Hydration Monitoring Using Urinary Biomarkers: A Hybrid Classical–Quantum Predictive Modeling Framework
Saúl González-Bermejo, Tommaso Albrigi, Borja Vázquez-Morado, Urko Regueiro-Ramos, Daniel Casado-Faulí, Sergi Cònsul-Pacareu, Laia Alentorn, Jordi Ferré, Valentino Asole, and Parfait Atchadé-Adelomou(1) Lighthouse Disruptive Innovation Group S.L.(2) Jungle Venture S.L · 2026
Read Hydration Monitoring Using Urinary Biomarkers: A Hybrid Classical–Quantum Predictive Modeling Framework by Saúl González-Bermejo, Tommaso Albrigi, Borja Vázquez-Morado, Urko Regueiro-Ramos, Daniel Casado-Faulí, Sergi Cònsul-Pacareu, Laia Alentorn, Jordi Ferré, Valentino Asole, and Parfait Atchadé-Adelomou(1) Lighthouse Disruptive Innovation Group S.L.(2) Jungle Venture S.L (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-15381.txt.
Introduction
Gonzalez-Bermejo and nine colleagues test whether quantum machine learning can help track hydration from urine biomarkers collected by a smart toilet system. Framed as regression on specific gravity, conductivity, and volume, the study compares classical models, quantum regressors, and a new modular Quantum Sequential Model. Results sketch both the promise and the present limits of near-term quantum computing for health data.
Catalog record
| CREATOR | Saúl González-Bermejo, Tommaso Albrigi, Borja Vázquez-Morado, Urko Regueiro-Ramos, Daniel Casado-Faulí, Sergi Cònsul-Pacareu, Laia Alentorn, Jordi Ferré, Valentino Asole, and Parfait Atchadé-Adelomou(1) Lighthouse Disruptive Innovation Group S.L.(2) Jungle Venture S.L |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 36198 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.15381
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Hydration Monitoring Using Urinary Biomarkers: A Hybrid Classical–Quantum Predictive Modeling Framework Saúl González-Bermejo, Tommaso Albrigi, Borja Vázquez-Morado, Urko Regueiro-Ramos, Daniel Casado-Faulí, Sergi Cònsul-Pacareu, Laia Alentorn, Jordi Ferré, Valentino Asole, and Parfait Atchadé-Adelomou(1) Lighthouse Disruptive Innovation Group S.L.(2) Jungle Venture S.L Abstract Hydration status is a key physiological indicator associated with cellular homeostasis, renal function, and overall health. Recent advances in smart sensing environments enable passive monitoring of urinary biomarkers that can provide continuous insight into hydration dynamics. In this work, we investigate predictive modeling approaches for hydration monitoring using biomarker data collected through the Predict Health Toilet (PHT) system. The problem is formulated as a regression task using urinary indicators such as urine specific gravity, conductivity, and volume. We evaluate classical machine learning models and quantum machine learning architectures based on variational quantum circuits. In particular, we introduce a modular Quantum Sequential Model (QSM) designed to construct flexible hybrid quantum–classical predictive pipelines. Experimental results compare classical regression models, symmetry-constrained quantum regressors, and QSM architectures. The results provide insights into the potential role of quantum machine learning in digital health monitoring systems and highlight the opportunities and current limitations of near-term quantum computing for physiological data analysis. I Introduction Hydration is a fundamental physiological process regulating cellular homeostasis, thermoregulation, cardiovascular stability, and renal function. Even mild levels of dehydration can negatively affect cognitive performance, physical endurance, and metabolic balance. Chronic or recurrent dehydration has been associated with increased risks of hospitalization, kidney dysfunction, and cognitive decline, particularly among elderly populations [ 17 , 9 ] . Assessing hydration status in everyday environments remains a significant challenge. Traditional monitoring methods typically rely on episodic laboratory measurements, such as plasma osmolality tests or urine strip analysis [ 1 ] . While these approaches provide reliable physiological indicators, they are inherently reactive and require active clinical intervention. As a consequence, early dehydration signals may remain undetected between clinical assessments. Recent advances in embedded sensing technologies and smart environments are enabling new paradigms for passive and continuous health monitoring [ 15 ] . Within this context, smart sanitation systems have emerged as a promising platform for non-invasive physiological monitoring. Smart toilets, in particular, offer the unique possibility of integrating biomarker sensing into routine daily activities without requiring active user participation [ 15 ] . These systems can capture urinary biomarkers such as urine specific gravity, conductivity, and volume, which provide valuable indicators of hydration status and renal concentration processes. The Predict Health Toilet (PHT) system follows this paradigm by transforming conventional sanitation infrastructure into a passive health monitoring platform capable of collecting physiological measurements during normal daily use. In this work, the system leverages urinary biomarker data obtained from the Smart Urinal platform, an IP-protected urinalysis solution that integrates spectrophotometry and electrochemical sensing to enable accurate, multi-parameter biochemical analysis directly at the point of use. This biosensing approach supports real-time, non-invasive, and non-supervised monitoring, allowing for continuous and scalable data collection in real-world settings. However, the effective interpretation of these biomarker signals requires predictive models capable of identifying relevant physiological patterns and translating them into clinically meaningful indicators. Machine learning techniques have been widely applied to biomedical prediction tasks and have demonstrated strong performance when analyzing structured physiological datasets [ 10 ] . These models are capable of capturing nonlinear relationships between biomarkers and physiological conditions, enabling early detection and risk prediction in digital health systems. More recently, quantum machine learning (QML) has emerged as a novel research direction exploring the potential advantages of quantum computational models for pattern recognition and regression tasks [ 19 , 5 ] . Variational quantum circuits and hybrid quantum–classical learning architectures provide flexible frameworks for modeling nonlinear relationships within high-dimensional feature spaces while remaining compatible with noisy intermediate-scale quantum (NISQ) hardware. Despite increasing interest in QML, its applicability to biomedical prediction problems remains largely unexplored. In particular, there is limited empirical evidence evaluating how quantum models compare with classical machine learning approaches when applied to physiological datasets derived from passive monitoring systems. In this work, we investigate predictive modeling approaches for hydration monitoring using urinary biomarker data collected through the Predict Health Toilet (PHT) system, a non-intrusive health monitoring platform designed to automatically capture physiological indicators from urine during routine bathroom use. By continuously collecting longitudinal biomarker measurements related to hydration and metabolic status, the system generates structured datasets suitable for predictive health analytics. Using these data, we evaluate both classical machine learning models and quantum machine learning architectures, including symmetry-constrained quantum regressors and a modular Quantum Sequential Model (QSM) designed to enable the flexible construction of hybrid quantum–classical predictive pipelines. The objective of this study is to analyze the potential role of quantum machine learning in digital health monitoring systems and to explore how emerging quantum architectures may complement classical predictive models in the analysis of physiological biomarker data. II State of the Art Hydration assessment traditionally relies on physiological biomarkers such as plasma osmolality, urine osmolality, and urine specific gravity (USG) [ 9 , 1 ] . Urine specific gravity is widely used because it provides a practical proxy for urine concentration and hydration status [ 1 ] . Although USG correlates with urine osmolality, the two measures represent different physical quantities. Recent work has explored the integration of biosensing technologies into smart sanitation infrastructure for health monitoring [ 15 ] . These systems aim to transform everyday environments into platforms capable of collecting physiological data passively. At the same time, machine learning methods have been widely applied in healthcare prediction and clinical decision support [ 10 ] . Emerging research has also explored the use of quantum machine learning models for biomedical prediction tasks [ 5 ] . III Methodology This section introduces the physiological modeling framework used for hydration monitoring as well as the predictive learning formulation used to estimate hydration-related variables from urinary biomarkers. III-A Biomarker Representation Hydration monitoring relies on physiological biomarkers that reflect the body’s fluid balance and renal concentration processes. Urinary biomarkers have been widely used in both clinical medicine and sports physiology as practical indicators of hydration status [ 9 , 1 ] . Let the urinary biomarker feature vector observed at time t be defined as X_{t}=\begin{bmatrix}USG_{t}\\ C_{t}\\ V_{t}\end{bmatrix} (1) where • USG_{t} denotes urine specific gravity • C_{t} denotes urine conductivity • V_{t} denotes urine volume The vector defined in ( 1 ) captures complementary biochemical properties related to hydration status. Urine specific gravity (USG) provides a proxy measurement of urine concentration, as it provides a rapid estimate of urine, while conductivity reflects ionic content within the urine [ 1 ] . Urine volume provide contextual physiological signals that contribute additional information about renal function and fluid balance. These biomarkers have been widely used in hydration assessment studies due to their ability to reflect short-term changes in body fluid regulation [ 9 , 17 ] . For example, clinical literature typically considers values between 1.000 and 1.040 as the physiological operating range for USG-based hydration assessment. III-B Predictive Learning Problem The objective of the predictive modeling framework is to estimate hydration-related physiological variables from urinary biomarker measurements. Let the dataset be defined as D=\{(X_{i},y_{i})\}_{i=1}^{N} (2) where X_{i} denotes the biomarker vector defined in ( 1 ) and y_{i}