Article
Mathematical & Computational Biology
Erin E. Gabriel, Michael C. Sachs, Torben Martinussen, Ingeborg Waernbaum, Els Goetghebeur, Stijn Vansteelandt, Arvid Sjolander
Summary: There is a concerning trend in the applied literature to believe that the combination of a propensity score and an adjusted outcome model automatically results in a doubly robust estimator and to misuse more complex established doubly robust estimators. This article introduces a simple alternative method and aims to help readers understand why it is doubly robust. It also provides examples and code for implementation.
STATISTICS IN MEDICINE
(2023)
Article
Computer Science, Interdisciplinary Applications
S. Karthick, N. Gomathi
Summary: The development of Internet of Medical Things (IoMT) with deep learning has opened up new possibilities in healthcare, particularly during the COVID-19 pandemic. This paper proposes RERNN-GEO, a deep learning method based on IoT, which utilizes the Gray-Level Co-Occurrence Matrix (GLCM) window adaptive algorithm for accurate diagnosis of COVID-19 using chest X-ray images.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Article
Computer Science, Interdisciplinary Applications
Indu Chhillar, Ajmer Singh
Summary: Lung and colon cancers are prevalent and lethal tumors. This study aims to develop an accurate and interpretable machine learning technique for automated classification of these cancers from histopathology images. Texture and color features are extracted using Haralick and Color histogram algorithms, then passed into the Light Gradient Boosting Machine (LightGBM) classifier. Experimental results show high accuracy and good performance of the proposed technique in classifying lung and colon cancer images.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Article
Health Care Sciences & Services
Samuel T. Savitz, Kristine Falk, Sally C. Stearns, Lexie R. Grove, Donald E. Pathman, Joseph S. Rossi
Summary: The existing literature on the differences in survival following percutaneous coronary intervention (PCI) based on patient sex, race-ethnicity, and socioeconomic characteristics (SEC) is limited. This study aimed to evaluate the differences in 1-year survival after PCI based on sex and race-ethnicity, and to explore the influence of SEC on these differences. The findings indicate that women are more likely to undergo PCI in the setting of acute myocardial infarction (AMI) and have less transition to outpatient care. Black patients experience higher 1-year mortality following PCI, which can be explained by differences in baseline comorbidities, county medical resources, and state of residence.
JOURNAL OF EVALUATION IN CLINICAL PRACTICE
(2023)
Review
Computer Science, Information Systems
Vishal Anand Shetty, Shauna Durbin, Meghan S. Weyrich, Airin Denise Martinez, Jing Qian, David L. Chin
Summary: This article provides a scoping review of studies on empathy recognition in text using natural language processing (NLP) and identifies an approach to identifying physician empathic communication over patient portal messages. The review finds variations in definitions of empathy and common settings where empathy is studied. NLP methods show promise in handling empathy-related tasks, but challenges remain in precisely defining and measuring empathy in text.
JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION
(2023)
Article
Health Care Sciences & Services
Lamin Juwara, Yi Archer Yang, Ana M. Velly, Paramita Saha-Chaudhuri
Summary: Analyses of distributed data networks of rare diseases are challenging due to privacy and ethical concerns. We propose a privacy-preserving data analysis technique by pooling individual records of covariates at recruiting sites. Our method shows good performance in simulations and analysis of real data.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2023)
Article
Mathematical & Computational Biology
Yasin Khadem Charvadeh, Grace Y. Yi
Summary: This article investigates the impact of ignoring misclassification in binary covariates on the determination of optimal decision rules in dynamic treatment regimes and demonstrates its negative effects on Q-learning through empirical studies. The authors propose two correction methods to address the misclassification effects and find that these methods successfully mitigate bias in parameter estimation.
STATISTICS IN MEDICINE
(2023)
Article
Computer Science, Interdisciplinary Applications
Kai Sun, Yuanjie Zheng, Xinbo Yang, Xinyuan Chen, Weikuan Jia
Summary: Breast cancer pathological image segmentation is important in quantifying tumor regions and providing treatment guidance. This study proposes a semi-supervised learning model based on classification-guided segmentation, which achieves outstanding segmentation performance.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Article
Computer Science, Interdisciplinary Applications
Yuansheng Ning, Hongbo Wang, Ying Liu, Qi Wang, Yu Rong, Jianye Niu
Summary: This paper proposes an upper limb exoskeleton rehabilitation robot system for shoulder-elbow-wrist joint rehabilitation training. The robot's model was guided by establishing a motion equivalent model, designing the robot mechanism configuration, and optimizing the structural parameters and human-machine compatibility. The results demonstrate that the rehabilitation robot can cover the motion range of upper limb joints, meet the needs of trajectory training, and maintain low human-machine interaction pressures during training.
MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING
(2023)
Review
Health Care Sciences & Services
Youn Sun Son, Ki Han Kwon
Summary: The combination of smart devices and big data enables personalized healthcare and promotes health, making it an important direction for the future healthcare industry.
Review
Health Care Sciences & Services
Liliane Zillner, Martin Andreas, Markus Mach
Summary: This systematic review highlights the potential of heart rate variability (HRV) and atrial fibrillation (AF) monitoring by wearable health monitoring devices as a diagnostic tool in cardiac surgery patients. HRV is found to be a predictor for sudden cardiac death, mortality after acute myocardial infarction (AMI), and postoperative atrial fibrillation (POAF). Continuous monitoring via wearables is also shown to be effective in capturing significant cardiac events that would otherwise have been missed.
Article
Computer Science, Information Systems
Ronak Etemadpour, Sonali Shintree, A. Duke Shereen
Summary: Visualization and visual analytic tools enhance data perception and aid decision making. Scatterplot matrices (SPLOM) are commonly used for visualizing multidimensional datasets, but they become unwieldy for high dimensional data. The Graph-theoretic Scatterplot Diagnostic (Scagnostics) method extracts a subset of scatterplots with salient features and manageable size to improve human decision making. In this study, EEG was used to observe brain activity during decision making based on scatterplots created with different visual measures.
JOURNAL OF HEALTHCARE INFORMATICS RESEARCH
(2023)
Article
Health Care Sciences & Services
Matteo Gadaleta, Patrick Harrington, Eric Barnhill, Evangelos Hytopoulos, Mintu P. Turakhia, Steven R. Steinhubl, Giorgio Quer
Summary: Early identification of atrial fibrillation (AF) is crucial in reducing the risk of serious cardiovascular outcomes. This study developed a model using a deep learning approach to predict the near-term risk of AF based on electrocardiogram (ECG) data. The model showed high accuracy in predicting AF over a two-week period, providing a potential digital strategy for improving diagnosis and treatment initiation.
NPJ DIGITAL MEDICINE
(2023)
Article
Mathematical & Computational Biology
An-Shun Tai, Sheng-Hsuan Lin
Summary: This study proposes a method of using multiply robust estimators to infer natural indirect effects in the presence of multiple ordered mediators. The method is shown to be robust to model misspecification and has been validated through simulations and an illustrative example.
STATISTICS IN MEDICINE
(2023)
Article
Mathematical & Computational Biology
Arvid Sjolander, Thomas Frisell, Sara Oberg, Yunzhang Wang, Sara Hagg
Summary: This article explores the combination of Mendelian randomization (MR) and sibling comparison design, discussing the feasibility and bias issues, and provides theoretical results and conclusions based on real data.
STATISTICS IN MEDICINE
(2023)
Article
Mathematical & Computational Biology
Dong Xi, Yao Chen
Summary: This article proposes an optimization algorithm for multiple objectives in clinical trials, focusing on the optimal weighted Bonferroni split. The study investigates the behavior of disjunctive power and conjunctive power as optimization objectives and proposes an efficient algorithm based on constrained nonlinear optimization and multiple starting points. The algorithm is applied to graphical approaches, providing a practical reference for optimal graphical strategies in clinical trials.
STATISTICS IN MEDICINE
(2023)
Correction
Medical Informatics
J. A. Andrews, M. P. Craven, A. R. Lang, B. Guo, R. Morriss, C. Hollis
BMC MEDICAL INFORMATICS AND DECISION MAKING
(2023)
Article
Medical Informatics
Qiang Fang, Yue Yang, Hao Wang, Hanxi Sun, Jiangming Chen, Zixiang Chen, Tian Pu, Xiaoqing Zhang, Fubao Liu
Summary: Liver cancer is the leading cause of mortality worldwide. In this paper, a novel neural network architecture called LCRNet is proposed to improve the accuracy of liver cancer diagnosis by considering global context features and building local dependencies between channels.
HEALTH INFORMATION SCIENCE AND SYSTEMS
(2023)
Review
Medical Informatics
Yan Zhang, Weiwei Xu, Ping Yang, An Zhang
Summary: Machine learning methods show relatively favorable accuracy in predicting mortality risk in sepsis patients, highlighting the need for updating prediction scoring systems based on existing machine learning approaches.
BMC MEDICAL INFORMATICS AND DECISION MAKING
(2023)
Article
Medical Informatics
Ziming Zhang, Wenjun Tan, Yuhang Sun, Juntao Han, Zhe Wang, Hongsheng Xue, Ruoyu Wang
Summary: Based on deep learning, this study addresses the problem of inconsistent brightness between frames in endoscopic 3D reconstruction, and introduces attention modules and inter-layer losses to handle the complexity of endoscopic scenes in clinical surgeries. The developed framework can better simulate the mapping relationship between adjacent frames during endoscope motion and exhibits good generalization performance.
HEALTH INFORMATION SCIENCE AND SYSTEMS
(2023)