Magn Reson Med Sci. 2024 Oct 29. doi: 10.2463/mrms.rev.2024-0056. Online ahead of print.

ABSTRACT

The integration of deep learning (DL) in breast MRI has revolutionized the field of medical imaging, notably enhancing diagnostic accuracy and efficiency. This review discusses the substantial influence of DL technologies across various facets of breast MRI, including image reconstruction, classification, object detection, segmentation, and prediction of clinical outcomes such as response to neoadjuvant chemotherapy and recurrence of breast cancer. Utilizing sophisticated models such as convolutional neural networks, recurrent neural networks, and generative adversarial networks, DL has improved image quality and precision, enabling more accurate differentiation between benign and malignant lesions and providing deeper insights into disease behavior and treatment responses. DL's predictive capabilities for patient-specific outcomes also suggest potential for more personalized treatment strategies. The advancements in DL are pioneering a new era in breast cancer diagnostics, promising more personalized and effective healthcare solutions. Nonetheless, the integration of this technology into clinical practice faces challenges, necessitating further research, validation, and development of legal and ethical frameworks to fully leverage its potential.

PMID:39477506 | DOI:10.2463/mrms.rev.2024-0056

Magn Reson Med Sci. 2024 Oct 29. doi: 10.2463/mrms.rev.2024-0077. Online ahead of print.

ABSTRACT

This review explores the significant progress and applications of artificial intelligence (AI) in obstetrics and gynecological MRI, charting its development from foundational algorithmic techniques to deep learning strategies and advanced radiomics. This review features research published over the last few years that has used AI with MRI to identify specific conditions such as uterine leiomyosarcoma, endometrial cancer, cervical cancer, ovarian tumors, and placenta accreta. In addition, it covers studies on the application of AI for segmentation and quality improvement in obstetrics and gynecology MRI. The review also outlines the existing challenges and envisions future directions for AI research in this domain. The growing accessibility of extensive datasets across various institutions and the application of multiparametric MRI are significantly enhancing the accuracy and adaptability of AI. This progress has the potential to enable more accurate and efficient diagnosis, offering opportunities for personalized medicine in the field of obstetrics and gynecology.

PMID:39477505 | DOI:10.2463/mrms.rev.2024-0077

Nihon Hoshasen Gijutsu Gakkai Zasshi. 2024 Oct 29. doi: 10.6009/jjrt.2024-1487. Online ahead of print.

ABSTRACT

PURPOSE: Using a magnetic resonance (MR) image generation technique with deep learning, we elucidated whether changing the training data patterns affected image generation accuracy.

METHODS: The pix2pix training model generated T1-weighted images from T2-weighted images or FLAIR images. Head MR images obtained at our hospital were used in this study. We prepared 300 cases for each model and four training data patterns for each model (a: 150 cases for one MR system, b: 300 cases for one MR system, c: 150 cases and augmentation data for one MR system, and d: 300 cases for two MR systems). The extension data were images of 150 cases rotated in the XY plane. The similarity between the images generated by the training and evaluation data in each group was evaluated using the peak signal-to-noise ratio (PSNR) and structural similarity (SSIM).

RESULTS: For both MR systems, the PSNR and SSIM were higher for training dataset b than training dataset a. The PSNR and SSIM were lower for training dataset d.

CONCLUSION: MR image generation accuracy varied among training data patterns.

PMID:39477465 | DOI:10.6009/jjrt.2024-1487

In Vivo. 2024 Nov-Dec;38(6):2864-2872. doi: 10.21873/invivo.13767.

ABSTRACT

BACKGROUND/AIM: Breast cancer remains a major global health concern. This study aimed to develop a deep-learning-based artificial intelligence (AI) model that predicts the malignancy of mammographic lesions and reduces unnecessary biopsies in patients with breast cancer.

PATIENTS AND METHODS: In this retrospective study, we used deep-learning-based AI to predict whether lesions in mammographic images are malignant. The AI model learned the malignancy as well as margins and shapes of mass lesions through multi-label training, similar to the diagnostic process of a radiologist. We used the Curated Breast Imaging Subset of Digital Database for Screening Mammography. This dataset includes annotations for mass lesions, and we developed an algorithm to determine the exact location of the lesions for accurate classification. A multi-label classification approach enabled the model to recognize malignancy and lesion attributes.

RESULTS: Our multi-label classification model, trained on both lesion shape and margin, demonstrated superior performance compared with models trained solely on malignancy. Gradient-weighted class activation mapping analysis revealed that by considering the margin and shape, the model assigned higher importance to border areas and analyzed pixels more uniformly when classifying malignant lesions. This approach improved diagnostic accuracy, particularly in challenging cases, such as American College of Radiology Breast Imaging-Reporting and Data System categories 3 and 4, where the breast density exceeded 50%.

CONCLUSION: This study highlights the potential of AI in improving the diagnosis of breast cancer. By integrating advanced techniques and modern neural network designs, we developed an AI model with enhanced accuracy for mammographic image analysis.

PMID:39477432 | DOI:10.21873/invivo.13767

In Vivo. 2024 Nov-Dec;38(6):3016-3021. doi: 10.21873/invivo.13785.

ABSTRACT

BACKGROUND/AIM: To evaluate efficacy of the AIxURO system, a deep learning-based artificial intelligence (AI) tool, in enhancing the accuracy and reliability of urine cytology for diagnosing upper urinary tract cancers.

MATERIALS AND METHODS: One hundred and eighty-five cytology samples of upper urine tract were collected and categorized according to The Paris System for Reporting Urinary Cytology (TPS), yielding 168 negative for High-Grade Urothelial Carcinoma (NHGUC), 14 atypical urothelial cells (AUC), 2 suspicious for high-grade urothelial carcinoma (SHGUC), and 1 high-grade urothelial carcinoma (HGUC). The AIxURO system, trained on annotated cytology images, was employed to analyze these samples. Independent assessments by a cytotechnologist and a cytopathologist were conducted to validate the initial AIxURO assessment.

RESULTS: AIxURO identified discrepancies in 37 of the 185 cases, resulting in a 20% discrepancy rate. The cytotechnologist achieved an accuracy of 85% for NHGUC and 21.4% for AUC, whereas the cytopathologist attained accuracies of 95% for NHGUC and 85.7% for AUC. The cytotechnologist exhibited overcall rates of roughly 15% and undercall rates of greater than 50%, while the cytopathologist showed profoundly lower miscall rates from both undercall and overcall. AIxURO significantly enhanced diagnostic accuracy and consistency, particularly in complex cases involving atypical cells.

CONCLUSION: AIxURO can improve the accuracy and reliability of cytology diagnosis for upper urine tract urothelial carcinomas by providing precise detection on atypical urothelial cells and reducing subjectivity in assessments. The integration of AIxURO into clinical practice can significantly ameliorate diagnostic outcomes, highlighting the synergistic potential of AI technology and human expertise in cytology.

PMID:39477382 | DOI:10.21873/invivo.13785

Child Health Nurs Res. 2024 Oct;30(4):277-287. doi: 10.4094/chnr.2024.020. Epub 2024 Oct 31.

ABSTRACT

PURPOSE: Hybrid simulation has emerged to increase the practicality of simulation training by combining simulators and standardized patient (SP) that implement realistic clinical environments at a high level. This study aimed to develop a hybrid simulation program focused on case of pediatric cardiac catheterization and to evaluate its effectiveness.

METHODS: The hybrid simulation program was developed according to the Analyze, Design, Develop, Implement, and Evaluate (ADDIE) model. And deep learning-based analysis program was used to analyze non-verbal communication with SP and applied it for debriefing sessions. To verify the effect of the program, a quasi-experimental study using a random assignment design was conducted. In total, 48 nursing students (n=24 in the experimental group; n=24 in the control group) participated in the study.

RESULTS: Knowledge (F=3.53, p=.038), confidence in clinical performance (F=9.73, p<.001), and communication self-efficacy (F=5.20, p=.007) showed a significant difference in both groups and interaction between time points, and the communication ability of the experimental group increased significantly (t=3.32, p=.003).

CONCLUSION: Hybrid simulation program developed in this study has been proven effective, it can be implemented in child nursing education. Future research should focus on developing and incorporating various hybrid simulation programs using SP into the nursing curriculum and evaluating their effectiveness.

PMID:39477234 | DOI:10.4094/chnr.2024.020

Biomed J. 2024 Oct 28:100803. doi: 10.1016/j.bj.2024.100803. Online ahead of print.

ABSTRACT

BACKGROUND: Recently, as a relatively novel technology, artificial intelligence (especially in the deep learning fields) has received more and more attention from researchers and has successfully been applied to many biomedical domains. Nonetheless, just a few research works use deep learning skills to predict the cardiac resynchronization therapy (CRT)-response of heart failure patients.

OBJECTIVE: We try to use the deep learning-based technique to construct a model which is used to predict the CRT response of patients with high prediction accuracy, precision, and sensitivity.

METHODS: Using two-dimensional echocardiographic strain traces from 131 patients, we pre-processed the data and synthesized 2,000 model inputs through the synthetic minority oversampling technique (SMOTE). These inputs trained and optimized deep neural networks (DNN) and one-dimensional convolution neural networks (1D-CNN). Visualization of prediction results was performed using t-distributed stochastic neighbor embedding (t-SNE), and model performance was evaluated using accuracy, precision, sensitivity, F1 score, and specificity. Variable importance was assessed using Shapley additive explanations (SHAP) analysis.

RESULTS: Both the optimal DNN and 1D-CNN models demonstrated exceptional predictive performance, with prediction accuracy, precision, and sensitivity all around 90%. Furthermore, the area under the receiver operating characteristic curve (AUROC) of the optimal 1D-CNN and DNN models achieved 0.8734 and 0.9217, respectively. Crucially, the most significant input variables for both models align well with clinical experience, further corroborating their robustness and applicability in real-world settings.

CONCLUSIONS: We believe that both the DL models could be an auxiliary to help in treatment response prediction for doctors because of the excellent prediction performance and the convenience of obtaining input data to predict the CRT response of patients clinically.

PMID:39477070 | DOI:10.1016/j.bj.2024.100803

J Arthroplasty. 2024 Oct 28:S0883-5403(24)01143-4. doi: 10.1016/j.arth.2024.10.103. Online ahead of print.

ABSTRACT

BACKGROUND: We present an automated image ingestion pipeline for a knee radiography registry, integrating a multilabel image-semantic classifier with conformal prediction-based uncertainty quantification and an object detection model for knee hardware.

METHODS: Annotators retrospectively classified 26,000 knee images detailing presence, laterality, prostheses, and radiographic views. They further annotated surgical construct locations in 11,841 knee radiographs. An uncertainty-aware multilabel EfficientNet-based classifier was trained to identify the knee laterality, implants, and radiographic view. A classifier trained with embeddings from the EfficientNet model detected out-of-domain images. An object detection model was trained to identify 20 different knee implants. Model performance was assessed against a held-out internal and an external dataset using per-class F1 score, accuracy, sensitivity, and specificity. Conformal prediction was evaluated with marginal coverage and efficiency.

RESULTS: Classification Model with Conformal Prediction: F1 scores for each label output > 0.98. Coverage of each label output was >0.99 and the average efficiency was 0.97.

DOMAIN DETECTION MODEL: The F1 score was 0.99, with precision and recall for knee radiographs of 0.99.

OBJECT DETECTION MODEL: Mean average precision across all classes was 0.945 and ranged from 0.695 to 1.000. Average precision and recall across all classes were 0.950 and 0.886.

CONCLUSIONS: We present a multilabel classifier with domain detection and an object detection model to characterize knee radiographs. Conformal prediction enhances transparency in cases when the model is uncertain.

PMID:39477040 | DOI:10.1016/j.arth.2024.10.103

Environ Pollut. 2024 Oct 28:125207. doi: 10.1016/j.envpol.2024.125207. Online ahead of print.

ABSTRACT

Despite the growing recognition of the impact of urban environments on social health, limited research explores the combined associations of multiple urban exposures, particularly in dense cities. This study examines the interplay between greenspace, air pollution, and social health as well as the underlying pathways and population heterogeneity in Hong Kong using cross-sectional survey data from 1,977 adults and residential environmental data. Social health includes social contacts, relations, and support. Greenspace used street-view greenness (SVG), park density, and the normalized difference vegetation index (NDVI). 100-m daily ground NO2 and O3, indicative of air pollution, were derived using a spatiotemporal deep learning model. Mediators involved physical activity and negative emotions. Main analyses were performed in a 1000-m buffer with multivariate logistical regressions, stratification, interaction, and Partial Lease Square - Structural Equation Modelling (PLS-SEM). Multi-exposure models revealed positive associations between park density/SVG and social contacts, as well as between SVG and social relations, while O3 was negatively associated with social relations/support. Significant moderators included age, birthplace, employment, and education. PLS-SEM indicated direct positive associations between SVG and social contacts/relations and significant indirect negative associations between NO2/O3 and social health via negative emotions. This study adds to urban health research by exploring complex relationships between greenspace, air pollution, and social health, highlighting the role of the environment in fostering social restoration.

PMID:39476997 | DOI:10.1016/j.envpol.2024.125207

Annu Rev Pathol. 2024 Oct 30. doi: 10.1146/annurev-pathmechdis-111523-023417. Online ahead of print.

ABSTRACT

Pathology has always been fueled by technological advances. Histology powered the study of tissue architecture at single-cell resolution and remains a cornerstone of clinical pathology today. In the last decade, next-generation sequencing has become informative for the targeted treatment of many diseases, demonstrating the importance of genome-scale molecular information for personalized medicine. Today, revolutionary developments in spatial transcriptomics technologies digitalize gene expression at subcellular resolution in intact tissue sections, enabling the computational analysis of cell types, cellular phenotypes, and cell-cell communication in routinely collected and archival clinical samples. Here we review how such molecular microscopes work, highlight their potential to identify disease mechanisms and guide personalized therapies, and provide guidance for clinical study design. Finally, we discuss remaining challenges to the swift translation of high-resolution spatial transcriptomics technologies and how integration of multimodal readouts and deep learning approaches is bringing us closer to a holistic understanding of tissue biology and pathology.

PMID:39476415 | DOI:10.1146/annurev-pathmechdis-111523-023417

Med Phys. 2024 Oct 30. doi: 10.1002/mp.17482. Online ahead of print.

ABSTRACT

BACKGROUND: Accurate segmentation of gliomas is crucial for diagnosis, treatment planning, and evaluating therapeutic efficacy. Physicians typically analyze and delineate tumor regions in brain magnetic resonance imaging (MRI) images based on personal experience, which is often time-consuming and subject to individual interpretation. Despite advancements in deep learning technology for image segmentation, current techniques still face challenges in clearly defining tumor boundary contours and enhancing segmentation accuracy.

PURPOSE: To address these issues, this paper proposes a conditional diffusion network (SF-Diff) with a spatial channel attention infusion (SCAI) module and a frequency spatial attention (FSA) mechanism to achieve accurate segmentation of the whole tumor (WT) region in brain tumors.

METHODS: SF-Diff initially extracts multiscale information from multimodal MRI images and subsequently employs a diffusion model to restore boundaries and details, thereby enabling accurate brain tumor segmentation (BraTS). Specifically, a SCAI module is developed to capture multiscale information within and between encoder layers. A dual-channel upsampling block (DUB) is designed to assist in detail recovery during upsampling. A FSA mechanism is introduced to better match the conditional features with the diffusion probability distribution information. Furthermore, a cross-model loss function has been implemented to supervise the feature extraction of the conditional model and the noise distribution of the diffusion model.

RESULTS: The dataset used in this paper is publicly available and includes 369 patient cases from the Multimodal BraTS Challenge 2020 (BraTS2020). The conducted experiments on BraTS2020 demonstrate that SF-Diff performs better than other state-of-the-art models. The method achieved a Dice score of 91.87%, a Hausdorff 95 of 5.47 mm, an IoU of 84.96%, a sensitivity of 92.29%, and a specificity of 99.95% on BraTS2020.

CONCLUSIONS: The proposed SF-Diff performs well in identifying the WT region of the brain tumors compared to other state-of-the-art models, especially in terms of boundary contours and non-contiguous lesion regions, which is clinically significant. In the future, we will further develop this method for brain tumor three-class segmentation task.

PMID:39476317 | DOI:10.1002/mp.17482

J Comput Chem. 2024 Oct 30. doi: 10.1002/jcc.27516. Online ahead of print.

ABSTRACT

Small molecule conformational sampling plays a pivotal role in molecular docking. Recent advancements have led to the emergence of various conformational sampling methods, each employing distinct algorithms. This study investigates the impact of different small molecule conformational sampling methods in molecular docking using UCSF DOCK 3.7. Specifically, six traditional sampling methods (Omega, BCL::Conf, CCDC Conformer Generator, ConfGenX, Conformator, RDKit ETKDGv3) and a deep learning-based model (Torsional Diffusion) for generating conformational ensembles are evaluated. These ensembles are subsequently docked against the Platinum Diverse Dataset, the PoseBusters dataset and the DUDE-Z dataset to assess binding pose reproducibility and screening power. Notably, different sampling methods exhibit varying performance due to their unique preferences, such as dihedral angle sampling ranges on rotatable bonds. Combining complementary methods may lead to further improvements in docking performance.

PMID:39476310 | DOI:10.1002/jcc.27516

Transl Vis Sci Technol. 2024 Oct 1;13(10):41. doi: 10.1167/tvst.13.10.41.

ABSTRACT

PURPOSE: To quantitatively characterize the posterior morphology of high myopia eyes with posterior staphyloma.

METHODS: Surface points of the eyeball were automatically extracted from magnetic resonance imaging scans using deep learning. Subsequently, the topography of posterior staphylomas was constructed to facilitate accurate visualization and quantification of their location and severity. In the three-dimensional Cartesian coordinate system established with surface points, measurements of distances (D) from each point to the hypothetical pre-elongation eye center within the eyeball and local curvatures (C) at each point on the posterior sclera were computed. Using this data, specific parameters were formulated. The concordance of these parameters with traditional staphyloma classification methods and their association with myopic traction maculopathy (MTM) grades based on the ATN classifications were investigated.

RESULTS: The study included 102 eyes from 52 participants. The measured parameters, particularly the variance of distance (Dvar) and the maximum value of the curvature and distance product (C · Dmax), demonstrated efficacy in differentiating various types of posterior staphyloma and exhibited strong correlations with the grades of MTM.

CONCLUSIONS: The automated generation of the posterior scleral topography facilitated visualization and quantification of staphyloma location and severity. Simple geometric parameters can quantify staphyloma shape and correlate well with retinal complications. Future works on expanding these measures to more imaging modalities could improve their clinical use and deepen insights into the link between posterior staphyloma and related retinal diseases.

TRANSLATIONAL RELEVANCE: This work has the potential to be translated into clinical practice, allowing for the accurate assessment of staphyloma severity and ultimately improving disease management.

PMID:39476086 | DOI:10.1167/tvst.13.10.41

PLoS One. 2024 Oct 30;19(10):e0310121. doi: 10.1371/journal.pone.0310121. eCollection 2024.

ABSTRACT

In order to assess risk of mosquito-vector borne disease and to effectively target and monitor vector control efforts, accurate information about mosquito vector population densities is needed. The traditional and still most common approach to this involves the use of traps along with manual counting and classification of mosquito species, but the costly and labor-intensive nature of this approach limits its widespread use. Numerous previous studies have sought to address this problem by developing machine learning models to automatically identify species and sex of mosquitoes based on their wingbeat sounds. Yet little work has addressed the issue of robust classification in the presence of environmental background noise, which is essential to making the approach practical. In this paper, we propose a new deep learning model, MosquitoSong+, to identify the species and sex of mosquitoes from raw wingbeat sounds so that it is robust to the environmental noise and the relative volume of the mosquito's flight tone. The proposed model extends the existing 1D-CNN model by adjusting its architecture and introducing two data augmentation techniques during model training: noise augmentation and wingbeat volume variation. Experiments show that the new model has very good generalizability, with species classification accuracy above 80% on several wingbeat datasets with various background noise. It also has an accuracy of 93.3% for species and sex classification on wingbeat sounds overlaid with various background noises. These results suggest that the proposed approach may be a practical means to develop classification models that can perform well in the field.

PMID:39475971 | DOI:10.1371/journal.pone.0310121

PLoS One. 2024 Oct 30;19(10):e0310097. doi: 10.1371/journal.pone.0310097. eCollection 2024.

ABSTRACT

PURPOSE: To establish a deep learning artificial intelligence model to predict the risk of long-term fellow eye neovascularization in unilateral type 3 macular neovascularization (MNV).

METHODS: This retrospective study included 217 patients (199 in the training/validation of the AI model and 18 in the testing set) with a diagnosis of unilateral type 3 MNV. The purpose of the AI model was to predict fellow eye neovascularization within 24 months after the initial diagnosis. The data used to train the AI model included a baseline fundus image and horizontal/vertical cross-hair scan optical coherence tomography images in the fellow eye. The neural network of this study for AI-learning was based on the visual geometry group with modification. The precision, recall, accuracy, and the area under the curve values of receiver operating characteristics (AUCROC) were calculated for the AI model. The accuracy of an experienced (examiner 1) and less experienced (examiner 2) human examiner was also evaluated.

RESULTS: The incidence of fellow eye neovascularization over 24 months was 28.6% in the training/validation set and 38.9% in the testing set (P = 0.361). In the AI model, precision was 0.562, recall was 0.714, accuracy was 0.667, and the AUCROC was 0.675. The sensitivity, specificity, and accuracy were 0.429, 0.727, and 0.611, respectively, for examiner 1, and 0.143, 0.636, and 0.444, respectively, for examiner 2.

CONCLUSIONS: This is the first AI study focusing on the clinical course of type 3 MNV. While our AI model exhibited accuracy comparable to that of human examiners, overall accuracy was not high. This may partly be a result of the relatively small number of patients used for AI training, suggesting the need for future multi-center studies to improve the accuracy of the model.

PMID:39475903 | DOI:10.1371/journal.pone.0310097

PLoS One. 2024 Oct 30;19(10):e0312022. doi: 10.1371/journal.pone.0312022. eCollection 2024.

ABSTRACT

Knowledge tracing is a technology that models students' changing knowledge state over learning time based on their historical answer records, thus predicting their learning ability. It is the core module that supports the intelligent education system. To address the problems of sparse input data, lack of interpretability and weak capacity to capture the relationship between exercises in the existing models, this paper build a deep knowledge tracing model DKVMN&MRI based on the Dynamic Key-Value Memory Network (DKVMN) that incorporates multiple relationship information including exercise-knowledge point relations, exercise-exercise relations, and learning-forgetting relations. In the model, firstly, the Q-matrix is utilized to map the link between knowledge points and exercises to the input layer; secondly, improved DKVMN and LSTM are used to model the learning process of learners, then the Ebbinghaus forgetting curve function is introduced to simulate the process of memory forgetting in learners, and finally, the prediction strategies of Item Response Theory (IRT) and attention mechanism are used to combine the similarity relationship between learners' knowledge state and exercises to calculate the probability that learners would correctly respond during the subsequent time step. Through extensive experiments on three real-world datasets, we demonstrate that DKVMN&MRI has significant improvements in both AUC and ACC metrics contrast with the latest models. Furthermore, the study provides explanations at both the exercise level and learner knowledge state level, demonstrating the interpretability and efficacy of the proposed model.

PMID:39475856 | DOI:10.1371/journal.pone.0312022

JMIR AI. 2024 Oct 30;3:e55059. doi: 10.2196/55059.

ABSTRACT

BACKGROUND: Global pandemics like COVID-19 put a high amount of strain on health care systems and health workers worldwide. These crises generate a vast amount of news information published online across the globe. This extensive corpus of articles has the potential to provide valuable insights into the nature of ongoing events and guide interventions and policies. However, the sheer volume of information is beyond the capacity of human experts to process and analyze effectively.

OBJECTIVE: The aim of this study was to explore how natural language processing (NLP) can be leveraged to build a system that allows for quick analysis of a high volume of news articles. Along with this, the objective was to create a workflow comprising human-computer symbiosis to derive valuable insights to support health workforce strategic policy dialogue, advocacy, and decision-making.

METHODS: We conducted a review of open-source news coverage from January 2020 to June 2022 on COVID-19 and its impacts on the health workforce from the World Health Organization (WHO) Epidemic Intelligence from Open Sources (EIOS) by synergizing NLP models, including classification and extractive summarization, and human-generated analyses. Our DeepCovid system was trained on 2.8 million news articles in English from more than 3000 internet sources across hundreds of jurisdictions.

RESULTS: Rules-based classification with hand-designed rules narrowed the data set to 8508 articles with high relevancy confirmed in the human-led evaluation. DeepCovid's automated information targeting component reached a very strong binary classification performance of 98.98 for the area under the receiver operating characteristic curve (ROC-AUC) and 47.21 for the area under the precision recall curve (PR-AUC). Its information extraction component attained good performance in automatic extractive summarization with a mean Recall-Oriented Understudy for Gisting Evaluation (ROUGE) score of 47.76. DeepCovid's final summaries were used by human experts to write reports on the COVID-19 pandemic.

CONCLUSIONS: It is feasible to synergize high-performing NLP models and human-generated analyses to benefit open-source health workforce intelligence. The DeepCovid approach can contribute to an agile and timely global view, providing complementary information to scientific literature.

PMID:39475833 | DOI:10.2196/55059

JMIR Med Inform. 2024 Oct 30;12:e57124. doi: 10.2196/57124.

ABSTRACT

BACKGROUND: Social determinants of health (SDOH) are critical drivers of health disparities and patient outcomes. However, accessing and collecting patient-level SDOH data can be operationally challenging in the emergency department (ED) clinical setting, requiring innovative approaches.

OBJECTIVE: This scoping review examines the potential of AI and data science for modeling, extraction, and incorporation of SDOH data specifically within EDs, further identifying areas for advancement and investigation.

METHODS: We conducted a standardized search for studies published between 2015 and 2022, across Medline (Ovid), Embase (Ovid), CINAHL, Web of Science, and ERIC databases. We focused on identifying studies using AI or data science related to SDOH within emergency care contexts or conditions. Two specialized reviewers in emergency medicine (EM) and clinical informatics independently assessed each article, resolving discrepancies through iterative reviews and discussion. We then extracted data covering study details, methodologies, patient demographics, care settings, and principal outcomes.

RESULTS: Of the 1047 studies screened, 26 met the inclusion criteria. Notably, 9 out of 26 (35%) studies were solely concentrated on ED patients. Conditions studied spanned broad EM complaints and included sepsis, acute myocardial infarction, and asthma. The majority of studies (n=16) explored multiple SDOH domains, with homelessness/housing insecurity and neighborhood/built environment predominating. Machine learning (ML) techniques were used in 23 of 26 studies, with natural language processing (NLP) being the most commonly used approach (n=11). Rule-based NLP (n=5), deep learning (n=2), and pattern matching (n=4) were the most commonly used NLP techniques. NLP models in the reviewed studies displayed significant predictive performance with outcomes, with F1-scores ranging between 0.40 and 0.75 and specificities nearing 95.9%.

CONCLUSIONS: Although in its infancy, the convergence of AI and data science techniques, especially ML and NLP, with SDOH in EM offers transformative possibilities for better usage and integration of social data into clinical care and research. With a significant focus on the ED and notable NLP model performance, there is an imperative to standardize SDOH data collection, refine algorithms for diverse patient groups, and champion interdisciplinary synergies. These efforts aim to harness SDOH data optimally, enhancing patient care and mitigating health disparities. Our research underscores the vital need for continued investigation in this domain.

PMID:39475815 | DOI:10.2196/57124

JMIR Dermatol. 2022 Sep 12;5(3):e35150. doi: 10.2196/35150.

ABSTRACT

BACKGROUND: Convolutional neural networks (CNNs) are a type of artificial intelligence that shows promise as a diagnostic aid for skin cancer. However, the majority are trained using retrospective image data sets with varying image capture standardization.

OBJECTIVE: The aim of our study was to use CNN models with the same architecture-trained on image sets acquired with either the same image capture device and technique (standardized) or with varied devices and capture techniques (nonstandardized)-and test variability in performance when classifying skin cancer images in different populations.

METHODS: In all, 3 CNNs with the same architecture were trained. CNN nonstandardized (CNN-NS) was trained on 25,331 images taken from the International Skin Imaging Collaboration (ISIC) using different image capture devices. CNN standardized (CNN-S) was trained on 177,475 MoleMap images taken with the same capture device, and CNN standardized number 2 (CNN-S2) was trained on a subset of 25,331 standardized MoleMap images (matched for number and classes of training images to CNN-NS). These 3 models were then tested on 3 external test sets: 569 Danish images, the publicly available ISIC 2020 data set consisting of 33,126 images, and The University of Queensland (UQ) data set of 422 images. Primary outcome measures were sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). Teledermatology assessments available for the Danish data set were used to determine model performance compared to teledermatologists.

RESULTS: When tested on the 569 Danish images, CNN-S achieved an AUROC of 0.861 (95% CI 0.830-0.889) and CNN-S2 achieved an AUROC of 0.831 (95% CI 0.798-0.861; standardized models), with both outperforming CNN-NS (nonstandardized model; P=.001 and P=.009, respectively), which achieved an AUROC of 0.759 (95% CI 0.722-0.794). When tested on 2 additional data sets (ISIC 2020 and UQ), CNN-S (P<.001 and P<.001, respectively) and CNN-S2 (P=.08 and P=.35, respectively) still outperformed CNN-NS. When the CNNs were matched to the mean sensitivity and specificity of the teledermatologists on the Danish data set, the models' resultant sensitivities and specificities were surpassed by the teledermatologists. However, when compared to CNN-S, the differences were not statistically significant (sensitivity: P=.10; specificity: P=.053). Performance across all CNN models as well as teledermatologists was influenced by image quality.

CONCLUSIONS: CNNs trained on standardized images had improved performance and, therefore, greater generalizability in skin cancer classification when applied to unseen data sets. This finding is an important consideration for future algorithm development, regulation, and approval.

PMID:39475778 | DOI:10.2196/35150

Assist Technol. 2024 Oct 30:1-16. doi: 10.1080/10400435.2024.2413650. Online ahead of print.

ABSTRACT

Visually impaired individuals actively utilize devices like computers, tablets, and smartphones, due to advancements in screen reader technologies. Integrating freely available deep learning models, image captioning can further enhance these readers, providing an affordable assistive tech solution. This research outlines the critical software requirements necessary for image captioning tools to effectively serve this demographic. Two qualitative investigations were conducted to determine these requirements. An online survey was first conducted to identify the main preferences of visually impaired users in relation to audio descriptive software, with findings visualized using word clouds. A subsequent study evaluated the proficiency of existing deep learning captioning models in addressing these stipulated requirements. Emphasizing the need for comprehensive image data, the results highlighted three primary areas: 1) characteristics of individuals, 2) color specifics of objects, and 3) the overall context of images. The research indicates that current captioning tools are not entirely effective for the visually impaired. Based on the delineated requirements and suggested future research paths, there is potential for the development of improved image captioning systems, advancing digital accessibility for the visually impaired.

PMID:39475401 | DOI:10.1080/10400435.2024.2413650