Advancing Healthcare Technology: From Edge-AI Diagnostics to Metagenomic Intelligence and Biomaterial Sensors

The intrinsic convergence of computational intelligence, material science, and molecular biology underpins modern healthcare engineering. At the nexus of this multidisciplinary interaction lies the urgent need for scalable, real-time diagnostic tools that can operating seamlessly from the planetary scale of metagenomic tracking to the micro-scale of wearable, edge-deployed physical sensors. Accelerating the translation of clinical workflows requires bridging the gap between algorithmic models and physical engineering realities. Addressing these complex healthcare demands necessitates robust software optimization frameworks, highly sensitive biomaterial substrates, assistive accessibility pipelines, and reference-free deep learning models that transcend classical alignment bottlenecks.

In response to these critical needs, we introduce the inaugural issue of the Journal of Biomedical Systems and Engineering - a peer-reviewed, open-access forum dedicated to advancing research at the intersection of medicine, hardware engineering, data science, and molecular diagnostics. Embracing the paradigm of translational technology, the journal serves as a multidisciplinary platform for high-impact original research, comprehensive reviews, and computational proof-of-concepts. Our inaugural issue features four contributions that collectively span the continuum from individual edge-hardware optimization and localized point-of-care biosensors to real-time assistive frameworks and deep-learning metagenomic architectures:

Odiba et al. provide a comprehensive review of the integration of Deep Learning (DL) with metagenomic sequencing for advanced microbial species differentiation.^[1] The authors evaluate four foundational neural architecture families - convolutional, recurrent, transformer-based genomic language, and graph neural networks - across clinical tasks including taxonomic classification, genome-resolved binning, and microbiome-phenotype association.^[2,3] The review contrasts these DL methodologies against conventional alignment-based limitations, specifically highlighting hurdles such as data leakage, domain shift across platforms, and database reference dependency. Ultimately, they outline the strategic necessity of reference-free, multi-modal foundation models to achieve biologically robust microbial identification. Basak and Sahu execute a comprehensive compression-aware ablation study aimed at continuous vital-sign anomaly detection on wearable and bedside Internet of Medical Things (IoMT) hardware.^[4] Utilizing a synthetic multi-variate vital-sign dataset mapped against National Early Warning Score 2 (NEWS2) thresholds, the authors systematically benchmark a baseline CNN-LSTM network against structured channel pruning, partial dynamic post-training 8-bit integer quantization, and knowledge distillation to a 1,662-parameter GRU student.^[5] While the baseline achieved an exceptional F1-score of 0.955, knowledge distillation yielded a massive 40.8-fold reduction in multiply-accumulate operations while preserving a high F1-score of 0.912. Crucially, the authors separate their software-sandbox metrics from a rigorous, self-contained implementation protocol designed to guide future physical edge-hardware deployment and clinical verification.

Shifting focus toward molecular-level real-time monitoring, Kumar et al. present the numerical design and optimization of a highly sensitive, multilayer Surface Plasmon Resonance (SPR) sensor for label-free urine glucose detection.^[6] The architecture integrates a calcium fluoride (CaF₂) prism coupled with a copper plasmonic layer, silicon dioxide (SiO₂), blue phosphorus (BlueP), and a 2D Franckeite mineral heterostructure to maximize the interaction of the evanescent field with the analyte.^[7,8] The optimized configuration demonstrates an extraordinary maximum angular sensitivity of 385.55°/RIU and a peak Figure of Merit (FoM) of 82.31/RIU. This computational proof-of-concept establishes a vital theoretical baseline for subsequent physical prototyping and clinical translation using real patient urine samples.

Moving from biochemical assays to human-computer interactions, Siddiqui et al. develop a real-time assistive pipeline combining Indian Sign Language (ISL) alphabet recognition with temporal stabilization.^[9] To accommodate users experiencing involuntary hand movements, such as Parkinson's disease tremors, the system tracks 21 three-dimensional skeletal joints via a webcam using MediaPipe Hands.^[10] These landmarks are mapped to a convolutional neural network (CNN) that achieves a macro F1-score of 99.27% across a 13,000-image dataset. Operating at an inference rate of ~9.23 frames per second, the prototype integrates a strict 20-frame consecutive prediction rule to smooth shaky-motion artifacts, delivering real-time text interpretation and augmented reality (AR) style feedback.

Collectively, these inaugural contributions demonstrate that safeguarding human health relies on a continuous synthesis of precise software optimization, sensitive material substrates, and scalable machine learning frameworks. They also delineate recurrent translational challenges - such as domain shift across clinical environments, laboratory-to-field hardware constraints, and the need for physical experimental validation - which we warmly encourage future contributors to address in upcoming issues.

The Editorial Office remains dedicated to a rigorous, transparent, and timely peer-review process adhering to the highest standards of scientific and ethical integrity. We enthusiastically invite biomedical engineers, computer scientists, material researchers, and clinical technologists to contribute their work and assist in establishing the Journal of Biomedical Systems and Engineering as a leading global venue for integrative health technology.

The author is a co-author of one article featured in this issue (Ref. 6); its peer review and editorial handling were carried out independently of the author.

The author declare that artificial intelligence (AI)-assisted tools were used only for language refinement, grammar improvement, and manuscript structuring purposes during the preparation of this work. All content and interpretations were reviewed and verified by the author.

1. [01] D. Odiba, F. C. Terna, O. G. Uyi, S. Anzaku, O. O. Orole, Integrating Metagenomics and Deep Learning for Accurate Microbial Species Differentiation, Journal of Biomedical Systems and Engineering, 2026, 1, 26803, doi: 10.64189/bse.26804.
2. [02] R. J. Wright, A. M. Comeau, M. G. I. Langille, From defaults to databases: Parameter and database choice dramatically impact the performance of metagenomic taxonomic classification tools, Microbial Genomics, 2023, 9, 000949, doi: 10.1099/mgen.0.000949.
3. [03] A. Escobar-Zepeda, E. E. Godoy-Lozano, L. Raggi, L. Segovia, E. Merino, R. M. Gutiérrez-Rios, K. Juarez, A. F. Licea-Navarro, L. Pardo-Lopez, A. Sanchez-Flores, Analysis of sequencing strategies and tools for taxonomic annotation: Defining standards for progressive metagenomics, Scientific Reports, 2018, 8, 12034, doi: 10.1038/s41598-018-30515-5.
4. [04] P. Basak, S. Sahu, Compression-aware vital-sign anomaly detection: A synthetic data ablation of pruning, quantization, and knowledge distillation, with a protocol for clinical and edge-hardware extension, Journal of Biomedical Systems and Engineering, 2026, 1, 26801, doi: 10.64189/bse.26801.
5. [05] T. Liang, J. Glossner, L. Wang, S. Shi, X. Zhang, Pruning and quantization for deep neural network acceleration: A survey, Neurocomputing, 2021, 461, 370–403, doi: 10.1016/j.neucom.2021.07.045.
6. [06] R. Kumar, L. Garia, B. Brahma, A. K. Bhoi, Enhanced surface plasmon resonance sensor employing blue phosphorus and franckeite for urine glucose detection, Journal of Biomedical Systems and Engineering, 2026, 1, 26802, doi: 10.64189/bse.26802.
7. [07] N. Mudgal, A. Saharia, A. Agarwal, J. Ali, P. Yupapin, G. Singh, Modeling of Highly Sensitive Surface Plasmon Resonance (SPR) Sensor for Urine Glucose Detection, Optical and Quantum Electronics, 2020, 52, 307, doi: 10.1007/s11082-020-02427-0.
8. [08] A. Shalabney, I. Abdulhalim, Electromagnetic Fields Distribution in Multilayer Thin Film Structures and the Origin of Sensitivity Enhancement in Surface Plasmon Resonance Sensors, Sensors and Actuators A: Physical, 2010, 159, 24–32, doi: 10.1016/j.sna.2010.02.005.
9. [09] F. Siddiqui, I. Ansari, N. Kazi, J. Khatri, Real-Time ISL to text interpretation and Parkinson’s tremor correction with augmented reality support, Journal of Biomedical Systems and Engineering, 2026, 1, 26803, doi: 10.64189/bse.26803.
10. [10] Z. Deng, Y. Leng, J. Hu, Z. Lin, X. Li, Q. Gao, SML: A Skeleton-based multi-feature learning method for sign language recognition, Knowledge-Based Systems, 2024, 301, 112288, doi: 10.1016/j.knosys.2024.112288.

Publisher Note: The views, statements, and data in all publications solely belong to the authors and contributors. GR Scholastic is not responsible for any injury resulting from the ideas, methods, or products mentioned. GR Scholastic remains neutral regarding jurisdictional claims in published maps and institutional affiliations.

This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits the non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as appropriate credit to the original author(s) and the source is given by providing a link to the Creative Commons License and changes need to be indicated if there are any. The images or other third-party material in this article are included in the article's Creative Commons License, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons License and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this License, visit: https://creativecommons.org/licenses/by-nc/4.0/

© The Author(s) 2026