From machine learning to deep learning: Advances of the recent data-driven paradigm shift in medicine and healthcare

Current Research in Biotechnology, Год журнала: 2023, Номер 7, С. 100164 - 100164

Опубликована: Ноя. 22, 2023

The medicine and healthcare sector has been evolving advancing very fast. advancement initiated shaped by the applications of data-driven, robust, efficient machine learning (ML) to deep (DL) technologies. ML in medical is developing quickly, causing rapid progress, reshaping medicine, improving clinician patient experiences. technologies evolved into data-hungry DL approaches, which are more robust dealing with data. This article reviews some critical data-driven aspects intelligence field. In this direction, illustrated recent progress science using two categories: firstly, development data uses and, secondly, Chabot particularly on ChatGPT. Here, we discuss ML, DL, transition requirements from DL. To science, illustrate prospective studies image data, newly interpretation EMR or EHR, big personalized dataset shifts artificial (AI). Simultaneously, recently developed DL-enabled ChatGPT technology. Finally, summarize broad role significant challenges for implementing healthcare. overview paradigm shift will benefit researchers immensely.

Язык: Английский

---

Medications and cognitive decline in Alzheimer's disease: Cohort cluster analysis of 15,428 patients

Journal of Alzheimer s Disease, Год журнала: 2025, Номер unknown

Опубликована: Янв. 8, 2025

Background Medications for comorbid conditions may affect cognition in Alzheimer's disease (AD). Objective To explore the association between common medications and cognition, measured with Mini-Mental State Examination. Methods Cohort study including persons AD from Swedish Registry Cognitive/Dementia Disorders (SveDem). were included if they used by ≥5% of patients (26 individual drugs). Each follow-up was analyzed independently performing 100 Monte-Carlo simulations two steps each 1) k-means clustering according to Examination at its decline since previous measure, 2) Identification presenting statistically significant differences proportion users different clusters. Results 15,428 (60.38% women) studied. Four clusters identified. associated best cluster (relative worse) atorvastatin (point estimate 1.44 95% confidence interval [1.15–1.83] first follow-up, simvastatin (1.41 [1.11–1.78] second follow-up), warfarin (1.56 [1.22–2.01] zopiclone (1.35 [1.15–1.58], metformin (2.08 [1.35–3.33] follow-up. Oxazepam (0.60 [0.50–0.73] paracetamol (0.83 [0.73–0.95] cyanocobalamin, felodipine furosemide worst cluster. Cholinesterase inhibitors clusters, whereas memantine appeared worse consistent indication moderate severe dementia. Conclusions We performed unsupervised classify based on their current cognitive testing. Atorvastatin, simvastatin, warfarin, metformin, presented a positive associations while oxazepam, felodipine, paracetamol,

Язык: Английский

---

The Evolving Landscape of Artificial Intelligence Applications in Animal Health

Indian Journal of Animal Research, Год журнала: 2024, Номер Of

Опубликована: Фев. 13, 2024

Background: This work explores the expansivetab realm of Artificial Intelligence (AI) applications in dynamic landscape animal health and veterinary sciences. Addressing challenges conventional approaches, we delve into how AI is transforming diagnosis, treatment healthcare practices for diverse species. Methods: Through a rigorous literature review methodology, study navigates current state health, identifying gaps emphasizing need further research. Looking ahead, paper outlines future directions opportunities, contributing to discourse on technology’s intersection with care. By providing comprehensive overview, this research paves way innovative solutions, promising brighter healthier our companions. Result: In domain emerges as powerful tool early disease detection intervention, offering personalized plans proactive management through continuous monitoring surveillance. sciences, accelerates drug discovery, enhances genetic reshapes surgical procedures robotic assistance. However, ethical considerations challenges, including data privacy AI-driven decision-making critical examination should be addressed to.

Язык: Английский

---

Machine learning in onco-pharmacogenomics: a path to precision medicine with many challenges

Frontiers in Pharmacology, Год журнала: 2024, Номер 14

Опубликована: Янв. 9, 2024

Over the past two decades, Next-Generation Sequencing (NGS) has revolutionized approach to cancer research. Applications of NGS include identification tumor specific alterations that can influence pathobiology and also impact diagnosis, prognosis therapeutic options. Pharmacogenomics (PGx) studies role inheritance individual genetic patterns in drug response taken advantage technology as it provides access high-throughput data can, however, be difficult manage. Machine learning (ML) recently been used life sciences discover hidden from complex solve various PGx problems. In this review, we provide a comprehensive overview approaches employed different implicating use data. We an excursus ML algorithms exert fundamental strategies field improve personalized medicine cancer.

Язык: Английский

---

Identifying diseases symptoms and general rules using supervised and unsupervised machine learning

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Авг. 2, 2024

The symptoms of diseases can vary among individuals and may remain undetected in the early stages. Detecting these is crucial initial stage to effectively manage treat cases varying severity. Machine learning has made major advances recent years, proving its effectiveness various healthcare applications. This study aims identify patterns general rules regarding patients using supervised unsupervised machine learning. integration a rule-based technique classification methods utilized extend prediction model. analyzes patient data that was available online through Kaggle repository. After preprocessing exploring descriptive statistics, Apriori algorithm applied frequent discovered rules. Additionally, several models for predicting diseases, including stepwise regression, support vector machine, bootstrap forest, boosted trees, neural-boosted methods. Several predictive were dataset predict diseases. It method fitting outperformed all competitors this study, as determined cross-validation conducted each model based on established criteria. Moreover, numerous significant decision extracted which streamline clinical applications without need additional expertise. These enable relationships between well different Therefore, results obtained have potential improve performance models. We discover dataset. Overall, proposed not only professionals but also who face cost time constraints diagnosing treating

Язык: Английский

---

Acrylic acid modified indapamide-based polymer as an effective inhibitor against carbon steel corrosion in CO2-saturated NaCl with variable H2S levels: An electrochemical, weight loss and machine learning study

Surfaces and Interfaces, Год журнала: 2024, Номер 53, С. 105065 - 105065

Опубликована: Сен. 6, 2024

Язык: Английский

---

Predicting loss of independence among geriatric patients following gastrointestinal surgery

Patient Safety in Surgery, Год журнала: 2025, Номер 19(1)

Опубликована: Янв. 9, 2025

Abstract Background While existing risk calculators focus on mortality and complications, elderly patients are concerned with how operations will affect their quality of life, especially independence. We sought to develop a novel clinically relevant easy-to-use score predict patients’ loss independence after gastrointestinal surgery. Methods This retrospective cohort study included age ≥ 65 years enrolled in the American College Surgeons National Surgical Quality Improvement Program database Geriatric Pilot Project who underwent pancreatic, colorectal, or hepatic surgery (January 1, 2014- December 31, 2018). Primary outcome was – discharge facility other than home decline functional status. Patients from 2014 2017 comprised training data set. A logistic regression (LR) model generated using variables p < 0.2 univariable analysis. The six factors most predictive composed short LR scoring system. system validated 2018. Results Of 6,510 operations, 841 (13%) lost Training validation datasets had 5,232 (80%) 1,278 (20%) patients, respectively. impactful predicting were age, preoperative mobility aid use, Society Anesthesiologists classification, albumin, non-elective surgery, race (all OR > 1.83; 0.001). odds ratio each these used create sixteen-point demonstrated satisfactory discrimination calibration across datasets, Receiver Operating Characteristic Area Under Curve 0.78 both Hosmer-Lemeshow statistic 0.16 0.34, Conclusions predicts for geriatric operations. Using readily available variables, this tool can be applied urgent setting contribute family discussions related prior high-risk applicability additional surgical sub-specialties external should explored future studies.

Язык: Английский

---

Integrating general practitioners’ and patients’ perspectives in the development of a digital tool supporting primary care for older patients with multimorbidity: a focus group study

Frontiers in Digital Health, Год журнала: 2025, Номер 7

Опубликована: Янв. 21, 2025

Introduction The web application gp-multitool.de is based on the German clinical practice guideline “multimorbidity” and supports mutual prioritisation of treatments by GPs (general practitioners) patients. facilitates sending hyperlinks to standardized assessments email, which can be completed patients any suitable digital device. document decisions. tool also a structured medication review. Aims this study were consider needs wants target groups in implementing tool, examine themes discussions order identify aspects considered most important for customising tool. Materials methods We conducted six focus with 32 33 Eight alongside programming four after finishing prototype. recruited mail asked invite up eligible from their participate. Focus semi-structured interview guides discussed assessments, functionalities, usability reliability gp-multitool.de. Discussions transcribed verbatim analysed using content analysis. Results wanted avoid unnecessary time-consuming functions did not want explore problems that they could provide solutions for. For some suggested simplifying scales or including residual categories. addressed possible misunderstandings due wording if items might too intimate overtax intellectually. In cases, participants confirmed usability, but changes default settings pointed out few minor bugs needed fixed. While data security an topic, unconcerned issue open share data. Conclusion Our indicates used customize according thus, improve content, functionality, tools. However, tools still need piloted evaluated everyday care. our groups, relevant approach overcoming deficits information

Язык: Английский

---

Real-Time Driver Drowsiness Detection Using Facial Analysis and Machine Learning Techniques

Sensors, Год журнала: 2025, Номер 25(3), С. 812 - 812

Опубликована: Янв. 29, 2025

Drowsy driving poses a significant challenge to road safety worldwide, contributing thousands of accidents and fatalities annually. Despite advancements in driver drowsiness detection (DDD) systems, many existing methods face limitations such as intrusiveness delayed reaction times. This research addresses these gaps by leveraging facial analysis state-of-the-art machine learning techniques develop real-time, non-intrusive DDD system. A distinctive aspect this is its systematic assessment various deep algorithms across three pivotal public datasets, the NTHUDDD, YawDD, UTA-RLDD, known for their widespread use studies. Our evaluation covered including K-Nearest Neighbors (KNNs), support vector machines (SVMs), convolutional neural networks (CNNs), advanced computer vision (CV) models YOLOv5, YOLOv8, Faster R-CNN. Notably, KNNs classifier reported highest accuracy 98.89%, precision 99.27%, an F1 score 98.86% on UTA-RLDD. Among CV methods, YOLOv5 YOLOv8 demonstrated exceptional performance, achieving 100% recall with [email protected] values 99.5% In contrast, R-CNN showed 81.0% 63.4% same dataset. These results demonstrate potential our system significantly enhance providing proactive alerts real time.

Язык: Английский

---

Phenotyping to predict 12-month health outcomes of older general medicine patients

Aging Clinical and Experimental Research, Год журнала: 2025, Номер 37(1)

Опубликована: Фев. 22, 2025

Abstract Background A variety of unsupervised learning algorithms have been used to phenotype older patients, enabling directed care and personalised treatment plans. However, the ability clusters accurately discriminate for risk may vary depending on methods employed. Aims To compare seven clustering in their develop patient phenotypes that predict health outcomes. Methods Data was collected N = 737 medical inpatients during hospital stay five different types data (ICD-10 codes, ATC drug laboratory, clinic frailty data). We trialled (K-means, K-modes, hierarchical clustering, latent class analysis (LCA), DBSCAN) two graph-based approaches create separate each method datatype. These were as input a random forest classifier eleven outcomes: mortality at one, three, six 12 months, in-hospital falls delirium, length-of-stay, outpatient visits, readmissions three months. Results The overall median area-under-the-curve (AUC) across outcomes (from highest lowest) 0.758 (hierarchical), 0.739 (K-means), 0.722 (KG-Louvain), 0.704 (KNN-Louvain), 0.698 0.694 (DBSCAN) 0.656 (K-modes). Overall, most important type predicting mortality, ICD-10 disease codes readmissions, laboratory falls. Conclusions Clusters created using hierarchical, K-means Louvain community detection identified well-separated consistently associated with age-related adverse Frailty valuable

Язык: Английский

---