Wee Jian ChinWei Yin LimSook Mei KhorNarayanan RamakrishnanPei Song, CHEECHEEPei Song0000-0002-8914-8524Choon-Hian Goh2026-10-152026-10-152026-0110.1016/j.cca.2025.120677https://dspace-cris.utar.edu.my/handle/123456789/11749Biosensors have emerged as transformative tools in modern diagnostics, enabling rapid, accurate, and sensitive detection of biological markers for disease diagnosis, real-time monitoring, and personalized healthcare. However, current biosensors still exhibit gaps in accuracy and sensitivity for various diagnostics applications, particularly in the detection of multiple analytes, non-invasive sensing, signal noise interference, and ultra-low concentration detection, which are crucial for diagnosing certain diseases. This review explores the recent advancements in Machine learning (ML) enhanced biosensors, highlighting their applications in health monitoring, disease diagnosis, and treatment evaluation. The novelty of this review lies in its comprehensive examination of how ML algorithms enhance biosensor technology by efficiently processing complex data and extracting actionable insights. Various types of biosensors, including electrochemical, optical, microfluidic, and wearable devices with the physiological signals they collect were introduced. The review empasize the role of supervised, unsupervised, and deep learning algorithms in improving biosensor performance, including tasks like classification, regression, clustering, and feature extraction. Despite significant progress, challenges remain in areas such as data privacy, ethical considerations, real-time data processing, computational requirement, and biosensor fabrication. Future prospects include the development of multifunctional biosensors, adaptive learning systems, and explainable artificial intelligence (AI) models to ensure the reliable, scalable, and ethical deployment in healthcare. Overall, this review underscores the potential of ML-reinforced biosensors to revolutionize diagnostics and personalized medicine while addressing key challenges that hinder their practical implementation. © 2025 Elsevier B.V.en-USBiosensorDisease diagnosisHealth monitoringMachine learningYTreatment evaluationAlgorithmsBiosensing TechniquesHumansMachine LearningC reactive proteinhigh density lipoproteinlow density lipoproteinuric acidaccuracyalgorithmArticleartificial neural networkbioelectrical impedance analysiscerebrospinal fluidclinical decision makingconvolutional neural networkdata qualitydecision treedeep learningelectroencephalographyfeature extractiongaitglucose blood levelgradient boosting algorithmhealthhealth careheart rate variabilityhierarchical clusteringhistogramk nearest neighborlong short term memory networkmachine learningmachine learning algorithmmicrofluidicsmonitoringMycobacterium tuberculosispersonalized medicinephysical activityprincipal component analysisrandom forestrecurrent neural networksimulationsupervised machine learningsupport vector machineunsupervised machine learninggenetic procedureshumanjournal-article