Deep Fusion Network with Feature Engineering for Discharge Risk Assessment

Communications in computer and information science, Год журнала: 2025, Номер unknown, С. 21 - 36

Опубликована: Янв. 1, 2025

Язык: Английский

---

A teaching proposal for a short course on biomedical data science

PLoS Computational Biology, Год журнала: 2025, Номер 21(4), С. e1012946 - e1012946

Опубликована: Апрель 14, 2025

As the availability of big biomedical data advances, there is a growing need university students trained professionally on analyzing these and correctly interpreting their results. We propose here study plan for master’s degree course science, by describing our experience during last academic year. In course, we explained how to find an open dataset, clean it prepare computational statistics or machine learning phase. By doing so, introduce common health science terms avoid mistakes in process. Moreover, clarified perform exploratory analysis (EDA) reasonably interpret its also described properly execute supervised unsupervised analysis, now understand outcomes. Eventually, validate findings obtained. illustrated all steps context principles, suggesting use only source programming languages (R Python particular), (if available), access scientific articles possible). believe teaching proposal can be useful interest anyone wanting start science.

Язык: Английский

---

Explanation and elaboration of MedinAI: guidelines for reporting artificial intelligence studies in medicines, pharmacotherapy, and pharmaceutical services

International Journal of Clinical Pharmacy, Год журнала: 2025, Номер unknown

Опубликована: Апрель 18, 2025

Язык: Английский

---

Investigation on the interpretable machine-learning model for predicting mechanical properties of DH auto-steel plates

Materials Today Communications, Год журнала: 2025, Номер unknown, С. 112674 - 112674

Опубликована: Апрель 1, 2025

Язык: Английский

---

RequestAtlas: Supporting the Slow and Iterative Process of Requesting Public Records

Proceedings of the ACM on Human-Computer Interaction, Год журнала: 2025, Номер 9(2), С. 1 - 35

Опубликована: Май 2, 2025

Public records requests are a central mechanism for government transparency. In practice, they slow, complex processes that require analyzing large amounts of messy, unstructured data. this paper, we introduce RequestAtlas, system helps investigative journalists review quantities data result from submitting many public requests. RequestAtlas was developed through year-long participatory design collaboration with the California Reporting Project (CRP), journalistic collective researching police use force and misconduct in California. evaluate results completeness negotiate agencies additional information. has had significant real-world impact. It been deployed more than year to identify missing response facilitate negotiation request officers. Through process designing observing explore technical challenges associated needs generally. We argue represent an instance adversarial relationship which two entities engage prolonged, iterative, often exchange Technologists can support information-gathering efforts within these relationships by building flexible local solutions help both account state ongoing information exchange. Additionally, offer insights on ways applications assist inevitably cleaning phase processing documents while supporting norms verification human review. Finally, reflect process, despite its success, lays bare some limitations inherent ''request respond'' model transparency

Язык: Английский

---

Predicting and Analyzing the Immobilization of MgAl and CaAl Layered Double Hydroxides for Cr(VI) in Polluted Soil Using Machine Learning

Colloids and Surfaces A Physicochemical and Engineering Aspects, Год журнала: 2025, Номер 720, С. 137138 - 137138

Опубликована: Май 6, 2025

Язык: Английский

---

Hyperdimensional computing in biomedical sciences: a brief review

PeerJ Computer Science, Год журнала: 2025, Номер 11, С. e2885 - e2885

Опубликована: Май 13, 2025

Hyperdimensional computing (HDC, also known as vector-symbolic architectures—VSA) is an emerging computational paradigm that relies on dealing with vectors in a high-dimensional space to represent and combine every kind of information. It finds applications wide array fields including bioinformatics, natural language processing, machine learning, artificial intelligence, many other scientific disciplines. Here we introduced the basic foundations HDC, focusing its application biomedical sciences, particular emphasis cheminformatics, medical informatics, providing critical comprehensive review current HDC landscape, highlighting pros cons applying this these specific domains. In study, first selected around forty articles hyperdimensional applied data existing literature, then analyzed key aspects their studies, such vector construction, encoding, programming employed, features. We counted how are open access, have public software code available, groups authors, journals, conferences most present among them. Finally, discussed advantages limitations approach, outlining potential future directions challenges for adoption sciences. To best our knowledge, brief survey topic therefore believe it can be interest useful readership.

Язык: Английский

---

Development and application of an early prediction model for risk of bloodstream infection based on real-world study

BMC Medical Informatics and Decision Making, Год журнала: 2025, Номер 25(1)

Опубликована: Май 14, 2025

Bloodstream Infection (BSI) is a severe systemic infectious disease that can lead to sepsis and Multiple Organ Dysfunction Syndrome (MODS), resulting in high mortality rates posing major public health burden globally. Early identification of BSI crucial for effective intervention, reducing mortality, improving patient outcomes. However, existing diagnostic methods are flawed by low specificity, long detection times demands on testing platforms. The development artificial intelligence provides new approach early identification. This study aims explore the optimal combination routine laboratory data clinical monitoring indicators, utilize machine learning algorithms construct an early, rapid, universally applicable risk prediction model, assist diagnosis practice. Clinical 2582 suspected patients admitted Chongqing University Central Hospital, from January 1, 2021 December 31, 2023 were collected this study. divided into modeling dataset external validation based chronological order, while was further training set internal set. occurrence rate BSI, distribution pathogens, microbial primary reporting time analyzed within During feature selection stage, univariate regression ML applied. First, Univariate logistic used screen predictive factors BSI. Then, Boruta algorithm, Lasso regression, Recursive Feature Elimination with Cross-validation (RFE-CV) employed determine predictors predicting Based combination, six model. best model selected models' performance, Shapley Additive Explanations (SHAP) method explain evaluate performance generalizability research findings ultimately applied incidence among inpatients at Hospital 12.91%. Following selection, 5 variables determined, including white blood cell count, standard bicarbonate, base excess extracellular fluid, interleukin-6, body temperature. models constructed using algorithms, XGBoost demonstrating achieving AUC value 0.782 0.776 made publicly available as online webpage tool use. successfully identified features analyzing indicators hospitalized patients. set, learning-based constructed. capable rapid differentiation between non-BSI inclusion minimal enhances its applicability settings, particularly care level. To improve model's real-world more convenient use, application could greatly efficiency patients' mortality.

Язык: Английский

---

Cross-Chain Abnormal Account Detection

Опубликована: Янв. 1, 2025

Язык: Английский

---

Clinical Feature Ranking Based on Ensemble Machine Learning Reveals Top Survival Factors for Glioblastoma Multiforme

Journal of Healthcare Informatics Research, Год журнала: 2023, Номер 8(1), С. 1 - 18

Опубликована: Сен. 20, 2023

Abstract Glioblastoma multiforme (GM) is a malignant tumor of the central nervous system considered to be highly aggressive and often carrying terrible survival prognosis. An accurate prognosis therefore pivotal for deciding good treatment plan patients. In this context, computational intelligence applied data electronic health records (EHRs) patients diagnosed with disease can useful predict patients’ time. study, we evaluated different machine learning models time in suffering from glioblastoma further investigated which features were most predictive We our methods three independent open datasets EHRs glioblastoma: Shieh dataset 84 patients, Berendsen 647 Lammer 60 Our prediction techniques obtained concordance index (C-index) = 0.583 dataset, C-index 0.776 0.64 as best results each dataset. Since original studies regarding analyzed here did not provide insights about clinical time, feature importance among these datasets. To end, then utilized Random Survival Forests, decision tree-based algorithm able model non-linear interaction between might better capture complex genetic status discoveries impact practice, aiding clinicians alike decide therapy suited their unique status.

Язык: Английский

---