used to develop predictive models that serve to improve the clinical processes in the healthcare industry. These models are also implemented in medical imaging machines to empower them with an intelligent decision system to aid physicians in their decisions and increase the efficiency of their routine clinical practices. The physicians who are going to work with these machines need to have an insight into what happens in the background of the implemented models and how they work. More importantly, they need to be able to interpret their predictions, assess their performance, and compare them to find the one with the best performance and fewer errors. This review aims to provide an accessible overview of key evaluation metrics for physicians without AI expertise. In this review, we developed four real-world diagnostic AI models (two ML and two DL models) for breast cancer diagnosis using ultrasound images. Then, 23 of the most commonly used evaluation metrics were reviewed uncomplicatedly for physicians. Finally, all metrics were calculated and used practically to interpret and evaluate the outputs of the models. Accessible explanations and practical applications empower physicians to effectively interpret, evaluate, and optimize AI models to ensure safety and efficacy when integrated into clinical practice.

AI-generated text in otolaryngology publications: a comparative analysis before and after the release of ChatGPT

PURPOSE

This study delves into the broader implications of artificial intelligence (AI) text generation technologies, including large language models (LLMs) and chatbots, on the scientific literature of otolaryngology. By observing trends in AI-generated text within published otolaryngology studies, this investigation aims to contextualize the impact of AI-driven tools that are reshaping scientific writing and communication.

METHODS

Text from 143 original articles published in JAMA Otolaryngology - Head and Neck Surgery was collected, representing periods before and after ChatGPT's release in November 2022. The text from each article's abstract, introduction, methods, results, and discussion were entered into ZeroGPT.com to estimate the percentage of AI-generated content. Statistical analyses, including T-Tests and Fligner-Killeen's tests, were conducted using R.

RESULTS

A significant increase was observed in the mean percentage of AI-generated text post-ChatGPT release, especially in the abstract (from 34.36 to 46.53%, p = 0.004), introduction (from 32.43 to 45.08%, p = 0.010), and discussion sections (from 15.73 to 25.03%, p = 0.015). Publications of authors from non-English speaking countries demonstrated a higher percentage of AI-generated text.

CONCLUSION

This study found that the advent of ChatGPT has significantly impacted writing practices among researchers publishing in JAMA Otolaryngology - Head and Neck Surgery, raising concerns over the accuracy of AI-created content and potential misinformation risks. This manuscript highlights the evolving dynamics between AI technologies, scientific communication, and publication integrity, emphasizing the urgent need for continued research in this dynamic field. The findings also suggest an increasing reliance on AI tools like ChatGPT, raising questions about their broader implications for scientific publishing.

Developing and testing a framework for coding general practitioners' free-text diagnoses in electronic medical records - a reliability study for generating training data in natural language processing

BACKGROUND

Diagnoses entered by general practitioners into electronic medical records have great potential for research and practice, but unfortunately, diagnoses are often in uncoded format, making them of little use. Natural language processing (NLP) could assist in coding free-text diagnoses, but NLP models require local training data to unlock their potential. The aim of this study was to develop a framework of research-relevant diagnostic codes, to test the framework using free-text diagnoses from a Swiss primary care database and to generate training data for NLP modelling.

METHODS

The framework of diagnostic codes was developed based on input from local stakeholders and consideration of epidemiological data. After pre-testing, the framework contained 105 diagnostic codes, which were then applied by two raters who independently coded randomly drawn lines of free text (LoFT) from diagnosis lists extracted from the electronic medical records of 3000 patients of 27 general practitioners. Coding frequency and mean occurrence rates (n and %) and inter-rater reliability (IRR) of coding were calculated using Cohen's kappa (Κ).

RESULTS

The sample consisted of 26,980 LoFT and in 56.3% no code could be assigned because it was not a specific diagnosis. The most common diagnostic codes were, 'dorsopathies' (3.9%, a code covering all types of back problems, including non-specific lower back pain, scoliosis, and others) and 'other diseases of the circulatory system' (3.1%). Raters were in almost perfect agreement (Κ ≥ 0.81) for 69 of the 105 diagnostic codes, and 28 codes showed a substantial agreement (K between 0.61 and 0.80). Both high coding frequency and almost perfect agreement were found in 37 codes, including codes that are particularly difficult to identify from components of the electronic medical record, such as musculoskeletal conditions, cancer or tobacco use.

CONCLUSION

The coding framework was characterised by a subset of very frequent and highly reliable diagnostic codes, which will be the most valuable targets for training NLP models for automated disease classification based on free-text diagnoses from Swiss general practice.

Information management of full-cycle inpatient bed reservation for cancer patients under the normalised prevention and control of the COVID-19 pandemic

BACKGROUND

During the prolonged COVID-19 pandemic, hospitals became focal points for normalised prevention and control. In this study, we investigated the feasibility of an inpatient bed reservation system for cancer patients that was developed in the department?s public WeChat account. We also explored its role in improving operational efficiency and nursing quality management, as well as in optimising nursing workforce deployment.

METHODS

We utilised WeChat to facilitate communication between cancer patients and health care professionals. Furthermore, we collected data on admissions, discharges, average number of hospitalisation days, bed utilisation rate, and the number of bed days occupied by hospitalised patients through the hospital information system and nurses? working hours and competency levels through the nurse scheduling system. The average nursing hours per patient per day were calculated. Through the inpatient bed reservation system, the number of accepted admissions, denied admissions, and cancelled admissions from the reservation system were collected. The impact of the bed reservation system on the department?s operational efficiency was analysed by comparing the number of hospitalisation discharges before and after reservations, as well as the average hospitalisation and bed utilisation rates. By comparing nurses? working hours per month and average nursing hours per patient per day, the system?s impact on nurses? working hours and nursing quality indicators was analysed.

RESULTS

The average hospitalisation length, bed utilisation rate, and nurses? working hours were significantly lower, and the average number of nursing hours per patient per day was significantly higher after the implementation of the reservation system. The full-cycle bed information management model for cancer patients did not affect the number of discharged patients.

CONCLUSION

Patients? ability to reserve bed types from home in advance using the department?s official WeChat-based inpatient bed reservation system allowed nurses to prepare for their work ahead of time. This in turn improved the operational efficiency of the department and nursing quality, and it optimised the deployment of the nursing workforce.

Integrating Clinical Data and Medical Imaging in Lung Cancer: Feasibility Study Using the Observational Medical Outcomes Partnership Common Data Model Extension

BACKGROUND

Digital transformation, particularly the integration of medical imaging with clinical data, is vital in personalized medicine. The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) standardizes health data. However, integrating medical imaging remains a challenge.

OBJECTIVE

This study proposes a method for combining medical imaging data with the OMOP CDM to improve multimodal research.

METHODS

Our approach included the analysis and selection of digital imaging and communications in medicine header tags, validation of data formats, and alignment according to the OMOP CDM framework. The Fast Healthcare Interoperability Resources ImagingStudy profile guided our consistency in column naming and definitions. Imaging Common Data Model (I-CDM), constructed using the entity-attribute-value model, facilitates scalable and efficient medical imaging data management. For patients with lung cancer diagnosed between 2010 and 2017, we introduced 4 new tables-IMAGING_STUDY, IMAGING_SERIES, IMAGING_ANNOTATION, and FILEPATH-to standardize various imaging-related data and link to clinical data.

RESULTS

This framework underscores the effectiveness of I-CDM in enhancing our understanding of lung cancer diagnostics and treatment strategies. The implementation of the I-CDM tables enabled the structured organization of a comprehensive data set, including 282,098 IMAGING_STUDY, 5,674,425 IMAGING_SERIES, and 48,536 IMAGING_ANNOTATION records, illustrating the extensive scope and depth of the approach. A scenario-based analysis using actual data from patients with lung cancer underscored the feasibility of our approach. A data quality check applying 44 specific rules confirmed the high integrity of the constructed data set, with all checks successfully passed, underscoring the reliability of our findings.

CONCLUSIONS

These findings indicate that I-CDM can improve the integration and analysis of medical imaging and clinical data. By addressing the challenges in data standardization and management, our approach contributes toward enhancing diagnostics and treatment strategies. Future research should expand the application of I-CDM to diverse disease populations and explore its wide-ranging utility for medical conditions.

Integrating Artificial Intelligence into Biomedical Science Curricula: Advancing Healthcare Education

The integration of artificial intelligence (AI) into healthcare practice has improved patient management and care. Many clinical laboratory specialties have already integrated AI in diagnostic specialties such as radiology and pathology, where it can assist in image analysis, diagnosis, and clinical reporting. As AI technologies continue to advance, it is crucial for biomedical science students to receive comprehensive education and training in AI concepts and applications and to understand the ethical consequences for such development. This review focus on the importance of integrating AI into biomedical science curricula and proposes strategies to enhance curricula for different specialties to prepare future healthcare workers. Improving the curriculum can be achieved by introducing specific subjects related to AI such as informatics, data sciences, and digital health. However, there are many challenges to enhancing the curriculum with AI. In this narrative review, we discuss these challenges and suggest mitigation strategies.

Development of a machine learning-based model for the prediction and progression of diabetic kidney disease: A single centred retrospective study

BACKGROUND

Diabetic kidney disease (DKD) is a diabetic microvascular complication often characterized by an unpredictable progression. Hence, early detection and recognition of patients vulnerable to progression is crucial.

OBJECTIVE

To develop a prediction model to identify the stages of DKD and the factors contributing to progression to each stage using machine learning.

METHODOLOGY

A retrospective study was conducted in a South Indian tertiary care hospital and collected the details of patients diagnosed with DKD from January 2017 to January 2022. Bayesian optimization-based machine learning techniques such as classification and regression were employed. The model was developed with the help of an optimization framework that effectively balances classification, prediction accuracy, and explainability.

RESULTS

Of the 311 patients diagnosed with DKD, 227 were selected for the study. A system for predicting DKD has been created for a patient dataset utilizing a variety of machine-learning approaches. The eXtreme gradient (XG) Boost method excelled, achieving 88.75% accuracy, 88.57% precision, 91.4% sensitivity,100% specificity, and 89.49% F1-score. An interpretable data-driven method highlights significant features for early DKD diagnosis. The best explainable prediction model uses the XG Boost classifier, revealing serum uric acid, urea, phosphorous, red blood cells, calcium, and absolute eosinophil count as the major predictors influencing the progression of DKD. In the case of regression models, the gradient boost regressor performed the best, with an R2 score of 0.97.

CONCLUSION

Machine learning algorithms can effectively predict the stages of DKD and thus help physicians in providing patients with personalized care at the right time.

Identification of key differentially expressed immune related genes in patients with persistent atrial fibrillation: an integrated bioinformation analysis

OBJECTIVE

We aimed to investigate key differentially expressed immune related genes in persistent atrial fibrillation.

METHODS

Gene expression profiles were downloaded from Gene Expression Omnibus (GEO) using "GEO query" package. "limma" package and "sva" package were used to conduct normalization and eliminate batch effects, respectively. We screened out differentially expressed genes (DEGs) based on "limma" package with the standard of |log fold change (FC)| ≥ 1.5 and false discovery rate (FDR) < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses of DEGs were performed by "clusterProfler" package. We further applied LASSO to select key DEGs, and intersected key DEGs with immune related genes from ImmPort database. The ROC curve of each DEIRG was constructed to evaluate its diagnostic efficiency for AF.

RESULTS

A total of 103 DEGs we were screened out, of them, 48 genes were down-regulated and 55 genes were up-regulated. Result of functional enrichment analysis show that, most of DEGs were related to immune response, inflammation, and oxidative stress. Ultimately, CYBB, RORB, S100A12, and CHGB were determined as key DEIRGs, each of which displayed a favor efficiency for diagnosing persistent AF.

CONCLUSION

CYBB, RORB, S100A12, and CHGB were identified as key DEIRGs in persistent AF, and future studies are needed to further explore the underlying roles of CYBB, RORB, S100A12, and CHGB in persistent AF.

Artificial intelligence empowering public health education: prospects and challenges

Artificial Intelligence (AI) is revolutionizing public health education through its capacity for intricate analysis of large-scale health datasets and the tailored dissemination of health-related information and interventions. This article conducts a profound exploration into the integration of AI within public health, accentuating its scientific foundations, prospective progress, and practical application scenarios. It underscores the transformative potential of AI in crafting individualized educational programs, developing sophisticated behavioral models, and informing the creation of health policies. The manuscript strives to thoroughly evaluate the extant landscape of AI applications in public health, scrutinizing critical challenges such as the propensity for data bias and the imperative of safeguarding privacy. By dissecting these issues, the article contributes to the conversation on how AI can be harnessed responsibly and effectively, ensuring that its application in public health education is both ethically grounded and equitable. The paper's significance is multifold: it aims to provide a blueprint for policy formulation, offer actionable insights for public health authorities, and catalyze the progression of health interventions toward increasingly sophisticated and precise approaches. Ultimately, this research anticipates fostering an environment where AI not only augments public health education but also does so with a steadfast commitment to the principles of justice and inclusivity, thereby elevating the standard and reach of health education initiatives globally.

Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review

PURPOSE OF REVIEW

Machine learning (ML) approaches are an emerging alternative for healthcare risk prediction. We aimed to synthesise the literature on ML and classical regression studies exploring potential prognostic factors and to compare prediction performance for pre-eclampsia.

RECENT FINDINGS

From 9382 studies retrieved, 82 were included. Sixty-six publications exclusively reported eighty-four classical regression models to predict variable timing of onset of pre-eclampsia. Another six publications reported purely ML algorithms, whilst another 10 publications reported ML algorithms and classical regression models in the same sample with 8 of 10 findings that ML algorithms outperformed classical regression models. The most frequent prognostic factors were age, pre-pregnancy body mass index, chronic medical conditions, parity, prior history of pre-eclampsia, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and pregnancy-associated plasma protein A. Top performing ML algorithms were random forest (area under the curve (AUC) = 0.94, 95% confidence interval (CI) 0.91-0.96) and extreme gradient boosting (AUC = 0.92, 95% CI 0.90-0.94). The competing risk model had similar performance (AUC = 0.92, 95% CI 0.91-0.92) compared with a neural network. Calibration performance was not reported in the majority of publications. ML algorithms had better performance compared to classical regression models in pre-eclampsia prediction. Random forest and boosting-type algorithms had the best prediction performance. Further research should focus on comparing ML algorithms to classical regression models using the same samples and evaluation metrics to gain insight into their performance. External validation of ML algorithms is warranted to gain insights into their generalisability.

Automated grading of anatomical objective structured practical examinations using decision trees: An artificial intelligence approach

An Objective Structured Practical Examination (OSPE) is an effective and robust, but resource-intensive, means of evaluating anatomical knowledge. Since most OSPEs employ short answer or fill-in-the-blank style questions, the format requires many people familiar with the content to mark the examinations. However, the increasing prevalence of online delivery for anatomy and physiology courses could result in students losing the OSPE practice that they would receive in face-to-face learning sessions. The purpose of this study was to test the accuracy of Decision Trees (DTs) in marking OSPE questions as a first step to creating an intelligent, online OSPE tutoring system. The study used the results of the winter 2020 semester final OSPE from McMaster University's anatomy and physiology course in the Faculty of Health Sciences (HTHSCI 2FF3/2LL3/1D06) as the data set. Ninety percent of the data set was used in a 10-fold validation algorithm to train a DT for each of the 54 questions. Each DT was comprised of unique words that appeared in correct, student-written answers. The remaining 10% of the data set was marked by the generated DTs. When the answers marked by the DT were compared to the answers marked by staff and faculty, the DT achieved an average accuracy of 94.49% across all 54 questions. This suggests that machine learning algorithms such as DTs are a highly effective option for OSPE grading and are suitable for the development of an intelligent, online OSPE tutoring system.

Perceptions of Surgeons in the Kurdistan Region of Iraq Regarding the Use of Artificial Intelligence

INTRODUCTION

Artificil intelligence (AI) is revolutionizing healthcare by seamlessly integrating into various aspects of human life. From robotic surgery to virtual nursing assistants and image analysis applications, AI is transforming the way we approach and deliver healthcare. By leveraging AI, patients can gain a deeper understanding of their symptoms, empowering them to make informed decisions about their health and ultimately improving their quality of life.  Methods: An online survey collected data from social media platforms regarding the surgeon society in the Kurdistan region of Iraq. All statistical analyses were carried out using IBM SPSS Statistics for Windows, Version 25 (Released 2017; IBM Corp., Armonk, New York).

RESULTS

A total of 316 surgeons responded to the survey. A significant majority believed that using artificial intelligence benefits patients, and a substantial number advocated for its avoidance as a matter of principle. More than half said that AI would always impact education, and half of the participants said that AI would always affect complication prediction.

CONCLUSIONS

This is the first study investigating surgeon attitudes and perceptions regarding the use of AI in the Kurdistan region. The surgeons who responded generally appreciated AI's use in their practice. Notably, general surgeons showed greater openness to integrating AI into their daily practices compared to those in other surgical specialties.

Machine learning algorithms for predicting COVID-19 mortality in Ethiopia

BACKGROUND

Coronavirus disease 2019 (COVID-19), a global public health crisis, continues to pose challenges despite preventive measures. The daily rise in COVID-19 cases is concerning, and the testing process is both time-consuming and costly. While several models have been created to predict mortality in COVID-19 patients, only a few have shown sufficient accuracy. Machine learning algorithms offer a promising approach to data-driven prediction of clinical outcomes, surpassing traditional statistical modeling. Leveraging machine learning (ML) algorithms could potentially provide a solution for predicting mortality in hospitalized COVID-19 patients in Ethiopia. Therefore, the aim of this study is to develop and validate machine-learning models for accurately predicting mortality in COVID-19 hospitalized patients in Ethiopia.

METHODS

Our study involved analyzing electronic medical records of COVID-19 patients who were admitted to public hospitals in Ethiopia. Specifically, we developed seven different machine learning models to predict COVID-19 patient mortality. These models included J48 decision tree, random forest (RF), k-nearest neighborhood (k-NN), multi-layer perceptron (MLP), Naïve Bayes (NB), eXtreme gradient boosting (XGBoost), and logistic regression (LR). We then compared the performance of these models using data from a cohort of 696 patients through statistical analysis. To evaluate the effectiveness of the models, we utilized metrics derived from the confusion matrix such as sensitivity, specificity, precision, and receiver operating characteristic (ROC).

RESULTS

The study included a total of 696 patients, with a higher number of females (440 patients, accounting for 63.2%) compared to males. The median age of the participants was 35.0 years old, with an interquartile range of 18-79. After conducting different feature selection procedures, 23 features were examined, and identified as predictors of mortality, and it was determined that gender, Intensive care unit (ICU) admission, and alcohol drinking/addiction were the top three predictors of COVID-19 mortality. On the other hand, loss of smell, loss of taste, and hypertension were identified as the three lowest predictors of COVID-19 mortality. The experimental results revealed that the k-nearest neighbor (k-NN) algorithm outperformed than other machine learning algorithms, achieving an accuracy of 95.25%, sensitivity of 95.30%, precision of 92.7%, specificity of 93.30%, F1 score 93.98% and a receiver operating characteristic (ROC) score of 96.90%. These findings highlight the effectiveness of the k-NN algorithm in predicting COVID-19 outcomes based on the selected features.

CONCLUSION

Our study has developed an innovative model that utilizes hospital data to accurately predict the mortality risk of COVID-19 patients. The main objective of this model is to prioritize early treatment for high-risk patients and optimize strained healthcare systems during the ongoing pandemic. By integrating machine learning with comprehensive hospital databases, our model effectively classifies patients' mortality risk, enabling targeted medical interventions and improved resource management. Among the various methods tested, the K-nearest neighbors (KNN) algorithm demonstrated the highest accuracy, allowing for early identification of high-risk patients. Through KNN feature identification, we identified 23 predictors that significantly contribute to predicting COVID-19 mortality. The top five predictors are gender (female), intensive care unit (ICU) admission, alcohol drinking, smoking, and symptoms of headache and chills. This advancement holds great promise in enhancing healthcare outcomes and decision-making during the pandemic. By providing services and prioritizing patients based on the identified predictors, healthcare facilities and providers can improve the chances of survival for individuals. This model provides valuable insights that can guide healthcare professionals in allocating resources and delivering appropriate care to those at highest risk.

Personalized prediction of mortality in patients with acute ischemic stroke using explainable artificial intelligence

BACKGROUND

Research into the acute kidney disease (AKD) after acute ischemic stroke (AIS) is rare, and how clinical features influence its prognosis remain unknown. We aim to employ interpretable machine learning (ML) models to study AIS and clarify its decision-making process in identifying the risk of mortality.

METHODS

We conducted a retrospective cohort study involving AIS patients from January 2020 to June 2021. Patient data were randomly divided into training and test sets. Eight ML algorithms were employed to construct predictive models for mortality. The performance of the best model was evaluated using various metrics. Furthermore, we created an artificial intelligence (AI)-driven web application that leveraged the top ten most crucial features for mortality prediction.

RESULTS

The study cohort consisted of 1633 AIS patients, among whom 257 (15.74%) developed subacute AKD, 173 (10.59%) experienced AKI recovery, and 65 (3.98%) met criteria for both AKI and AKD. The mortality rate stood at 4.84%. The LightGBM model displayed superior performance, boasting an AUROC of 0.96 for mortality prediction. The top five features linked to mortality were ACEI/ARE, renal function trajectories, neutrophil count, diuretics, and serum creatinine. Moreover, we designed a web application using the LightGBM model to estimate mortality risk.

CONCLUSIONS

Complete renal function trajectories, including AKI and AKD, are vital for fitting mortality in AIS patients. An interpretable ML model effectively clarified its decision-making process for identifying AIS patients at risk of mortality. The AI-driven web application has the potential to contribute to the development of personalized early mortality prevention.

Using diffusion models to generate synthetic labeled data for medical image segmentation

PURPOSE

Medical image analysis has become a prominent area where machine learning has been applied. However, high-quality, publicly available data are limited either due to patient privacy laws or the time and cost required for experts to annotate images. In this retrospective study, we designed and evaluated a pipeline to generate synthetic labeled polyp images for augmenting medical image segmentation models with the aim of reducing this data scarcity.

METHODS

We trained diffusion models on the HyperKvasir dataset, comprising 1000 images of polyps in the human GI tract from 2008 to 2016. Qualitative expert review, Fréchet Inception Distance (FID), and Multi-Scale Structural Similarity (MS-SSIM) were tested for evaluation. Additionally, various segmentation models were trained with the generated data and evaluated using Dice score (DS) and Intersection over Union (IoU).

RESULTS

Our pipeline produced images more akin to real polyp images based on FID scores. Segmentation model performance also showed improvements over GAN methods when trained entirely, or partially, with synthetic data, despite requiring less compute for training. Moreover, the improvement persists when tested on different datasets, showcasing the transferability of the generated images.

CONCLUSIONS

The proposed pipeline produced realistic image and mask pairs which could reduce the need for manual data annotation when performing a machine learning task. We support this use case by showing that the methods proposed in this study enhanced segmentation model performance, as measured by Dice and IoU scores, when trained fully or partially on synthetic data.

DeepLeish: a deep learning based support system for the detection of Leishmaniasis parasite from Giemsa-stained microscope images

BACKGROUND

Leishmaniasis is a vector-born neglected parasitic disease belonging to the genus Leishmania. Out of the 30 Leishmania species, 21 species cause human infection that affect the skin and the internal organs. Around, 700,000 to 1,000,000 of the newly infected cases and 26,000 to 65,000 deaths are reported worldwide annually. The disease exhibits three clinical presentations, namely, the cutaneous, muco-cutaneous and visceral Leishmaniasis which affects the skin, mucosal membrane and the internal organs, respectively. The relapsing behavior of the disease limits its diagnosis and treatment efficiency. The common diagnostic approaches follow subjective, error-prone, repetitive processes. Despite, an ever pressing need for an accurate detection of Leishmaniasis, the research conducted so far is scarce. In this regard, the main aim of the current research is to develop an artificial intelligence based detection tool for the Leishmaniasis from the Geimsa-stained microscopic images using deep learning method.

METHODS

Stained microscopic images were acquired locally and labeled by experts. The images were augmented using different methods to prevent overfitting and improve the generalizability of the system. Fine-tuned Faster RCNN, SSD, and YOLOV5 models were used for object detection. Mean average precision (MAP), precision, and Recall were calculated to evaluate and compare the performance of the models.

RESULTS

The fine-tuned YOLOV5 outperformed the other models such as Faster RCNN and SSD, with the MAP scores, of 73%, 54% and 57%, respectively.

CONCLUSION

The currently developed YOLOV5 model can be tested in the clinics to assist the laboratorists in diagnosing Leishmaniasis from the microscopic images. Particularly, in low-resourced healthcare facilities, with fewer qualified medical professionals or hematologists, our AI support system can assist in reducing the diagnosing time, workload, and misdiagnosis. Furthermore, the dataset collected by us will be shared with other researchers who seek to improve upon the detection system of the parasite. The current model detects the parasites even in the presence of the monocyte cells, but sometimes, the accuracy decreases due to the differences in the sizes of the parasite cells alongside the blood cells. The incorporation of cascaded networks in future and the quantification of the parasite load, shall overcome the limitations of the currently developed system.

Smart solutions in hypertension diagnosis and management: a deep dive into artificial intelligence and modern wearables for blood pressure monitoring

Hypertension, a widespread cardiovascular issue, presents a major global health challenge. Traditional diagnosis and treatment methods involve periodic blood pressure monitoring and prescribing antihypertensive drugs. Smart technology integration in healthcare offers promising results in optimizing the diagnosis and treatment of various conditions. We investigate its role in improving hypertension diagnosis and treatment effectiveness using machine learning algorithms for early and accurate detection. Intelligent models trained on diverse datasets (encompassing physiological parameters, lifestyle factors, and genetic information) to detect subtle hypertension risk patterns. Adaptive algorithms analyze patient-specific data, optimizing treatment plans based on medication responses and lifestyle habits. This personalized approach ensures effective, minimally invasive interventions tailored to each patient. Wearables and smart sensors provide real-time health insights for proactive treatment adjustments and early complication detection.

Artificial Intelligence in Otology and Neurotology

Clinical applications of artificial intelligence (AI) have grown exponentially with increasing computational power and Big Data. Data rich fields such as Otology and Neurotology are still in the infancy of harnessing the power of AI but are increasingly involved in training and developing ways to incorporate AI into patient care. Current studies involving AI are focused on accessible datasets; health care wearables, tabular data from electronic medical records, electrophysiologic measurements, imaging, and "omics" provide huge amounts of data to utilize. Health care wearables, such as hearing aids and cochlear implants, are a ripe environment for AI implementation.

Tissue clearing and imaging approaches for in toto analysis of the reproductive system†

New microscopy techniques in combination with tissue clearing protocols and emerging analytical approaches have presented researchers with the tools to understand dynamic biological processes in a three-dimensional context. This paves the road for the exploration of new research questions in reproductive biology, for which previous techniques have provided only approximate resolution. These new methodologies now allow for contextualized analysis of far-larger volumes than was previously possible. Tissue optical clearing and three-dimensional imaging techniques posit the bridging of molecular mechanisms, macroscopic morphogenic development, and maintenance of reproductive function into one cohesive and comprehensive understanding of the biology of the reproductive system. In this review, we present a survey of the various tissue clearing techniques and imaging systems, as they have been applied to the developing and adult reproductive system. We provide an overview of tools available for analysis of experimental data, giving particular attention to the emergence of artificial intelligence-assisted methods and their applicability to image analysis. We conclude with an evaluation of how novel image analysis approaches that have been applied to other organ systems could be incorporated into future experimental evaluation of reproductive biology.

Evaluating the readiness of healthcare administration students to utilize AI for sustainable leadership: a survey study

PURPOSE

This paper explores how healthcare administration students perceive the integration of Artificial Intelligence (AI) in healthcare leadership, mainly focusing on the sustainability aspects involved. It aims to identify gaps in current educational curricula and suggests enhancements to better prepare future healthcare professionals for the evolving demands of AI-driven healthcare environments.

DESIGN/METHODOLOGY/APPROACH

This study utilized a cross-sectional survey design to understand healthcare administration students' perceptions regarding integrating AI in healthcare leadership. An online questionnaire, developed from an extensive literature review covering fundamental AI knowledge and its role in sustainable leadership, was distributed to students majoring and minoring in healthcare administration. This methodological approach garnered participation from 62 students, providing insights and perspectives crucial for the study's objectives.

FINDINGS

The research revealed that while a significant majority of healthcare administration students (70%) recognize the potential of AI in fostering sustainable leadership in healthcare, only 30% feel adequately prepared to work in AI-integrated environments. Additionally, students were interested in learning more about AI applications in healthcare and the role of AI in sustainable leadership, underscoring the need for comprehensive AI-focused education in their curriculum.

RESEARCH LIMITATIONS/IMPLICATIONS

The research is limited by its focus on a single academic institution, which may not fully represent the diversity of perspectives in healthcare administration.

PRACTICAL IMPLICATIONS

This study highlights the need for healthcare administration curricula to incorporate AI education, aligning theoretical knowledge with practical applications, to effectively prepare future professionals for the evolving demands of AI-integrated healthcare environments.

ORIGINALITY/VALUE

This research paper presents insights into healthcare administration students' readiness and perspectives toward AI integration in healthcare leadership, filling a critical gap in understanding the educational needs in the evolving landscape of AI-driven healthcare.

Class imbalance on medical image classification: towards better evaluation practices for discrimination and calibration performance

PURPOSE

This work aims to assess standard evaluation practices used by the research community for evaluating medical imaging classifiers, with a specific focus on the implications of class imbalance. The analysis is performed on chest X-rays as a case study and encompasses a comprehensive model performance definition, considering both discriminative capabilities and model calibration.

MATERIALS AND METHODS

We conduct a concise literature review to examine prevailing scientific practices used when evaluating X-ray classifiers. Then, we perform a systematic experiment on two major chest X-ray datasets to showcase a didactic example of the behavior of several performance metrics under different class ratios and highlight how widely adopted metrics can conceal performance in the minority class.

RESULTS

Our literature study confirms that: (1) even when dealing with highly imbalanced datasets, the community tends to use metrics that are dominated by the majority class; and (2) it is still uncommon to include calibration studies for chest X-ray classifiers, albeit its importance in the context of healthcare. Moreover, our systematic experiments confirm that current evaluation practices may not reflect model performance in real clinical scenarios and suggest complementary metrics to better reflect the performance of the system in such scenarios.

CONCLUSION

Our analysis underscores the need for enhanced evaluation practices, particularly in the context of class-imbalanced chest X-ray classifiers. We recommend the inclusion of complementary metrics such as the area under the precision-recall curve (AUC-PR), adjusted AUC-PR, and balanced Brier score, to offer a more accurate depiction of system performance in real clinical scenarios, considering metrics that reflect both, discrimination and calibration performance.

CLINICAL RELEVANCE STATEMENT

This study underscores the critical need for refined evaluation metrics in medical imaging classifiers, emphasizing that prevalent metrics may mask poor performance in minority classes, potentially impacting clinical diagnoses and healthcare outcomes.

KEY POINTS

Common scientific practices in papers dealing with X-ray computer-assisted diagnosis (CAD) systems may be misleading. We highlight limitations in reporting of evaluation metrics for X-ray CAD systems in highly imbalanced scenarios. We propose adopting alternative metrics based on experimental evaluation on large-scale datasets.

Computer Vision Foundation Models in Endoscopy: Proof of Concept in Oropharyngeal Cancer

OBJECTIVES

To evaluate the performance of vision transformer-derived image embeddings for distinguishing between normal and neoplastic tissues in the oropharynx and to investigate the potential of computer vision (CV) foundation models in medical imaging.

METHODS

Computational study using endoscopic frames with a focus on the application of a self-supervised vision transformer model (DINOv2) for tissue classification. High-definition endoscopic images were used to extract image patches that were then normalized and processed using the DINOv2 model to obtain embeddings. These embeddings served as input for a standard support vector machine (SVM) to classify the tissues as neoplastic or normal. The model's discriminative performance was validated using an 80-20 train-validation split.

RESULTS

From 38 endoscopic NBI videos, 327 image patches were analyzed. The classification results in the validation cohort demonstrated high accuracy (92%) and precision (89%), with a perfect recall (100%) and an F1-score of 94%. The receiver operating characteristic (ROC) curve yielded an area under the curve (AUC) of 0.96.

CONCLUSION

The use of large vision model-derived embeddings effectively differentiated between neoplastic and normal oropharyngeal tissues. This study supports the feasibility of employing CV foundation models like DINOv2 in the endoscopic evaluation of mucosal lesions, potentially augmenting diagnostic precision in Otorhinolaryngology.

LEVEL OF EVIDENCE

4 Laryngoscope, 2024.

Towards regulatory generative AI in ophthalmology healthcare: a security and privacy perspective

As the healthcare community increasingly harnesses the power of generative artificial intelligence (AI), critical issues of security, privacy and regulation take centre stage. In this paper, we explore the security and privacy risks of generative AI from model-level and data-level perspectives. Moreover, we elucidate the potential consequences and case studies within the domain of ophthalmology. Model-level risks include knowledge leakage from the model and model safety under AI-specific attacks, while data-level risks involve unauthorised data collection and data accuracy concerns. Within the healthcare context, these risks can bear severe consequences, encompassing potential breaches of sensitive information, violating privacy rights and threats to patient safety. This paper not only highlights these challenges but also elucidates governance-driven solutions that adhere to AI and healthcare regulations. We advocate for preparedness against potential threats, call for transparency enhancements and underscore the necessity of clinical validation before real-world implementation. The objective of security and privacy improvement in generative AI warrants emphasising the role of ophthalmologists and other healthcare providers, and the timely introduction of comprehensive regulations.

Application of machine learning techniques in population pharmacokinetics/pharmacodynamics modeling

Population pharmacokinetics/pharmacodynamics (pop-PK/PD) consolidates pharmacokinetic and pharmacodynamic data from many subjects to understand inter- and intra-individual variability due to patient backgrounds, including disease state and genetics. The typical workflow in pop-PK/PD analysis involves the determination of the structure model, selection of the error model, analysis based on the base model, covariate modeling, and validation of the final model. Machine learning is gaining considerable attention in the medical and various fields because, in contrast to traditional modeling, which often assumes linear or predefined relationships, machine learning modeling learns directly from data and accommodates complex patterns. Machine learning has demonstrated excellent capabilities for prescreening covariates and developing predictive models. This review introduces various applications of machine learning techniques in pop-PK/PD research.

A Klotho-Based Machine Learning Model for Prediction of both Kidney and Cardiovascular Outcomes in Chronic Kidney Disease

Introduction

This study aimed to develop and validate machine learning (ML) models based on serum Klotho for predicting end-stage kidney disease (ESKD) and cardiovascular disease (CVD) in patients with chronic kidney disease (CKD).

Methods

Five different ML models were trained to predict the risk of ESKD and CVD at three different time points (3, 5, and 8 years) using a cohort of 400 non-dialysis CKD patients. The dataset was divided into a training set (70%) and an internal validation set (30%). These models were informed by data comprising 47 clinical features, including serum Klotho. The best-performing model was selected and used to identify risk factors for each outcome. Model performance was assessed using various metrics.

Results

The findings showed that the least absolute shrinkage and selection operator regression model had the highest accuracy (C-index = 0.71) in predicting ESKD. The features mainly included in this model were estimated glomerular filtration rate, 24-h urinary microalbumin, serum albumin, phosphate, parathyroid hormone, and serum Klotho, which achieved the highest area under the curve (AUC) of 0.930 (95% CI: 0.897-0.962). In addition, for the CVD risk prediction, the random survival forest model with the highest accuracy (C-index = 0.66) was selected and achieved the highest AUC of 0.782 (95% CI: 0.633-0.930). The features mainly included in this model were age, history of primary hypertension, calcium, tumor necrosis factor-alpha, and serum Klotho.

Conclusion

We successfully developed and validated Klotho-based ML risk prediction models for CVD and ESKD in CKD patients with good performance, indicating their high clinical utility.

Application of Machine Learning for Classification of Brain Tumors: A Systematic Review and Meta-Analysis

BACKGROUND

Classifying brain tumors accurately is crucial for treatment and prognosis. Machine learning (ML) shows great promise in improving tumor classification accuracy. This study evaluates ML algorithms for differentiating various brain tumor types.

METHODS

A systematic review and meta-analysis were conducted, searching PubMed, Embase, and Web of Science up to March 14, 2023. Studies that only investigated image segmentation accuracy or brain tumor detection instead of classification were excluded. We extracted binary diagnostic accuracy data, constructing contingency tables to derive sensitivity and specificity.

RESULTS

Fifty-one studies were included. The pooled area under the curve for glioblastoma versus lymphoma and low-grade versus high-grade gliomas were 0.99 (95% confidence interval [CI]: 0.98-1.00) and 0.89, respectively. The pooled sensitivity and specificity for benign versus malignant tumors were 0.90 (95% CI: 0.85-0.93) and 0.93 (95% CI: 0.90-0.95), respectively. The pooled sensitivity and specificity for low-grade versus high-grade gliomas were 0.99 (95% CI: 0.97-1.00) and 0.94, (95% CI: 0.79-0.99), respectively. Primary versus metastatic tumor identification yields sensitivity and specificity of 0.89, (95% CI: 0.83-0.93) and 0.87 (95% CI: 0.82-0.91), correspondingly. The differentiation of gliomas from pituitary tumors yielded the highest results among primary brain tumor classifications: sensitivity of 0.99 (95% CI: 0.99-1.00) and specificity of 0.99 (95% CI: 0.98-1.00).

CONCLUSIONS

ML demonstrated excellent performance in classifying brain tumor images, with near-maximum area under the curves, sensitivity, and specificity.

A comparative study of English and Japanese ChatGPT responses to anaesthesia-related medical questions

Background

The expansion of artificial intelligence (AI) within large language models (LLMs) has the potential to streamline healthcare delivery. Despite the increased use of LLMs, disparities in their performance particularly in different languages, remain underexplored. This study examines the quality of ChatGPT responses in English and Japanese, specifically to questions related to anaesthesiology.

Methods

Anaesthesiologists proficient in both languages were recruited as experts in this study. Ten frequently asked questions in anaesthesia were selected and translated for evaluation. Three non-sequential responses from ChatGPT were assessed for content quality (accuracy, comprehensiveness, and safety) and communication quality (understanding, empathy/tone, and ethics) by expert evaluators.

Results

Eight anaesthesiologists evaluated English and Japanese LLM responses. The overall quality for all questions combined was higher in English compared with Japanese responses. Content and communication quality were significantly higher in English compared with Japanese LLMs responses (both P<0.001) in all three responses. Comprehensiveness, safety, and understanding were higher scores in English LLM responses. In all three responses, more than half of the evaluators marked overall English responses as better than Japanese responses.

Conclusions

English LLM responses to anaesthesia-related frequently asked questions were superior in quality to Japanese responses when assessed by bilingual anaesthesia experts in this report. This study highlights the potential for language-related disparities in healthcare information and the need to improve the quality of AI responses in underrepresented languages. Future studies are needed to explore these disparities in other commonly spoken languages and to compare the performance of different LLMs.

The Role of Artificial Intelligence in Healthcare: Enhancing Coronary Computed Tomography Angiography for Coronary Artery Disease Management

This review aims to explore the potential of artificial intelligence (AI) in coronary CT angiography (CCTA), a key tool for diagnosing coronary artery disease (CAD). Because CAD is still a major cause of death worldwide, effective and accurate diagnostic methods are required to identify and manage the condition. CCTA certainly is a noninvasive alternative for diagnosing CAD, but it requires a large amount of data as input. We intend to discuss the idea of incorporating AI into CCTA, which enhances its diagnostic accuracy and operational efficiency. Using such AI technologies as machine learning (ML) and deep learning (DL) tools, CCTA images are automated to perfection and the analysis is significantly refined. It enables the characterization of a plaque, assesses the severity of the stenosis, and makes more accurate risk stratifications than traditional methods, with pinpoint accuracy. Automating routine tasks through AI-driven CCTA will reduce the radiologists' workload considerably, which is a standard benefit of such technologies. More importantly, it would enable radiologists to allocate more time and expertise to complex cases, thereby improving overall patient care. However, the field of AI in CCTA is not without its challenges, which include data protection, algorithm transparency, as well as criteria for standardization encoding. Despite such obstacles, it appears that the integration of AI technology into CCTA in the future holds great promise for keeping CAD itself in check, thereby aiding the fight against this disease and begetting better clinical outcomes and more optimized modes of healthcare. Future research on AI algorithms for CCTA, making ethical use of AI, and thereby overcoming the technical and clinical barriers to widespread adoption of this new tool, will hopefully pave the way for profound AI-driven transformations in healthcare.

Broadening Perspectives of Artificial Intelligence in Echocardiography

Echocardiography frequently serves as the first-line treatment of diagnostic imaging for several pathological entities in cardiology. Artificial intelligence (AI) has been growing substantially in information technology and various commercial industries. Machine learning (ML), a branch of AI, has been shown to expand the capabilities and potential of echocardiography. ML algorithms expand the field of echocardiography by automated assessment of the ejection fraction and left ventricular function, integrating novel approaches such as speckle tracking or tissue Doppler echocardiography or vector flow mapping, improved phenotyping, distinguishing between cardiac conditions, and incorporating information from mobile health and genomics. In this review article, we assess the impact of AI and ML in echocardiography.

Leveraging technological advances to assess dyadic visual cognition during infancy in high- and low-resource settings

Caregiver-infant interactions shape infants' early visual experience; however, there is limited work from low-and middle-income countries (LMIC) in characterizing the visual cognitive dynamics of these interactions. Here, we present an innovative dyadic visual cognition pipeline using machine learning methods which captures, processes, and analyses the visual dynamics of caregiver-infant interactions across cultures. We undertake two studies to examine its application in both low (rural India) and high (urban UK) resource settings. Study 1 develops and validates the pipeline to process caregiver-infant interaction data captured using head-mounted cameras and eye-trackers. We use face detection and object recognition networks and validate these tools using 12 caregiver-infant dyads (4 dyads from a 6-month-old UK cohort, 4 dyads from a 6-month-old India cohort, and 4 dyads from a 9-month-old India cohort). Results show robust and accurate face and toy detection, as well as a high percent agreement between processed and manually coded dyadic interactions. Study 2 applied the pipeline to a larger data set (25 6-month-olds from the UK, 31 6-month-olds from India, and 37 9-month-olds from India) with the aim of comparing the visual dynamics of caregiver-infant interaction across the two cultural settings. Results show remarkable correspondence between key measures of visual exploration across cultures, including longer mean look durations during infant-led joint attention episodes. In addition, we found several differences across cultures. Most notably, infants in the UK had a higher proportion of infant-led joint attention episodes consistent with a child-centered view of parenting common in western middle-class families. In summary, the pipeline we report provides an objective assessment tool to quantify the visual dynamics of caregiver-infant interaction across high- and low-resource settings.

Semi-automated seizure detection using interpretable machine learning models

Despite the vast number of seizure detection publications there are no validated open-source tools for automating seizure detection based on electrographic recordings. Researchers instead rely on manual curation of seizure detection that is highly laborious, inefficient, error prone, and heavily biased. Here we developed an open-source software called SeizyML that uses sensitive machine learning models coupled with manual validation of detected events reducing bias and promoting efficient and accurate detection of electrographic seizures. We compared the validity of four interpretable machine learning models (decision tree, gaussian naïve bayes, passive aggressive classifier, and stochastic gradient descent classifier) on an extensive electrographic seizure dataset that we collected from chronically epileptic mice. We find that the gaussian naïve bayes and stochastic gradient descent models achieved the highest precision and f1 scores, while also detecting all seizures in our mouse dataset and only require a small amount of data to train the model and achieve good performance. Further, we demonstrate the utility of this approach to detect electrographic seizures in a human EEG dataset. This approach has the potential to be a transformative research tool overcoming the analysis bottleneck that slows research progress.