Low-cost and convenient screening of disease using analysis of physical measurements and recordings.

In recent years, there has been substantial work in low-cost medical diagnostics based on the physical manifestations of disease. This is due to advancements in data analysis techniques and classification algorithms and the increased availability of computing power through smart devices. Smartphones and their ability to interface with simple sensors such as inertial measurement units (IMUs), microphones, piezoelectric sensors, etc., or with convenient attachments such as lenses have revolutionized the ability collect medically relevant data easily. Even if the data has relatively low resolution or signal to noise ratio, newer algorithms have made it possible to identify disease with this data. Many low-cost diagnostic tools have been created in medical fields spanning from neurology to dermatology to obstetrics. These tools are particularly useful in low-resource areas where access to expensive diagnostic equipment may not be possible. The ultimate goal would be the creation of a "diagnostic toolkit" consisting of a smartphone and a set of sensors and attachments that can be used to screen for a wide set of diseases in a community healthcare setting. However, there are a few concerns that still need to be overcome in low-cost diagnostics: lack of incentives to bring these devices to market, algorithmic bias, "black box" nature of the algorithms, and data storage/transfer concerns.

Assessment of Clinical Metadata on the Accuracy of Retinal Fundus Image Labels in Diabetic Retinopathy in Uganda: Case-Crossover Study Using the Multimodal Database of Retinal Images in Africa.

BACKGROUND: Labeling color fundus photos (CFP) is an important step in the development of artificial intelligence screening algorithms for the detection of diabetic retinopathy (DR). Most studies use the International Classification of Diabetic Retinopathy (ICDR) to assign labels to CFP, plus the presence or absence of macular edema (ME). Images can be grouped as referrable or nonreferrable according to these classifications. There is little guidance in the literature about how to collect and use metadata as a part of the CFP labeling process. OBJECTIVE: This study aimed to improve the quality of the Multimodal Database of Retinal Images in Africa (MoDRIA) by determining whether the availability of metadata during the image labeling process influences the accuracy, sensitivity, and specificity of image labels. MoDRIA was developed as one of the inaugural research projects of the Mbarara University Data Science Research Hub, part of the Data Science for Health Discovery and Innovation in Africa (DS-I Africa) initiative. METHODS: This is a crossover assessment with 2 groups and 2 phases. Each group had 10 randomly assigned labelers who provided an ICDR score and the presence or absence of ME for each of the 50 CFP in a test image with and without metadata including blood pressure, visual acuity, glucose, and medical history. Specificity and sensitivity of referable retinopathy were based on ICDR scores, and ME was calculated using a 2-sided t test. Comparison of sensitivity and specificity for ICDR scores and ME with and without metadata for each participant was calculated using the Wilcoxon signed rank test. Statistical significance was set at P<.05. RESULTS: The sensitivity for identifying referrable DR with metadata was 92.8% (95% CI 87.6-98.0) compared with 93.3% (95% CI 87.6-98.9) without metadata, and the specificity was 84.9% (95% CI 75.1-94.6) with metadata compared with 88.2% (95% CI 79.5-96.8) without metadata. The sensitivity for identifying the presence of ME was 64.3% (95% CI 57.6-71.0) with metadata, compared with 63.1% (95% CI 53.4-73.0) without metadata, and the specificity was 86.5% (95% CI 81.4-91.5) with metadata compared with 87.7% (95% CI 83.9-91.5) without metadata. The sensitivity and specificity of the ICDR score and the presence or absence of ME were calculated for each labeler with and without metadata. No findings were statistically significant. CONCLUSIONS: The sensitivity and specificity scores for the detection of referrable DR were slightly better without metadata, but the difference was not statistically significant. We cannot make definitive conclusions about the impact of metadata on the sensitivity and specificity of image labels in our study. Given the importance of metadata in clinical situations, we believe that metadata may benefit labeling quality. A more rigorous study to determine the sensitivity and specificity of CFP labels with and without metadata is recommended.

A Clinician's Guide to Understanding Bias in Critical Clinical Prediction Models.

This narrative review focuses on the role of clinical prediction models in supporting informed decision-making in critical care, emphasizing their 2 forms: traditional scores and artificial intelligence (AI)-based models. Acknowledging the potential for both types to embed biases, the authors underscore the importance of critical appraisal to increase our trust in models. The authors outline recommendations and critical care examples to manage risk of bias in AI models. The authors advocate for enhanced interdisciplinary training for clinicians, who are encouraged to explore various resources (books, journals, news Web sites, and social media) and events (Datathons) to deepen their understanding of risk of bias.

Assessment of fluid responsiveness using pulse pressure variation, stroke volume variation, plethysmographic variability index, central venous pressure, and inferior vena cava variation in patients undergoing mechanical ventilation: a systematic review and meta-analysis.

IMPORTANCE: Maneuvers assessing fluid responsiveness before an intravascular volume expansion may limit useless fluid administration, which in turn may improve outcomes. OBJECTIVE: To describe maneuvers for assessing fluid responsiveness in mechanically ventilated patients. REGISTRATION: The protocol was registered at PROSPERO: CRD42019146781. INFORMATION SOURCES AND SEARCH: PubMed, EMBASE, CINAHL, SCOPUS, and Web of Science were search from inception to 08/08/2023. STUDY SELECTION AND DATA COLLECTION: Prospective and intervention studies were selected. STATISTICAL ANALYSIS: Data for each maneuver were reported individually and data from the five most employed maneuvers were aggregated. A traditional and a Bayesian meta-analysis approach were performed. RESULTS: A total of 69 studies, encompassing 3185 fluid challenges and 2711 patients were analyzed. The prevalence of fluid responsiveness was 49.9%. Pulse pressure variation (PPV) was studied in 40 studies, mean threshold with 95% confidence intervals (95% CI) = 11.5 (10.5-12.4)%, and area under the receiver operating characteristics curve (AUC) with 95% CI was 0.87 (0.84-0.90). Stroke volume variation (SVV) was studied in 24 studies, mean threshold with 95% CI = 12.1 (10.9-13.3)%, and AUC with 95% CI was 0.87 (0.84-0.91). The plethysmographic variability index (PVI) was studied in 17 studies, mean threshold = 13.8 (12.3-15.3)%, and AUC was 0.88 (0.82-0.94). Central venous pressure (CVP) was studied in 12 studies, mean threshold with 95% CI = 9.0 (7.7-10.1) mmHg, and AUC with 95% CI was 0.77 (0.69-0.87). Inferior vena cava variation (∆IVC) was studied in 8 studies, mean threshold = 15.4 (13.3-17.6)%, and AUC with 95% CI was 0.83 (0.78-0.89). CONCLUSIONS: Fluid responsiveness can be reliably assessed in adult patients under mechanical ventilation. Among the five maneuvers compared in predicting fluid responsiveness, PPV, SVV, and PVI were superior to CVP and ∆IVC. However, there is no data supporting any of the above mentioned as being the best maneuver. Additionally, other well-established tests, such as the passive leg raising test, end-expiratory occlusion test, and tidal volume challenge, are also reliable.

Applied artificial intelligence for global child health: Addressing biases and barriers.

Given the potential benefits of artificial intelligence and machine learning (AI/ML) within healthcare, it is critical to consider how these technologies can be deployed in pediatric research and practice. Currently, healthcare AI/ML has not yet adapted to the specific technical considerations related to pediatric data nor adequately addressed the specific vulnerabilities of children and young people (CYP) in relation to AI. While the greatest burden of disease in CYP is firmly concentrated in lower and middle-income countries (LMICs), existing applied pediatric AI/ML efforts are concentrated in a small number of high-income countries (HICs). In LMICs, use-cases remain primarily in the proof-of-concept stage. This narrative review identifies a number of intersecting challenges that pose barriers to effective AI/ML for CYP globally and explores the shifts needed to make progress across multiple domains. Child-specific technical considerations throughout the AI/ML lifecycle have been largely overlooked thus far, yet these can be critical to model effectiveness. Governance concerns are paramount, with suitable national and international frameworks and guidance required to enable the safe and responsible deployment of advanced technologies impacting the care of CYP and using their data. An ambitious vision for child health demands that the potential benefits of AI/Ml are realized universally through greater international collaboration, capacity building, strong oversight, and ultimately diffusing the AI/ML locus of power to empower researchers and clinicians globally. In order that AI/ML systems that do not exacerbate inequalities in pediatric care, teams researching and developing these technologies in LMICs must ensure that AI/ML research is inclusive of the needs and concerns of CYP and their caregivers. A broad, interdisciplinary, and human-centered approach to AI/ML is essential for developing tools for healthcare workers delivering care, such that the creation and deployment of ML is grounded in local systems, cultures, and clinical practice. Decisions to invest in developing and testing pediatric AI/ML in resource-constrained settings must always be part of a broader evaluation of the overall needs of a healthcare system, considering the critical building blocks underpinning effective, sustainable, and cost-efficient healthcare delivery for CYP.

A health equity monitoring framework based on process mining.

In the United States, there is a proposal to link hospital Medicare payments with health equity measures, signaling a need to precisely measure equity in healthcare delivery. Despite significant research demonstrating disparities in health care outcomes and access, there is a noticeable gap in tools available to assess health equity across various health conditions and treatments. The available tools often focus on a single area of patient care, such as medication delivery, but fail to examine the entire health care process. The objective of this study is to propose a process mining framework to provide a comprehensive view of health equity. Using event logs which track all actions during patient care, this method allows us to look at disparities in single and multiple treatment steps, but also in the broader strategy of treatment delivery. We have applied this framework to the management of patients with sepsis in the Intensive Care Unit (ICU), focusing on sex and English language proficiency. We found no significant differences between treatments of male and female patients. However, for patients who don't speak English, there was a notable delay in starting their treatment, even though their illness was just as severe and subsequent treatments were similar. This framework subsumes existing individual approaches to measure health inequities and offers a comprehensive approach to pinpoint and delve into healthcare disparities, providing a valuable tool for research and policy-making aiming at more equitable healthcare.

The TRIPOD-LLM Statement: A Targeted Guideline For Reporting Large Language Models Use.

Large Language Models (LLMs) are rapidly being adopted in healthcare, necessitating standardized reporting guidelines. We present TRIPOD-LLM, an extension of the TRIPOD+AI statement, addressing the unique challenges of LLMs in biomedical applications. TRIPOD-LLM provides a comprehensive checklist of 19 main items and 50 subitems, covering key aspects from title to discussion. The guidelines introduce a modular format accommodating various LLM research designs and tasks, with 14 main items and 32 subitems applicable across all categories. Developed through an expedited Delphi process and expert consensus, TRIPOD-LLM emphasizes transparency, human oversight, and task-specific performance reporting. We also introduce an interactive website ( https://tripod-llm.vercel.app/ ) facilitating easy guideline completion and PDF generation for submission. As a living document, TRIPOD-LLM will evolve with the field, aiming to enhance the quality, reproducibility, and clinical applicability of LLM research in healthcare through comprehensive reporting. COI: DSB: Editorial, unrelated to this work: Associate Editor of Radiation Oncology, HemOnc.org (no financial compensation); Research funding, unrelated to this work: American Association for Cancer Research; Advisory and consulting, unrelated to this work: MercurialAI. DDF: Editorial, unrelated to this work: Associate Editor of JAMIA, Editorial Board of Scientific Data, Nature; Funding, unrelated to this work: the intramural research program at the U.S. National Library of Medicine, National Institutes of Health. JWG: Editorial, unrelated to this work: Editorial Board of Radiology: Artificial Intelligence, British Journal of Radiology AI journal and NEJM AI. All other authors declare no conflicts of interest.

Competing interests: digital health and indigenous data sovereignty.

Digital health is increasingly promoting open health data. Although this open approach promises a number of benefits, it also leads to tensions with Indigenous data sovereignty movements led by Indigenous peoples around the world who are asserting control over the use of health data as a part of self-determination. Digital health has a role in improving access to services and delivering improved health outcomes for Indigenous communities. However, we argue that in order to be effective and ethical, it is essential that the field engages more with Indigenous peoples´ rights and interests. We discuss challenges and possible improvements for data acquisition, management, analysis, and integration as they pertain to the health of Indigenous communities around the world.

Diversity and inclusion: A hidden additional benefit of Open Data.

The recent imperative by the National Institutes of Health to share scientific data publicly underscores a significant shift in academic research. Effective as of January 2023, it emphasizes that transparency in data collection and dedicated efforts towards data sharing are prerequisites for translational research, from the lab to the bedside. Given the role of data access in mitigating potential bias in clinical models, we hypothesize that researchers who leverage open-access datasets rather than privately-owned ones are more diverse. In this brief report, we proposed to test this hypothesis in the transdisciplinary and expanding field of artificial intelligence (AI) for critical care. Specifically, we compared the diversity among authors of publications leveraging open datasets, such as the commonly used MIMIC and eICU databases, with that among authors of publications relying exclusively on private datasets, unavailable to other research investigators (e.g., electronic health records from ICU patients accessible only to Mayo Clinic analysts). To measure the extent of author diversity, we characterized gender balance as well as the presence of researchers from low- and middle-income countries (LMIC) and minority-serving institutions (MSI) located in the United States (US). Our comparative analysis revealed a greater contribution of authors from LMICs and MSIs among researchers leveraging open critical care datasets (treatment group) than among those relying exclusively on private data resources (control group). The participation of women was similar between the two groups, albeit slightly larger in the former. Notably, although over 70% of all articles included at least one author inferred to be a woman, less than 25% had a woman as a first or last author. Importantly, we found that the proportion of authors from LMICs was substantially higher in the treatment than in the control group (10.1% vs. 6.2%, p<0.001), including as first and last authors. Moreover, we found that the proportion of US-based authors affiliated with a MSI was 1.5 times higher among articles in the treatment than in the control group, suggesting that open data resources attract a larger pool of participants from minority groups (8.6% vs. 5.6%, p<0.001). Thus, our study highlights the valuable contribution of the Open Data strategy to underrepresented groups, while also quantifying persisting gender gaps in academic and clinical research at the intersection of computer science and healthcare. In doing so, we hope our work points to the importance of extending open data practices in deliberate and systematic ways.

Lost in Transplantation: Characterizing Racial Gaps in Physician Organ Offer Acceptance.

Background: There are known racial disparities in the organ transplant allocation system in the United States. However, prior work has yet to establish if transplant center decisions on offer acceptance-the final step in the allocation process-contribute to these disparities. Objective: To estimate racial differences in the acceptance of organ offers by transplant center physicians on behalf of their patients. Design: Retrospective cohort analysis using data from the Scientific Registry of Transplant Recipients (SRTR) on patients who received an offer for a heart, liver, or lung transplant between January 1, 2010 and December 31, 2020. Setting: Nationwide, waitlist-based. Patients: 32,268 heart transplant candidates, 102,823 liver candidates, and 25,780 lung candidates, all aged 18 or older. Measurements: 1) Association between offer acceptance and two race-based variables: candidate race and donor-candidate race match; 2) association between offer rejection and time to patient mortality. Results: Black race was associated with significantly lower odds of offer acceptance for livers (OR=0.93, CI: 0.88-0.98) and lungs (OR=0.80, CI: 0.73-0.87). Donor-candidate race match was associated with significantly higher odds of offer acceptance for hearts (OR=1.11, CI: 1.06-1.16), livers (OR=1.10, CI: 1.06-1.13), and lungs (OR=1.13, CI: 1.07-1.19). Rejecting an offer was associated with lower survival times for all three organs (heart hazard ratio=1.16, CI: 1.09-1.23; liver HR=1.74, CI: 1.66-1.82; lung HR=1.21, CI: 1.15-1.28). Limitations: Our study analyzed the observational SRTR dataset, which has known limitations. Conclusion: Offer acceptance decisions are associated with inequity in the organ allocation system. Our findings demonstrate the additional barriers that Black patients face in accessing organ transplants and demonstrate the need for standardized practice, continuous distribution policies, and better organ procurement.

BRSET: A Brazilian Multilabel Ophthalmological Dataset of Retina Fundus Photos.

INTRODUCTION: The Brazilian Multilabel Ophthalmological Dataset (BRSET) addresses the scarcity of publicly available ophthalmological datasets in Latin America. BRSET comprises 16,266 color fundus retinal photos from 8,524 Brazilian patients, aiming to enhance data representativeness, serving as a research and teaching tool. It contains sociodemographic information, enabling investigations into differential model performance across demographic groups. METHODS: Data from three São Paulo outpatient centers yielded demographic and medical information from electronic records, including nationality, age, sex, clinical history, insulin use, and duration of diabetes diagnosis. A retinal specialist labeled images for anatomical features (optic disc, blood vessels, macula), quality control (focus, illumination, image field, artifacts), and pathologies (e.g., diabetic retinopathy). Diabetic retinopathy was graded using International Clinic Diabetic Retinopathy and Scottish Diabetic Retinopathy Grading. Validation used a ConvNext model trained during 50 epochs using a weighted cross entropy loss to avoid overfitting, with 70% training (20% validation), and 30% testing subsets. Performance metrics included area under the receiver operating curve (AUC) and Macro F1-score. Saliency maps were calculated for interpretability. RESULTS: BRSET comprises 65.1% Canon CR2 and 34.9% Nikon NF5050 images. 61.8% of the patients are female, and the average age is 57.6 (± 18.26) years. Diabetic retinopathy affected 15.8% of patients, across a spectrum of disease severity. Anatomically, 20.2% showed abnormal optic discs, 4.9% abnormal blood vessels, and 28.8% abnormal macula. A ConvNext V2 model was trained and evaluated BRSET in four prediction tasks: "binary diabetic retinopathy diagnosis (Normal vs Diabetic Retinopathy)" (AUC: 97, F1: 89); "3 class diabetic retinopathy diagnosis (Normal, Proliferative, Non-Proliferative)" (AUC: 97, F1: 82); "diabetes diagnosis" (AUC: 91, F1: 83); "sex classification" (AUC: 87, F1: 70). DISCUSSION: BRSET is the first multilabel ophthalmological dataset in Brazil and Latin America. It provides an opportunity for investigating model biases by evaluating performance across demographic groups. The model performance of three prediction tasks demonstrates the value of the dataset for external validation and for teaching medical computer vision to learners in Latin America using locally relevant data sources.

Epidemiology of Renal Replacement Therapy for Critically Ill Patients across Seven Health Jurisdictions.

INTRODUCTION: Acute kidney injury (AKI) requiring treatment with renal replacement therapy (RRT) is a common complication after admission to an intensive care unit (ICU) and is associated with significant morbidity and mortality. However, the prevalence of RRT use and the associated outcomes in critically patients across the globe are not well described. Therefore, we describe the epidemiology and outcomes of patients receiving RRT for AKI in ICUs across several large health system jurisdictions. METHODS: Retrospective cohort analysis using nationally representative and comparable databases from seven health jurisdictions in Australia, Brazil, Canada, Denmark, New Zealand, Scotland, and the USA between 2006 and 2023, depending on data availability of each dataset. Patients with a history of end-stage kidney disease receiving chronic RRT and patients with a history of renal transplant were excluded. RESULTS: A total of 4,104,480 patients in the ICU cohort and 3,520,516 patients in the mechanical ventilation cohort were included. Overall, 156,403 (3.8%) patients in the ICU cohort and 240,824 (6.8%) patients in the mechanical ventilation cohort were treated with RRT for AKI. In the ICU cohort, the proportion of patients treated with RRT was lowest in Australia and Brazil (3.3%) and highest in Scotland (9.2%). The in-hospital mortality for critically ill patients treated with RRT was almost fourfold higher (57.1%) than those not receiving RRT (16.8%). The mortality of patients treated with RRT varied across the health jurisdictions from 37 to 65%. CONCLUSION: The outcomes of patients who receive RRT in ICUs throughout the world vary widely. Our research suggests that differences in access to and provision of this therapy are contributing factors.

INSPIRE, a publicly available research dataset for perioperative medicine.

We present the INSPIRE dataset, a publicly available research dataset in perioperative medicine, which includes approximately 130,000 surgical operations at an academic institution in South Korea over a ten-year period between 2011 and 2020. This comprehensive dataset includes patient characteristics such as age, sex, American Society of Anesthesiologists physical status classification, diagnosis, surgical procedure code, department, and type of anaesthesia. The dataset also includes vital signs in the operating theatre, general wards, and intensive care units (ICUs), laboratory results from six months before admission to six months after discharge, and medication during hospitalisation. Complications include total hospital and ICU length of stay and in-hospital death. We hope this dataset will inspire collaborative research and development in perioperative medicine and serve as a reproducible external validation dataset to improve surgical outcomes.

A multimodal framework for extraction and fusion of satellite images and public health data.

In low- and middle-income countries, the substantial costs associated with traditional data collection pose an obstacle to facilitating decision-making in the field of public health. Satellite imagery offers a potential solution, but the image extraction and analysis can be costly and requires specialized expertise. We introduce SatelliteBench, a scalable framework for satellite image extraction and vector embeddings generation. We also propose a novel multimodal fusion pipeline that utilizes a series of satellite imagery and metadata. The framework was evaluated generating a dataset with a collection of 12,636 images and embeddings accompanied by comprehensive metadata, from 81 municipalities in Colombia between 2016 and 2018. The dataset was then evaluated in 3 tasks: including dengue case prediction, poverty assessment, and access to education. The performance showcases the versatility and practicality of SatelliteBench, offering a reproducible, accessible and open tool to enhance decision-making in public health.

Towards long-tailed, multi-label disease classification from chest X-ray: Overview of the CXR-LT challenge.

Many real-world image recognition problems, such as diagnostic medical imaging exams, are "long-tailed" - there are a few common findings followed by many more relatively rare conditions. In chest radiography, diagnosis is both a long-tailed and multi-label problem, as patients often present with multiple findings simultaneously. While researchers have begun to study the problem of long-tailed learning in medical image recognition, few have studied the interaction of label imbalance and label co-occurrence posed by long-tailed, multi-label disease classification. To engage with the research community on this emerging topic, we conducted an open challenge, CXR-LT, on long-tailed, multi-label thorax disease classification from chest X-rays (CXRs). We publicly release a large-scale benchmark dataset of over 350,000 CXRs, each labeled with at least one of 26 clinical findings following a long-tailed distribution. We synthesize common themes of top-performing solutions, providing practical recommendations for long-tailed, multi-label medical image classification. Finally, we use these insights to propose a path forward involving vision-language foundation models for few- and zero-shot disease classification.