Dynamic brain networks are crucial for diagnosing brain disorders, as they reveal changes in brain activity and connectivity over time. Previous methods exploit the sliding window approach on fMRI data to construct these networks. However, this approach encounters two major issues: fixed temporal length, which inadequately captures the temporal dynamics of brain activity, and global spatial scope, which introduces noise and reduces sensitivity to localized dysfunctions when applied across the entire brain. These issues can lead to inaccurate brain network representations, potentially resulting in misdiagnosis. To overcome these challenges, we propose BrainDGT, a dynamic Graph Transformer model that adaptively captures and analyzes modular brain activities for improved diagnosis of brain disorders. BrainDGT addresses the fixed temporal length issue by estimating adaptive brain states through deconvolution of the Hemodynamic Response Function (HRF), avoiding the constraints of fixed-size windows. It also addresses the global spatial scope issue by segmenting fMRI scans into functional modules based on established brain networks for detailed, module-specific analysis. The model employs a dual attention mechanism: graph-attention extracts structural features from dynamic brain network snapshots, while self-attention identifies significant temporal dependencies. These spatio-temporal features are adaptively fused into a unified representation for disorder classification. BrainDGT’s effectiveness is validated through classification experiments on three real fMRI datasets ADNI, PPMI, and ABIDE demonstrating superior performance compared to state-of-the-art methods. BrainDGT improves brain disorder diagnosis by offering adaptive, localized analysis of dynamic brain networks, advancing neuroimaging and enabling more precise treatments in biomedical research.
{"title":"Dynamic Graph Transformer for Brain Disorder Diagnosis","authors":"Ahsan Shehzad, Dongyu Zhang, Shuo Yu, Shagufta Abid, Feng Xia","doi":"10.1101/2024.09.05.24313048","DOIUrl":"https://doi.org/10.1101/2024.09.05.24313048","url":null,"abstract":"Dynamic brain networks are crucial for diagnosing brain disorders, as they reveal changes in brain activity and connectivity over time. Previous methods exploit the sliding window approach on fMRI data to construct these networks. However, this approach encounters two major issues: fixed temporal length, which inadequately captures the temporal dynamics of brain activity, and global spatial scope, which introduces noise and reduces sensitivity to localized dysfunctions when applied across the entire brain. These issues can lead to inaccurate brain network representations, potentially resulting in misdiagnosis. To overcome these challenges, we propose BrainDGT, a dynamic Graph Transformer model that adaptively captures and analyzes modular brain activities for improved diagnosis of brain disorders. BrainDGT addresses the fixed temporal length issue by estimating adaptive brain states through deconvolution of the Hemodynamic Response Function (HRF), avoiding the constraints of fixed-size windows. It also addresses the global spatial scope issue by segmenting fMRI scans into functional modules based on established brain networks for detailed, module-specific analysis. The model employs a dual attention mechanism: graph-attention extracts structural features from dynamic brain network snapshots, while self-attention identifies significant temporal dependencies. These spatio-temporal features are adaptively fused into a unified representation for disorder classification. BrainDGT’s effectiveness is validated through classification experiments on three real fMRI datasets ADNI, PPMI, and ABIDE demonstrating superior performance compared to state-of-the-art methods. BrainDGT improves brain disorder diagnosis by offering adaptive, localized analysis of dynamic brain networks, advancing neuroimaging and enabling more precise treatments in biomedical research.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211702","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-06DOI: 10.1101/2024.09.04.24312955
Natalia Castano-Villegas, Isabella Llano, Maria Camila Villa, Julian Martinez, Jose Zea, Tatiana Urrea, Alejandra Maria Bañol, Carlos Bohorquez, Nelson Martinez
Background Conversational Agents have attracted attention for personal and professional use. Their specialisation in the medical field is being explored. Conversational Agents (CA) have accomplished passing-level performance in medical school examinations and shown empathy when responding to patient questions. Alzheimer’s disease is characterized by the progression of cognitive and somatic decline. As the leading cause of dementia in the elderly, it is the subject of continuous investigations, which result in a constant stream of new information. Physicians are expected to keep up with the latest clinical guidelines; however, they aren’t always able to do so due to the large amount of information and their busy schedules.
{"title":"Development and initial evaluation of a conversational agent for Alzheimer’s disease","authors":"Natalia Castano-Villegas, Isabella Llano, Maria Camila Villa, Julian Martinez, Jose Zea, Tatiana Urrea, Alejandra Maria Bañol, Carlos Bohorquez, Nelson Martinez","doi":"10.1101/2024.09.04.24312955","DOIUrl":"https://doi.org/10.1101/2024.09.04.24312955","url":null,"abstract":"<strong>Background</strong> Conversational Agents have attracted attention for personal and professional use. Their specialisation in the medical field is being explored. Conversational Agents (CA) have accomplished passing-level performance in medical school examinations and shown empathy when responding to patient questions. Alzheimer’s disease is characterized by the progression of cognitive and somatic decline. As the leading cause of dementia in the elderly, it is the subject of continuous investigations, which result in a constant stream of new information. Physicians are expected to keep up with the latest clinical guidelines; however, they aren’t always able to do so due to the large amount of information and their busy schedules.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"46 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-04DOI: 10.1101/2024.09.04.24313058
Juan C. Rojas, Patrick G. Lyons, Kaveri Chhikara, Vaishvik Chaudhari, Sivasubramanium V. Bhavani, Muna Nour, Kevin G. Buell, Kevin D. Smith, Catherine A. Gao, Saki Amagai, Chengsheng Mao, Yuan Luo, Anna K Barker, Mark Nuppnau, Haley Beck, Rachel Baccile, Michael Hermsen, Zewei Liao, Brenna Park-Egan, Kyle A Carey, XuanHan, Chad H Hochberg, Nicholas E Ingraham, William F Parker
Background Critical illness, or acute organ failure requiring life support, threatens over five million American lives annually. Electronic health record (EHR) data are a source of granular information that could generate crucial insights into the nature and optimal treatment of critical illness. However, data management, security, and standardization are barriers to large-scale critical illness EHR studies.
{"title":"A Common Longitudinal Intensive Care Unit data Format (CLIF) to enable multi-institutional federated critical illness research","authors":"Juan C. Rojas, Patrick G. Lyons, Kaveri Chhikara, Vaishvik Chaudhari, Sivasubramanium V. Bhavani, Muna Nour, Kevin G. Buell, Kevin D. Smith, Catherine A. Gao, Saki Amagai, Chengsheng Mao, Yuan Luo, Anna K Barker, Mark Nuppnau, Haley Beck, Rachel Baccile, Michael Hermsen, Zewei Liao, Brenna Park-Egan, Kyle A Carey, XuanHan, Chad H Hochberg, Nicholas E Ingraham, William F Parker","doi":"10.1101/2024.09.04.24313058","DOIUrl":"https://doi.org/10.1101/2024.09.04.24313058","url":null,"abstract":"<strong>Background</strong> Critical illness, or acute organ failure requiring life support, threatens over five million American lives annually. Electronic health record (EHR) data are a source of granular information that could generate crucial insights into the nature and optimal treatment of critical illness. However, data management, security, and standardization are barriers to large-scale critical illness EHR studies.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"40 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211705","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-04DOI: 10.1101/2024.09.03.24312994
Louis Rebaud, Nicolò Capobianco, Nicolas Captier, Thibault Escobar, Bruce Spottiswoode, Irène Buvat
In the analysis of medical data with censored outcomes, identifying the optimal machine learning pipeline is a challenging task, often requiring extensive preprocessing, feature selection, model testing, and tuning. To investigate the impact of the choice of pipeline on prediction performance, we evaluated 9 machine learning models on 71 medical datasets with censored targets. Only the decision tree model was consistently underperforming, while the other 8 models performed similarly across datasets, with little to no improvement from preprocessing optimization and hyperparameter tuning. Interestingly, more complex models did not outperform simpler ones, and reciprocally. ICARE, a straightforward model univariately learning only the sign of each feature instead of a weight, demonstrated similar performance to other models across most datasets while exhibiting lower overfitting, particularly in high-dimensional datasets. These findings suggest that using the ICARE model to build signatures between centers could improve reproducibility. Our findings also challenge the traditional approach of extensive model testing and tuning to improve performance.
{"title":"Similar performance of 8 machine learning models on 71 censored medical datasets: a case for simplicity","authors":"Louis Rebaud, Nicolò Capobianco, Nicolas Captier, Thibault Escobar, Bruce Spottiswoode, Irène Buvat","doi":"10.1101/2024.09.03.24312994","DOIUrl":"https://doi.org/10.1101/2024.09.03.24312994","url":null,"abstract":"In the analysis of medical data with censored outcomes, identifying the optimal machine learning pipeline is a challenging task, often requiring extensive preprocessing, feature selection, model testing, and tuning. To investigate the impact of the choice of pipeline on prediction performance, we evaluated 9 machine learning models on 71 medical datasets with censored targets. Only the decision tree model was consistently underperforming, while the other 8 models performed similarly across datasets, with little to no improvement from preprocessing optimization and hyperparameter tuning. Interestingly, more complex models did not outperform simpler ones, and reciprocally. ICARE, a straightforward model univariately learning only the sign of each feature instead of a weight, demonstrated similar performance to other models across most datasets while exhibiting lower overfitting, particularly in high-dimensional datasets. These findings suggest that using the ICARE model to build signatures between centers could improve reproducibility. Our findings also challenge the traditional approach of extensive model testing and tuning to improve performance.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"58 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-04DOI: 10.1101/2024.09.02.24311997
Tianhao Zhu, Kexin Xu, Wonchan Son, Kristofer Linton-Reid, Marc Boubnovski-Martell, Matt Grech-Sollars, Antoine D. Lain, Joram M. Posma
Chest X-ray (CXR) is a conventional diagnostic tool for cardiothoracic assessment, boasting a high degree of costeffectiveness and versatility. However, with an increasing number of scans to be evaluated by radiologists, they can suffer from fatigue which might impede diagnostic accuracy and slow down report generation. We describe a prototype computer-assisted diagnosis (CAD) pipeline employing computer vision (CV) and Natural Language Processing (NLP) trained on the publicly available MIMICCXR dataset. We perform image quality assessment, view labelling, segmentation-based cardiomegaly severity classification, and use the output of the severity classification for large language model-based report generation. Four certified radiologists assessed the output accuracy of the CAD pipeline. Across the dataset composed of 377,100 CXR images and 227,827 free-text radiology reports, our system identified 0.18% of cases with mixedsex mentions, 0.02% of poor quality images (F1=0.81), and 0.28% of wrongly labelled views (accuracy 99.4%), furthermore it assigned views for 4.18% of images which have unlabelled views. For binary cardiomegaly classification, we achieve state-of-the-art performance of 95.2% accuracy. The inter-radiologist agreement on evaluating the report’s semantics and correctness for radiologistMIMIC is 0.62 (strict agreement) and 0.85 (relaxed agreement) similar to the radiologist-CAD agreement of 0.55 (strict) and 0.93 (relaxed). Our work found and corrected several incorrect or missing metadata annotations for the MIMIC-CXR dataset, and the performance of our CAD system suggests performance on par with human radiologists. Future improvements revolve around improved text generation and the development of CV tools for other diseases.
{"title":"Designing a computer-assisted diagnosis system for cardiomegaly detection and radiology report generation","authors":"Tianhao Zhu, Kexin Xu, Wonchan Son, Kristofer Linton-Reid, Marc Boubnovski-Martell, Matt Grech-Sollars, Antoine D. Lain, Joram M. Posma","doi":"10.1101/2024.09.02.24311997","DOIUrl":"https://doi.org/10.1101/2024.09.02.24311997","url":null,"abstract":"Chest X-ray (CXR) is a conventional diagnostic tool for cardiothoracic assessment, boasting a high degree of costeffectiveness and versatility. However, with an increasing number of scans to be evaluated by radiologists, they can suffer from fatigue which might impede diagnostic accuracy and slow down report generation. We describe a prototype computer-assisted diagnosis (CAD) pipeline employing computer vision (CV) and Natural Language Processing (NLP) trained on the publicly available MIMICCXR dataset. We perform image quality assessment, view labelling, segmentation-based cardiomegaly severity classification, and use the output of the severity classification for large language model-based report generation. Four certified radiologists assessed the output accuracy of the CAD pipeline. Across the dataset composed of 377,100 CXR images and 227,827 free-text radiology reports, our system identified 0.18% of cases with mixedsex mentions, 0.02% of poor quality images (F1=0.81), and 0.28% of wrongly labelled views (accuracy 99.4%), furthermore it assigned views for 4.18% of images which have unlabelled views. For binary cardiomegaly classification, we achieve state-of-the-art performance of 95.2% accuracy. The inter-radiologist agreement on evaluating the report’s semantics and correctness for radiologistMIMIC is 0.62 (strict agreement) and 0.85 (relaxed agreement) similar to the radiologist-CAD agreement of 0.55 (strict) and 0.93 (relaxed). Our work found and corrected several incorrect or missing metadata annotations for the MIMIC-CXR dataset, and the performance of our CAD system suggests performance on par with human radiologists. Future improvements revolve around improved text generation and the development of CV tools for other diseases.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"58 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211707","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-03DOI: 10.1101/2024.08.31.24312878
Reem Agbareia, Mahmud Omar, Shelly Soffer, Benjamin S Glicksberg, Girish N Nadkarni, Eyal Klang
Background and Aim Visual data from images is essential for many medical diagnoses. This study evaluates the performance of multimodal Large Language Models (LLMs) in integrating textual and visual information for diagnostic purposes.
{"title":"Visual-Textual Integration in LLMs for Medical Diagnosis: A Quantitative Analysis","authors":"Reem Agbareia, Mahmud Omar, Shelly Soffer, Benjamin S Glicksberg, Girish N Nadkarni, Eyal Klang","doi":"10.1101/2024.08.31.24312878","DOIUrl":"https://doi.org/10.1101/2024.08.31.24312878","url":null,"abstract":"<strong>Background and Aim</strong> Visual data from images is essential for many medical diagnoses. This study evaluates the performance of multimodal Large Language Models (LLMs) in integrating textual and visual information for diagnostic purposes.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"119 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-03DOI: 10.1101/2024.09.02.24312934
Yang Xiao, Guilherme Soares, Leonardo Bastos, Rafael Izbicki, Paula Moraga
Dengue is a mosquito-borne viral disease that poses significant public health challenges in tropical and sub-tropical regions worldwide. Surveillance systems are essential for dengue prevention and control. However, traditional systems often rely on delayed data, limiting their effectiveness. To address this, nowcasting methods are needed to estimate underreported cases, enabling more timely decision-making. This study evaluates the value of using Google Trends indices of dengue-related keywords to complement official dengue data for nowcasting dengue in Brazil, a country frequently affected by this disease. We compare various nowcasting approaches that incorporate autoregressive features from official dengue cases, Google Trends data, and a combination of both, using a naive approach as a baseline. The performance of these methods is evaluated by nowcasting weekly dengue cases from March to June 2024 across Brazilian states. Error measures and 95% coverage probabilities reveal that models incorporating Google Trends data enhance the accuracy of weekly nowcasts across states and offer valuable insights into dengue activity levels. To support real-time decision-making, we also present Dengue Tracker, a website that displays weekly dengue nowcasts and trends to inform both decision-makers and the public, improving situational awareness of dengue activity. In conclusion, the study demonstrates the value of digital data sources in enhancing dengue nowcasting, and emphasizes the value of integrating alternative data streams into traditional surveillance systems for better-informed decision-making.
{"title":"Dengue nowcasting in Brazil by combining official surveillance data and Google Trends information","authors":"Yang Xiao, Guilherme Soares, Leonardo Bastos, Rafael Izbicki, Paula Moraga","doi":"10.1101/2024.09.02.24312934","DOIUrl":"https://doi.org/10.1101/2024.09.02.24312934","url":null,"abstract":"Dengue is a mosquito-borne viral disease that poses significant public health challenges in tropical and sub-tropical regions worldwide. Surveillance systems are essential for dengue prevention and control. However, traditional systems often rely on delayed data, limiting their effectiveness. To address this, nowcasting methods are needed to estimate underreported cases, enabling more timely decision-making. This study evaluates the value of using Google Trends indices of dengue-related keywords to complement official dengue data for nowcasting dengue in Brazil, a country frequently affected by this disease. We compare various nowcasting approaches that incorporate autoregressive features from official dengue cases, Google Trends data, and a combination of both, using a naive approach as a baseline. The performance of these methods is evaluated by nowcasting weekly dengue cases from March to June 2024 across Brazilian states. Error measures and 95% coverage probabilities reveal that models incorporating Google Trends data enhance the accuracy of weekly nowcasts across states and offer valuable insights into dengue activity levels. To support real-time decision-making, we also present Dengue Tracker, a website that displays weekly dengue nowcasts and trends to inform both decision-makers and the public, improving situational awareness of dengue activity. In conclusion, the study demonstrates the value of digital data sources in enhancing dengue nowcasting, and emphasizes the value of integrating alternative data streams into traditional surveillance systems for better-informed decision-making.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211499","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-03DOI: 10.1101/2024.09.02.24312874
Magdalena Z. Raban, Alison Merchant, Erin Fitzpatrick, Melissa T. Baysari, Ling Li, Peter J. Gates, Johanna I. Westbrook
Objectives Technology-related prescribing errors curtail the positive impacts of computerised provider order entry (CPOE) on medication safety. Understanding how technology-related errors occur can inform CPOE optimisation. Previously, we developed a classification of the underlying mechanisms of technology-related errors using prescribing error data from two adult hospitals. Our objective was to update the classification using paediatric prescribing error data, and to assess the reliability with which reviewers could independently apply the classification.
{"title":"Understanding technology-related prescribing errors for system optimisation: the Technology-Related Error Mechanism (TREM) classification","authors":"Magdalena Z. Raban, Alison Merchant, Erin Fitzpatrick, Melissa T. Baysari, Ling Li, Peter J. Gates, Johanna I. Westbrook","doi":"10.1101/2024.09.02.24312874","DOIUrl":"https://doi.org/10.1101/2024.09.02.24312874","url":null,"abstract":"<strong>Objectives</strong> Technology-related prescribing errors curtail the positive impacts of computerised provider order entry (CPOE) on medication safety. Understanding how technology-related errors occur can inform CPOE optimisation. Previously, we developed a classification of the underlying mechanisms of technology-related errors using prescribing error data from two adult hospitals. Our objective was to update the classification using paediatric prescribing error data, and to assess the reliability with which reviewers could independently apply the classification.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"79 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211715","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-03DOI: 10.1101/2024.09.02.24312917
Isabella Catharina Wiest, Fabian Wolf, Marie-Elisabeth Leßmann, Marko van Treeck, Dyke Ferber, Jiefu Zhu, Heiko Boehme, Keno K. Bressem, Hannes Ulrich, Matthias P. Ebert, Jakob Nikolas Kather
In clinical science and practice, text data, such as clinical letters or procedure reports, is stored in an unstructured way. This type of data is not a quantifiable resource for any kind of quantitative investigations and any manual review or structured information retrieval is time-consuming and costly. The capabilities of Large Language Models (LLMs) mark a paradigm shift in natural language processing and offer new possibilities for structured Information Extraction (IE) from medical free text. This protocol describes a workflow for LLM based information extraction (LLM-AIx), enabling extraction of predefined entities from unstructured text using privacy preserving LLMs. By converting unstructured clinical text into structured data, LLM-AIx addresses a critical barrier in clinical research and practice, where the efficient extraction of information is essential for improving clinical decision-making, enhancing patient outcomes, and facilitating large-scale data analysis.
{"title":"LLM-AIx: An open source pipeline for Information Extraction from unstructured medical text based on privacy preserving Large Language Models","authors":"Isabella Catharina Wiest, Fabian Wolf, Marie-Elisabeth Leßmann, Marko van Treeck, Dyke Ferber, Jiefu Zhu, Heiko Boehme, Keno K. Bressem, Hannes Ulrich, Matthias P. Ebert, Jakob Nikolas Kather","doi":"10.1101/2024.09.02.24312917","DOIUrl":"https://doi.org/10.1101/2024.09.02.24312917","url":null,"abstract":"In clinical science and practice, text data, such as clinical letters or procedure reports, is stored in an unstructured way. This type of data is not a quantifiable resource for any kind of quantitative investigations and any manual review or structured information retrieval is time-consuming and costly. The capabilities of Large Language Models (LLMs) mark a paradigm shift in natural language processing and offer new possibilities for structured Information Extraction (IE) from medical free text. This protocol describes a workflow for LLM based information extraction (LLM-AIx), enabling extraction of predefined entities from unstructured text using privacy preserving LLMs. By converting unstructured clinical text into structured data, LLM-AIx addresses a critical barrier in clinical research and practice, where the efficient extraction of information is essential for improving clinical decision-making, enhancing patient outcomes, and facilitating large-scale data analysis.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"256 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-02DOI: 10.1101/2024.09.01.24312016
Felix Busch, Lena Hoffmann, Lina Xu, Longjiang Zhang, Bin Hu, Ignacio García-Juárez, Liz N Toapanta-Yanchapaxi, Natalia Gorelik, Valérie Gorelik, Gaston A Rodriguez-Granillo, Carlos Ferrarotti, Nguyen N Cuong, Chau AP Thi, Murat Tuncel, Gürsan Kaya, Sergio M Solis-Barquero, Maria C Mendez Avila, Nevena G Ivanova, Felipe C Kitamura, Karina YI Hayama, Monserrat L Puntunet Bates, Pedro Iturralde Torres, Esteban Ortiz-Prado, Juan S Izquierdo-Condoy, Gilbert M Schwarz, Jochen G Hofstaetter, Michihiro Hide, Konagi Takeda, Barbara Perić, Gašper Pilko, Hans O Thulesius, Thomas A Lindow, Israel K Kolawole, Samuel Adegboyega Olatoke, Andrzej Grzybowski, Alexandru Corlateanu, Oana-Simina Iaconi, Ting Li, Izabela Domitrz, Katarzyna Kępczyńska, Matúš Mihalčin, Lenka Fašaneková, Tomasz Zatoński, Katarzyna Fułek, András Molnár, Stefani Maihoub, Zenewton A da Silva Gama, Luca Saba, Petros Sountoulides, Marcus R Makowski, Hugo JWL Aerts, Lisa C Adams, Keno K Bressem, COMFORT consortium
The successful implementation of artificial intelligence (AI) in healthcare is dependent upon the acceptance of this technology by key stakeholders, particularly patients, who are the primary beneficiaries of AI-driven outcomes. This international, multicenter, cross-sectional study assessed the attitudes of hospital patients towards AI in healthcare across 43 countries. A total of 13806 patients at 74 hospitals were surveyed between February and November 2023, with 64.8% from the Global North and 35.2% from the Global South. The findings indicate a predominantly favorable general view of AI in healthcare, with 57.6% of respondents expressing a positive attitude. However, attitudes exhibited notable variation based on demographic characteristics, health status, and technological literacy. Female respondents and those with poorer health status exhibited fewer positive attitudes towards AI use in medicine. Conversely, higher levels of AI knowledge and frequent use of technology devices were associated with more positive attitudes. It is noteworthy that less than half of the participants expressed positive attitudes regarding all items pertaining to trust in AI. The lowest level of trust was observed for the accuracy of AI in providing information regarding treatment responses. Patients exhibited a strong preference for explainable AI and physician-led decision-making, even if it meant slightly compromised accuracy. This large-scale, multinational study provides a comprehensive perspective on patient attitudes towards AI in healthcare across six continents. Findings suggest a need for tailored AI implementation strategies that consider patient demographics, health status, and preferences for explainable AI and physician oversight. All study data has been made publicly available to encourage replication and further investigation.
{"title":"Multinational attitudes towards AI in healthcare and diagnostics among hospital patients","authors":"Felix Busch, Lena Hoffmann, Lina Xu, Longjiang Zhang, Bin Hu, Ignacio García-Juárez, Liz N Toapanta-Yanchapaxi, Natalia Gorelik, Valérie Gorelik, Gaston A Rodriguez-Granillo, Carlos Ferrarotti, Nguyen N Cuong, Chau AP Thi, Murat Tuncel, Gürsan Kaya, Sergio M Solis-Barquero, Maria C Mendez Avila, Nevena G Ivanova, Felipe C Kitamura, Karina YI Hayama, Monserrat L Puntunet Bates, Pedro Iturralde Torres, Esteban Ortiz-Prado, Juan S Izquierdo-Condoy, Gilbert M Schwarz, Jochen G Hofstaetter, Michihiro Hide, Konagi Takeda, Barbara Perić, Gašper Pilko, Hans O Thulesius, Thomas A Lindow, Israel K Kolawole, Samuel Adegboyega Olatoke, Andrzej Grzybowski, Alexandru Corlateanu, Oana-Simina Iaconi, Ting Li, Izabela Domitrz, Katarzyna Kępczyńska, Matúš Mihalčin, Lenka Fašaneková, Tomasz Zatoński, Katarzyna Fułek, András Molnár, Stefani Maihoub, Zenewton A da Silva Gama, Luca Saba, Petros Sountoulides, Marcus R Makowski, Hugo JWL Aerts, Lisa C Adams, Keno K Bressem, COMFORT consortium","doi":"10.1101/2024.09.01.24312016","DOIUrl":"https://doi.org/10.1101/2024.09.01.24312016","url":null,"abstract":"The successful implementation of artificial intelligence (AI) in healthcare is dependent upon the acceptance of this technology by key stakeholders, particularly patients, who are the primary beneficiaries of AI-driven outcomes. This international, multicenter, cross-sectional study assessed the attitudes of hospital patients towards AI in healthcare across 43 countries. A total of 13806 patients at 74 hospitals were surveyed between February and November 2023, with 64.8% from the Global North and 35.2% from the Global South. The findings indicate a predominantly favorable general view of AI in healthcare, with 57.6% of respondents expressing a positive attitude. However, attitudes exhibited notable variation based on demographic characteristics, health status, and technological literacy. Female respondents and those with poorer health status exhibited fewer positive attitudes towards AI use in medicine. Conversely, higher levels of AI knowledge and frequent use of technology devices were associated with more positive attitudes. It is noteworthy that less than half of the participants expressed positive attitudes regarding all items pertaining to trust in AI. The lowest level of trust was observed for the accuracy of AI in providing information regarding treatment responses. Patients exhibited a strong preference for explainable AI and physician-led decision-making, even if it meant slightly compromised accuracy. This large-scale, multinational study provides a comprehensive perspective on patient attitudes towards AI in healthcare across six continents. Findings suggest a need for tailored AI implementation strategies that consider patient demographics, health status, and preferences for explainable AI and physician oversight. All study data has been made publicly available to encourage replication and further investigation.","PeriodicalId":501454,"journal":{"name":"medRxiv - Health Informatics","volume":"3 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142211500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：