Pub Date : 2025-01-02DOI: 10.1016/j.artmed.2024.103063
Inês Won Sampaio , Emma Tassi , Marcella Bellani , Francesco Benedetti , Igor Nenadić , Mary L. Phillips , Fabrizio Piras , Lakshmi Yatham , Anna Maria Bianchi , Paolo Brambilla , Eleonora Maggioni
The heterogeneity of psychiatric disorders makes researching disorder-specific neurobiological markers an ill-posed problem. Here, we face the need for disease stratification models by presenting a generalizable multivariate normative modelling framework for characterizing brain morphology, applied to bipolar disorder (BD). We used deep autoencoders in an anomaly detection framework, combined for the first time with a confounder removal step that integrates training and external validation.
The model was trained with healthy control (HC) data from the human connectome project and applied to multi-site external data of HC and BD individuals. We found that brain deviating scores were greater, more heterogeneous, and with increased extreme values in the BD group, with volumes prominently from the basal ganglia, hippocampus, and adjacent regions emerging as significantly deviating. Similarly, individual brain deviating maps based on modified z scores expressed higher abnormalities occurrences, but their overall spatial overlap was lower compared to HCs.
Our generalizable framework enabled the identification of brain deviating patterns differing between the subject and the group levels, a step forward towards the development of more effective and personalized clinical decision support systems and patient stratification in psychiatry.
{"title":"A generalizable normative deep autoencoder for brain morphological anomaly detection: application to the multi-site StratiBip dataset on bipolar disorder in an external validation framework","authors":"Inês Won Sampaio , Emma Tassi , Marcella Bellani , Francesco Benedetti , Igor Nenadić , Mary L. Phillips , Fabrizio Piras , Lakshmi Yatham , Anna Maria Bianchi , Paolo Brambilla , Eleonora Maggioni","doi":"10.1016/j.artmed.2024.103063","DOIUrl":"10.1016/j.artmed.2024.103063","url":null,"abstract":"<div><div>The heterogeneity of psychiatric disorders makes researching disorder-specific neurobiological markers an ill-posed problem. Here, we face the need for disease stratification models by presenting a generalizable multivariate normative modelling framework for characterizing brain morphology, applied to bipolar disorder (BD). We used deep autoencoders in an anomaly detection framework, combined for the first time with a confounder removal step that integrates training and external validation.</div><div>The model was trained with healthy control (HC) data from the human connectome project and applied to multi-site external data of HC and BD individuals. We found that brain deviating scores were greater, more heterogeneous, and with increased extreme values in the BD group, with volumes prominently from the basal ganglia, hippocampus, and adjacent regions emerging as significantly deviating. Similarly, individual brain deviating maps based on modified z scores expressed higher abnormalities occurrences, but their overall spatial overlap was lower compared to HCs.</div><div>Our generalizable framework enabled the identification of brain deviating patterns differing between the subject and the group levels, a step forward towards the development of more effective and personalized clinical decision support systems and patient stratification in psychiatry.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"161 ","pages":"Article 103063"},"PeriodicalIF":6.1,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143016909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-23DOI: 10.1016/j.artmed.2024.103025
Muhammad Ahsan, Robertas Damaševičius
The global burden of infectious diseases significantly affects mortality rates, with their varying symptoms making it challenging to assess and determine the severity of infections. Different countries face unique challenges related to these diseases. This study introduces innovative Artificial Intelligence (AI) based methodologies to enhance diagnostic accuracy through the analysis of medical imagery. It achieves this by developing a mathematical model capable of identifying potential infectious diseases from images, utilizing a Multi-Criteria Decision-Making (MCDM) framework. This cutting-edge approach combines Hypersoft Set (HSS) within a fuzzy context, pioneering in AI-driven diagnostic processes. The decision-making process might suggest actions such as isolation, quarantine in either domestic settings or specialized facilities, or admission to a hospital for further treatment. The use of visual aids in this research not only improves understanding but also highlights the effectiveness and significance of the proposed methods. The foundational theory and the results from this novel approach demonstrate its potential for widespread application in fields like machine learning, deep learning, and pattern recognition, indicating a significant stride in the fight against infectious diseases through advanced diagnostic techniques.
{"title":"Artificial intelligence-powered image analysis: A paradigm shift in infectious disease detection","authors":"Muhammad Ahsan, Robertas Damaševičius","doi":"10.1016/j.artmed.2024.103025","DOIUrl":"10.1016/j.artmed.2024.103025","url":null,"abstract":"<div><div>The global burden of infectious diseases significantly affects mortality rates, with their varying symptoms making it challenging to assess and determine the severity of infections. Different countries face unique challenges related to these diseases. This study introduces innovative Artificial Intelligence (AI) based methodologies to enhance diagnostic accuracy through the analysis of medical imagery. It achieves this by developing a mathematical model capable of identifying potential infectious diseases from images, utilizing a Multi-Criteria Decision-Making (MCDM) framework. This cutting-edge approach combines Hypersoft Set (HSS) within a fuzzy context, pioneering in AI-driven diagnostic processes. The decision-making process might suggest actions such as isolation, quarantine in either domestic settings or specialized facilities, or admission to a hospital for further treatment. The use of visual aids in this research not only improves understanding but also highlights the effectiveness and significance of the proposed methods. The foundational theory and the results from this novel approach demonstrate its potential for widespread application in fields like machine learning, deep learning, and pattern recognition, indicating a significant stride in the fight against infectious diseases through advanced diagnostic techniques.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103025"},"PeriodicalIF":6.1,"publicationDate":"2024-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142719592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-21DOI: 10.1016/j.artmed.2024.103032
Susanne Ibing , Julian Hugo , Florian Borchert , Linea Schmidt , Caroline Benson , Allison A. Marshall , Colleen Chasteau , Ujunwa Korie , Diana Paguay , Jan Philipp Sachs , Bernhard Y. Renard , Judy H. Cho , Erwin P. Böttinger , Ryan C. Ungaro
Background:
Early diagnosis and treatment of Crohn’s Disease are associated with decreased risk of surgery and complications. However, diagnostic delay is frequently seen in clinical practice. To better understand Crohn’s Disease risk factors and disease indicators, we identified, described, and predicted incident Crohn’s Disease patients based on the Electronic Health Record data of the Mount Sinai Health System.
Methods:
We developed two phenotyping algorithms based on structured Electronic Health Record data (i.e., coded diagnosis, medication prescription, and healthcare utilization), and a more simple and advanced approach of information extraction from clinical notes, including data between 2011 and 2023. We conducted an ablation study for the classification task using different models, prediction time points, data inputs, text encoding methods, and case-control matching variables.
Results:
We identified 247 incident Crohn’s Disease cases and 1221 matched controls and validated our cohorts through manual chart review. A second control cohort (n = 1235) was created without matching on race. Gastrointestinal symptoms were significantly overrepresented in cases at least 180 days before the first coded Crohn’s Disease diagnosis. Adding text-based features to the clinical prediction models increased their overall performances. However, adding race as a matching variable had more effects on the model performance than the choice of modeling algorithm or input data, with an area under the receiver operating characteristic difference of 0.09 between the best-performing models.
Conclusion:
We demonstrate the feasibility of identifying newly diagnosed Crohn’s Disease patients within a United States health system using Electronic Health Records. For the predictive modeling task, cases and controls were distinguished only with modest performance, even though various state-of-the-art methods were applied based on features from structured and unstructured data. Our findings suggest the benefit of adding information from clinical notes in a supervised or unsupervised manner for cohort creation and predictive modeling.
{"title":"Electronic Health Records-based identification of newly diagnosed Crohn’s Disease cases","authors":"Susanne Ibing , Julian Hugo , Florian Borchert , Linea Schmidt , Caroline Benson , Allison A. Marshall , Colleen Chasteau , Ujunwa Korie , Diana Paguay , Jan Philipp Sachs , Bernhard Y. Renard , Judy H. Cho , Erwin P. Böttinger , Ryan C. Ungaro","doi":"10.1016/j.artmed.2024.103032","DOIUrl":"10.1016/j.artmed.2024.103032","url":null,"abstract":"<div><h3>Background:</h3><div>Early diagnosis and treatment of Crohn’s Disease are associated with decreased risk of surgery and complications. However, diagnostic delay is frequently seen in clinical practice. To better understand Crohn’s Disease risk factors and disease indicators, we identified, described, and predicted incident Crohn’s Disease patients based on the Electronic Health Record data of the Mount Sinai Health System.</div></div><div><h3>Methods:</h3><div>We developed two phenotyping algorithms based on structured Electronic Health Record data (i.e., coded diagnosis, medication prescription, and healthcare utilization), and a more simple and advanced approach of information extraction from clinical notes, including data between 2011 and 2023. We conducted an ablation study for the classification task using different models, prediction time points, data inputs, text encoding methods, and case-control matching variables.</div></div><div><h3>Results:</h3><div>We identified 247 incident Crohn’s Disease cases and 1221 matched controls and validated our cohorts through manual chart review. A second control cohort (n = 1235) was created without matching on race. Gastrointestinal symptoms were significantly overrepresented in cases at least 180 days before the first coded Crohn’s Disease diagnosis. Adding text-based features to the clinical prediction models increased their overall performances. However, adding race as a matching variable had more effects on the model performance than the choice of modeling algorithm or input data, with an area under the receiver operating characteristic difference of 0.09 between the best-performing models.</div></div><div><h3>Conclusion:</h3><div>We demonstrate the feasibility of identifying newly diagnosed Crohn’s Disease patients within a United States health system using Electronic Health Records. For the predictive modeling task, cases and controls were distinguished only with modest performance, even though various state-of-the-art methods were applied based on features from structured and unstructured data. Our findings suggest the benefit of adding information from clinical notes in a supervised or unsupervised manner for cohort creation and predictive modeling.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103032"},"PeriodicalIF":6.1,"publicationDate":"2024-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703915","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-21DOI: 10.1016/j.artmed.2024.103024
K. Naveen Kumar , C. Krishna Mohan , Linga Reddy Cenkeramaddi , Navchetan Awasthi
The privacy-sensitive nature of medical image data is often bounded by strict data sharing regulations that necessitate the need for novel modeling and analysis techniques. Federated learning (FL) enables multiple medical institutions to collectively train a deep neural network without sharing sensitive patient information. In addition, FL uses its collaborative approach to address challenges related to the scarcity and non-uniform distribution of heterogeneous medical domain data. Nevertheless, the data-opaque nature and distributed setup make FL susceptible to data poisoning attacks. There are diverse FL data poisoning attacks for classification models on natural image data in the literature. But their primary focus is on the impact of the attack and they do not consider the attack budget and attack visibility. The attack budget is essential for adversaries to optimize resource utilization in real-world scenarios, which determines the number of manipulations or perturbations they can apply. Simultaneously, attack visibility is crucial to ensure covert execution, allowing attackers to achieve their objectives without triggering detection mechanisms. Generally, an attacker’s aim is to create maximum attack impact with minimal resources and low visibility. So, considering these three entities can effectively comprehend the adversary’s perspective in designing an attack for real-world scenarios. Further, data poisoning attacks on medical images are challenging compared to natural images due to the subjective nature of medical data. Hence, we develop an attack with a low budget, low visibility, and high impact for medical image classification in FL. We propose a federated learning attention guided minimal attack (FL-AGMA), that uses class attention maps to identify specific medical image regions for perturbation. We introduce image distortion degree (IDD) as a metric to assess the attack budget. Also, we develop a feedback mechanism to regulate the attack coefficient for low attack visibility. Later, we optimize the attack budget by adaptively changing the IDD based on attack visibility. We extensively evaluate three large-scale datasets, namely, Covid-chestxray, Camelyon17, and HAM10000, covering three different data modalities. We observe that our FL-AGMA method has resulted in 44.49% less test accuracy with only 24% of IDD attack budget and lower attack visibility compared to the other attacks.
{"title":"Minimal data poisoning attack in federated learning for medical image classification: An attacker perspective","authors":"K. Naveen Kumar , C. Krishna Mohan , Linga Reddy Cenkeramaddi , Navchetan Awasthi","doi":"10.1016/j.artmed.2024.103024","DOIUrl":"10.1016/j.artmed.2024.103024","url":null,"abstract":"<div><div>The privacy-sensitive nature of medical image data is often bounded by strict data sharing regulations that necessitate the need for novel modeling and analysis techniques. Federated learning (FL) enables multiple medical institutions to collectively train a deep neural network without sharing sensitive patient information. In addition, FL uses its collaborative approach to address challenges related to the scarcity and non-uniform distribution of heterogeneous medical domain data. Nevertheless, the data-opaque nature and distributed setup make FL susceptible to data poisoning attacks. There are diverse FL data poisoning attacks for classification models on natural image data in the literature. But their primary focus is on the impact of the attack and they do not consider the attack budget and attack visibility. The attack budget is essential for adversaries to optimize resource utilization in real-world scenarios, which determines the number of manipulations or perturbations they can apply. Simultaneously, attack visibility is crucial to ensure covert execution, allowing attackers to achieve their objectives without triggering detection mechanisms. Generally, an attacker’s aim is to create maximum attack impact with minimal resources and low visibility. So, considering these three entities can effectively comprehend the adversary’s perspective in designing an attack for real-world scenarios. Further, data poisoning attacks on medical images are challenging compared to natural images due to the subjective nature of medical data. Hence, we develop an attack with a low budget, low visibility, and high impact for medical image classification in FL. We propose a federated learning attention guided minimal attack (FL-AGMA), that uses class attention maps to identify specific medical image regions for perturbation. We introduce image distortion degree (IDD) as a metric to assess the attack budget. Also, we develop a feedback mechanism to regulate the attack coefficient for low attack visibility. Later, we optimize the attack budget by adaptively changing the IDD based on attack visibility. We extensively evaluate three large-scale datasets, namely, Covid-chestxray, Camelyon17, and HAM10000, covering three different data modalities. We observe that our FL-AGMA method has resulted in 44.49% less test accuracy with only 24% of IDD attack budget and lower attack visibility compared to the other attacks.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103024"},"PeriodicalIF":6.1,"publicationDate":"2024-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703916","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Time-to-event prediction, e.g. cancer survival analysis or hospital length of stay, is a highly prominent machine learning task in medical and healthcare applications. However, only a few interpretable machine learning methods comply with its challenges. To facilitate a comprehensive explanatory analysis of survival models, we formally introduce time-dependent feature effects and global feature importance explanations. We show how post-hoc interpretation methods allow for finding biases in AI systems predicting length of stay using a novel multi-modal dataset created from 1235 X-ray images with textual radiology reports annotated by human experts. Moreover, we evaluate cancer survival models beyond predictive performance to include the importance of multi-omics feature groups based on a large-scale benchmark comprising 11 datasets from The Cancer Genome Atlas (TCGA). Model developers can use the proposed methods to debug and improve machine learning algorithms, while physicians can discover disease biomarkers and assess their significance. We contribute open data and code resources to facilitate future work in the emerging research direction of explainable survival analysis.
时间到事件预测，例如癌症存活率分析或住院时间预测，是医疗保健应用中一项非常突出的机器学习任务。然而，只有少数可解释的机器学习方法能应对其挑战。为了促进生存模型的综合解释分析，我们正式引入了时间依赖特征效应和全局特征重要性解释。我们展示了如何利用事后解释方法发现人工智能系统在预测住院时间方面存在的偏差，该方法使用了一个新颖的多模态数据集，该数据集由 1235 张 X 光图像和人类专家注释的文本放射学报告创建而成。此外，我们对癌症生存模型进行了评估，除了预测性能外，还包括基于癌症基因组图谱（TCGA）11 个数据集的大规模基准的多组学特征组的重要性。模型开发人员可以利用提出的方法调试和改进机器学习算法，而医生则可以发现疾病生物标志物并评估其重要性。我们提供开放的数据和代码资源，以促进可解释生存分析这一新兴研究方向的未来工作。
{"title":"Interpretable machine learning for time-to-event prediction in medicine and healthcare","authors":"Hubert Baniecki , Bartlomiej Sobieski , Patryk Szatkowski , Przemyslaw Bombinski , Przemyslaw Biecek","doi":"10.1016/j.artmed.2024.103026","DOIUrl":"10.1016/j.artmed.2024.103026","url":null,"abstract":"<div><div>Time-to-event prediction, e.g. cancer survival analysis or hospital length of stay, is a highly prominent machine learning task in medical and healthcare applications. However, only a few interpretable machine learning methods comply with its challenges. To facilitate a comprehensive explanatory analysis of survival models, we formally introduce time-dependent feature effects and global feature importance explanations. We show how post-hoc interpretation methods allow for finding biases in AI systems predicting length of stay using a novel multi-modal dataset created from 1235 X-ray images with textual radiology reports annotated by human experts. Moreover, we evaluate cancer survival models beyond predictive performance to include the importance of multi-omics feature groups based on a large-scale benchmark comprising 11 datasets from The Cancer Genome Atlas (TCGA). Model developers can use the proposed methods to debug and improve machine learning algorithms, while physicians can discover disease biomarkers and assess their significance. We contribute open data and code resources to facilitate future work in the emerging research direction of explainable survival analysis.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103026"},"PeriodicalIF":6.1,"publicationDate":"2024-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142696234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-20DOI: 10.1016/j.artmed.2024.103027
Jinghui Liu , Bevan Koopman , Nathan J. Brown , Kevin Chu , Anthony Nguyen
Large language models (LLMs) demonstrate impressive capabilities in generating human-like content and have much potential to improve the performance and efficiency of healthcare. An important application of LLMs is to generate synthetic clinical reports that could alleviate the burden of annotating and collecting real-world data in training AI models. Meanwhile, there could be concerns and limitations in using commercial LLMs to handle sensitive clinical data. In this study, we examined the use of open-source LLMs as an alternative to generate synthetic radiology reports to supplement real-world annotated data. We found LLMs hosted locally can achieve similar performance compared to ChatGPT and GPT-4 in augmenting training data for the downstream report classification task of identifying misdiagnosed fractures. We also examined the predictive value of using synthetic reports alone for training downstream models, where our best setting achieved more than 90 % of the performance using real-world data. Overall, our findings show that open-source, local LLMs can be a favourable option for creating synthetic clinical reports for downstream tasks.
{"title":"Generating synthetic clinical text with local large language models to identify misdiagnosed limb fractures in radiology reports","authors":"Jinghui Liu , Bevan Koopman , Nathan J. Brown , Kevin Chu , Anthony Nguyen","doi":"10.1016/j.artmed.2024.103027","DOIUrl":"10.1016/j.artmed.2024.103027","url":null,"abstract":"<div><div>Large language models (LLMs) demonstrate impressive capabilities in generating human-like content and have much potential to improve the performance and efficiency of healthcare. An important application of LLMs is to generate synthetic clinical reports that could alleviate the burden of annotating and collecting real-world data in training AI models. Meanwhile, there could be concerns and limitations in using commercial LLMs to handle sensitive clinical data. In this study, we examined the use of open-source LLMs as an alternative to generate synthetic radiology reports to supplement real-world annotated data. We found LLMs hosted locally can achieve similar performance compared to ChatGPT and GPT-4 in augmenting training data for the downstream report classification task of identifying misdiagnosed fractures. We also examined the predictive value of using synthetic reports alone for training downstream models, where our best setting achieved more than 90 % of the performance using real-world data. Overall, our findings show that open-source, local LLMs can be a favourable option for creating synthetic clinical reports for downstream tasks.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103027"},"PeriodicalIF":6.1,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142703914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-20DOI: 10.1016/j.artmed.2024.103031
Gennaro Percannella, Umberto Petruzzello, Francesco Tortorella, Mario Vento
Antinuclear Antibody (ANA) testing is pivotal to help diagnose patients with a suspected autoimmune disease. The Indirect Immunofluorescence (IIF) microscopy performed with human epithelial type 2 (HEp-2) cells as the substrate is the reference method for ANA screening. It allows for the detection of antibodies binding to specific intracellular targets, resulting in various staining patterns that should be identified for diagnosis purposes. In recent years, there has been an increasing interest in devising deep learning methods for automated cell segmentation and classification of staining patterns, as well as for other tasks related to this diagnostic technique (such as intensity classification). However, little attention has been devoted to architectures aimed at simultaneously managing multiple interrelated tasks, via a shared representation.
In this paper, we propose a deep neural network model that extends U-Net in a Multi-Task Learning (MTL) fashion, thus offering an end-to-end approach to tackle three fundamental tasks of the diagnostic procedure, i.e., HEp-2 cell specimen intensity classification, specimen segmentation, and pattern classification. The experiments were conducted on one of the largest publicly available datasets of HEp-2 images. The results showed that the proposed approach significantly outperformed the competing state-of-the-art methods for all the considered tasks.
{"title":"A Multi-task learning U-Net model for end-to-end HEp-2 cell image analysis","authors":"Gennaro Percannella, Umberto Petruzzello, Francesco Tortorella, Mario Vento","doi":"10.1016/j.artmed.2024.103031","DOIUrl":"10.1016/j.artmed.2024.103031","url":null,"abstract":"<div><div>Antinuclear Antibody (<em>ANA</em>) testing is pivotal to help diagnose patients with a suspected autoimmune disease. The Indirect Immunofluorescence (<em>IIF</em>) microscopy performed with human epithelial type 2 (HEp-2) cells as the substrate is the reference method for ANA screening. It allows for the detection of antibodies binding to specific intracellular targets, resulting in various staining patterns that should be identified for diagnosis purposes. In recent years, there has been an increasing interest in devising deep learning methods for automated cell segmentation and classification of staining patterns, as well as for other tasks related to this diagnostic technique (such as intensity classification). However, little attention has been devoted to architectures aimed at simultaneously managing multiple interrelated tasks, <em>via</em> a shared representation.</div><div>In this paper, we propose a deep neural network model that extends U-Net in a Multi-Task Learning (MTL) fashion, thus offering an end-to-end approach to tackle three fundamental tasks of the diagnostic procedure, i.e., HEp-2 cell specimen intensity classification, specimen segmentation, and pattern classification. The experiments were conducted on one of the largest publicly available datasets of HEp-2 images. The results showed that the proposed approach significantly outperformed the competing state-of-the-art methods for all the considered tasks.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103031"},"PeriodicalIF":6.1,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142748331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-20DOI: 10.1016/j.artmed.2024.103029
Hong Wang , Luhe Zhuang , Yijie Ding , Prayag Tiwari , Cheng Liang
Predicting drug–drug interactions (DDIs) is crucial for understanding and preventing adverse drug reactions (ADRs). However, most existing methods inadequately explore the interactive information between drugs in a self-supervised manner, limiting our comprehension of drug–drug associations. This paper introduces EDDINet: Enhancing Drug-Drug Interaction Prediction via Information Flow and Consensus-Constrained Multi-Graph Contrastive Learning for precise DDI prediction. We first present a cross-modal information-flow mechanism to integrate diverse drug features, enriching the structural insights conveyed by the drug feature vector. Next, we employ contrastive learning to filter various biological networks, enhancing the model’s robustness. Additionally, we propose a consensus regularization framework that collaboratively trains multi-view models, producing high-quality drug representations. To unify drug representations derived from different biological information, we utilize an attention mechanism for DDI prediction. Extensive experiments demonstrate that EDDINet surpasses state-of-the-art unsupervised models and outperforms some supervised baseline models in DDI prediction tasks. Our approach shows significant advantages and holds promising potential for advancing DDI research and improving drug safety assessments. Our codes are available at: https://github.com/95LY/EDDINet_code.
预测药物间相互作用（DDI）对于了解和预防药物不良反应（ADR）至关重要。然而，大多数现有方法都没有以自我监督的方式充分探索药物之间的交互信息，从而限制了我们对药物关联的理解。本文介绍了 EDDINet：EDDINet: Enhancing Drug-Drug Interaction Prediction via Information Flow and Consensus-Constrained Multi-Graph Contrastive Learning（通过信息流和共识约束多图对比学习增强药物间相互作用预测），用于精确的 DDI 预测。我们首先提出了一种跨模态信息流机制，用于整合不同的药物特征，丰富药物特征向量所传达的结构洞察力。接下来，我们利用对比学习过滤各种生物网络，增强了模型的鲁棒性。此外，我们还提出了一种共识正则化框架，可协同训练多视角模型，从而生成高质量的药物表征。为了统一来自不同生物信息的药物表征，我们利用注意力机制进行 DDI 预测。广泛的实验证明，EDDINet 在 DDI 预测任务中超越了最先进的无监督模型，并优于一些有监督基线模型。我们的方法显示出显著的优势，在推进 DDI 研究和改进药物安全性评估方面具有广阔的前景。我们的代码可在以下网址获取：https://github.com/95LY/EDDINet_code。
{"title":"EDDINet: Enhancing drug–drug interaction prediction via information flow and consensus constrained multi-graph contrastive learning","authors":"Hong Wang , Luhe Zhuang , Yijie Ding , Prayag Tiwari , Cheng Liang","doi":"10.1016/j.artmed.2024.103029","DOIUrl":"10.1016/j.artmed.2024.103029","url":null,"abstract":"<div><div>Predicting drug–drug interactions (DDIs) is crucial for understanding and preventing adverse drug reactions (ADRs). However, most existing methods inadequately explore the interactive information between drugs in a self-supervised manner, limiting our comprehension of drug–drug associations. This paper introduces <strong>EDDINet</strong>: <strong>E</strong>nhancing <strong>D</strong>rug-<strong>D</strong>rug <strong>I</strong>nteraction Prediction via Information Flow and Consensus-Constrained Multi-Graph Contrastive Learning for precise DDI prediction. We first present a cross-modal information-flow mechanism to integrate diverse drug features, enriching the structural insights conveyed by the drug feature vector. Next, we employ contrastive learning to filter various biological networks, enhancing the model’s robustness. Additionally, we propose a consensus regularization framework that collaboratively trains multi-view models, producing high-quality drug representations. To unify drug representations derived from different biological information, we utilize an attention mechanism for DDI prediction. Extensive experiments demonstrate that EDDINet surpasses state-of-the-art unsupervised models and outperforms some supervised baseline models in DDI prediction tasks. Our approach shows significant advantages and holds promising potential for advancing DDI research and improving drug safety assessments. Our codes are available at: <span><span>https://github.com/95LY/EDDINet_code</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103029"},"PeriodicalIF":6.1,"publicationDate":"2024-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142719593","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-18DOI: 10.1016/j.artmed.2024.103028
Pei-Yan Li , Yu-Wen Huang , Vin-Cent Wu , Jeff S. Chueh , Chi-Shin Tseng , Chung-Ming Chen
Background and objective
Predicting postoperative prognosis is vital for clinical decision making in patients undergoing adrenalectomy (ADX). This study introduced GAPPA, a novel GNN-based approach, to predict post-ADX outcomes in patients with unilateral primary aldosteronism (UPA). The objective was to leverage the intricate dependencies between clinico-biochemical features and clinical outcomes using GNNs integrated into a bipartite graph structure to enhance prognostic prediction accuracy.
Methods
We conceptualized prognostic prediction as a link prediction task on a bipartite graph, with nodes representing patients, clinico-biochemical features, and clinical outcomes, and edges denoting the connections between them. GAPPA utilizes GNNs to capture these dependencies and seamlessly integrates the outcome predictions into a graph structure. This approach was evaluated using a dataset of 640 patients with UPA who underwent unilateral ADX (uADX) between 1990 and 2022. We conducted a comparative analysis using repeated stratified five-fold cross-validation and paired t-tests to evaluate the performance of GAPPA against conventional machine learning methods and previous studies across various metrics.
Results
GAPPA significantly outperformed conventional machine learning methods and previous studies (p < 0.05) across various metrics. It achieved F1-score, accuracy, sensitivity, and specificity of 71.3 % ± 3.1 %, 71.1 % ± 3.4 %, 69.9 % ± 4.3 %, and 72.4 % ± 7.2 %, respectively, with an AUC of 0.775 ± 0.030. We also investigated the impact of different initialization schemes on GAPPA outcome-edge embeddings, highlighting their robustness and stability.
Conclusion
GAPPA aids in preoperative prognosis assessment and facilitates patient counseling, contributing to prognostic prediction and advancing the applications of GNNs in the biomedical domain.
{"title":"GAPPA: Enhancing prognosis prediction in primary aldosteronism post-adrenalectomy using graph-based modeling","authors":"Pei-Yan Li , Yu-Wen Huang , Vin-Cent Wu , Jeff S. Chueh , Chi-Shin Tseng , Chung-Ming Chen","doi":"10.1016/j.artmed.2024.103028","DOIUrl":"10.1016/j.artmed.2024.103028","url":null,"abstract":"<div><h3>Background and objective</h3><div>Predicting postoperative prognosis is vital for clinical decision making in patients undergoing adrenalectomy (ADX). This study introduced GAPPA, a novel GNN-based approach, to predict post-ADX outcomes in patients with unilateral primary aldosteronism (UPA). The objective was to leverage the intricate dependencies between clinico-biochemical features and clinical outcomes using GNNs integrated into a bipartite graph structure to enhance prognostic prediction accuracy.</div></div><div><h3>Methods</h3><div>We conceptualized prognostic prediction as a link prediction task on a bipartite graph, with nodes representing patients, clinico-biochemical features, and clinical outcomes, and edges denoting the connections between them. GAPPA utilizes GNNs to capture these dependencies and seamlessly integrates the outcome predictions into a graph structure. This approach was evaluated using a dataset of 640 patients with UPA who underwent unilateral ADX (uADX) between 1990 and 2022. We conducted a comparative analysis using repeated stratified five-fold cross-validation and paired <em>t</em>-tests to evaluate the performance of GAPPA against conventional machine learning methods and previous studies across various metrics.</div></div><div><h3>Results</h3><div>GAPPA significantly outperformed conventional machine learning methods and previous studies (<em>p</em> < 0.05) across various metrics. It achieved F1-score, accuracy, sensitivity, and specificity of 71.3 % ± 3.1 %, 71.1 % ± 3.4 %, 69.9 % ± 4.3 %, and 72.4 % ± 7.2 %, respectively, with an AUC of 0.775 ± 0.030. We also investigated the impact of different initialization schemes on GAPPA outcome-edge embeddings, highlighting their robustness and stability.</div></div><div><h3>Conclusion</h3><div>GAPPA aids in preoperative prognosis assessment and facilitates patient counseling, contributing to prognostic prediction and advancing the applications of GNNs in the biomedical domain.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103028"},"PeriodicalIF":6.1,"publicationDate":"2024-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142696233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-11-17DOI: 10.1016/j.artmed.2024.103023
Qiao Ning , Yue Wang , Yaomiao Zhao , Jiahao Sun , Lu Jiang , Kaidi Wang , Minghao Yin
Accurate identification of drug-target interactions (DTIs) plays a crucial role in drug discovery. Compared with traditional experimental methods that are labor-intensive and time-consuming, computational methods for drug-target interactions prediction are more popular in recent years. Conventional computational methods almost simply view heterogeneous network constructed by the drug-related and protein-related dataset instead of comprehensively exploring drug-protein pair (DPP) information. To address this limitation, we proposed a Double Multi-view Heterogeneous Graph Neural Network framework for drug-target interaction prediction (DMHGNN). In DMHGNN, one multi-view heterogeneous graph neural network is based on meta-paths and denoising autoencoder for protein-, drug-related heterogeneous network learning, and another multi-view heterogeneous graph neural network is based on multi-channel graph convolutional network for drug-protein pair similarity network learning. First, a meta-path-based graph encoder with the attention mechanism is used for substructure learning of complex relationships from heterogeneous network constructed by proteins, drugs, side-effects and diseases, obtaining key information that is easy to be ignored in global learning of heterogeneous networks, and multi-source neighbouring features for drugs and proteins are learned from heterogeneous network via denoising auto-encoder model. Then, multi-view graphs of drug-protein pairs (DPPs) including the topology graph, semantics graph and collaborative graph with shared weights are constructed, and the multi-channel graph convolutional network (GCN) is utilized to learn the deep representation of DPPs. Finally, a multi-layer fully connection network is trained to predict drug-target interactions. Experiments have demonstrated its effectiveness and better performance than state-of-the-art methods.
{"title":"DMHGNN: Double multi-view heterogeneous graph neural network framework for drug-target interaction prediction","authors":"Qiao Ning , Yue Wang , Yaomiao Zhao , Jiahao Sun , Lu Jiang , Kaidi Wang , Minghao Yin","doi":"10.1016/j.artmed.2024.103023","DOIUrl":"10.1016/j.artmed.2024.103023","url":null,"abstract":"<div><div>Accurate identification of drug-target interactions (DTIs) plays a crucial role in drug discovery. Compared with traditional experimental methods that are labor-intensive and time-consuming, computational methods for drug-target interactions prediction are more popular in recent years. Conventional computational methods almost simply view heterogeneous network constructed by the drug-related and protein-related dataset instead of comprehensively exploring drug-protein pair (DPP) information. To address this limitation, we proposed a <strong>D</strong>ouble <strong>M</strong>ulti-view <strong>H</strong>eterogeneous <strong>G</strong>raph <strong>N</strong>eural <strong>N</strong>etwork framework for drug-target interaction prediction (DMHGNN). In DMHGNN, one multi-view heterogeneous graph neural network is based on meta-paths and denoising autoencoder for protein-, drug-related heterogeneous network learning, and another multi-view heterogeneous graph neural network is based on multi-channel graph convolutional network for drug-protein pair similarity network learning. First, a meta-path-based graph encoder with the attention mechanism is used for substructure learning of complex relationships from heterogeneous network constructed by proteins, drugs, side-effects and diseases, obtaining key information that is easy to be ignored in global learning of heterogeneous networks, and multi-source neighbouring features for drugs and proteins are learned from heterogeneous network via denoising auto-encoder model. Then, multi-view graphs of drug-protein pairs (DPPs) including the topology graph, semantics graph and collaborative graph with shared weights are constructed, and the multi-channel graph convolutional network (GCN) is utilized to learn the deep representation of DPPs. Finally, a multi-layer fully connection network is trained to predict drug-target interactions. Experiments have demonstrated its effectiveness and better performance than state-of-the-art methods.</div></div>","PeriodicalId":55458,"journal":{"name":"Artificial Intelligence in Medicine","volume":"159 ","pages":"Article 103023"},"PeriodicalIF":6.1,"publicationDate":"2024-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142696232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号