The future of intelligent biosystems is bright, with major strides projected as a result of new discoveries and innovations. A number of leading multinational companies have integrated intelligent biosystems to uplift food quality. For example, Nestlé uses biosensing in production to track microbial contamination. IBM is adopting blockchain to ensure transparent food traceability, while BluWrap monitors the oxygen and temperature of fresh fish pallets with smart packaging to maximize shelf life and reduce carbon footprint. At the same time, biosensors supported by the Internet of Things (IoT) can help farmers, stakeholders, and the agri-food industry through rapid testing and predictive analytics based on sensor-generated computing using Artificial Intelligence (AI). The integration of AI methodologies, including cluster analysis and classification algorithms, with biosensors can bridge the gap between data collection and analysis and advance the accuracy of data handling throughout the food supply chain. The methodology of this work follows a systematic review of the literature on intelligent biosystems and AI tools in food safety, evaluating biosensing techniques, challenges, and scalability, and exploring future directions. The potential of these tools is conspicuous, although their application in real-world scenarios is still limited due to lack of focus, implementation costs, scalability, and well-adapted and regulatory framework research. In this review, we search for results that examine the state of the art of AI for food quality control, highlighting the impact of smart biosensors that offer advanced real-time monitoring, predictive analytics, optimization, enhanced traceability, and consumer empowerment to improve risk management and ensure high standards of food processing and safety, as well as public health and economic integrity.

Wearable technologies are rapidly expanding, creating a demand for the real-time monitoring of molecular biomarkers. Non-invasive samples, such as sweat and saliva, are particularly promising for this purpose. However, achieving selectivity and specificity in sensor measurements remains a challenge due to the complexity of biomarkers and the stability of captured molecules. Laser-Scribed (LS) electrodes, fabricated using a CO₂ laser cutter on polyimide substrates, offer a cost-effective and promising alternative for wearable electrochemical sensors and biosensors. This study investigates the optimization of LS electrode manufacturing parameters using a 60 W CO₂ laser cutter and explores their application for the label-free detection and classification of biomarkers in sweat. Cyclic voltammetry (CV) and Electrochemical Impedance Spectroscopy (EIS) experiments were conducted to characterize the electrochemical performance of LS electrodes, with a focus on detecting L-histidine in artificial sweat. The optimized LS electrodes exhibited high sensitivity, demonstrating a linear relationship (R² = 0.987) between the current peak and L-histidine concentration in the range from 8.3 mM to 50 mM. Additionally, an MLPNN (Multilayer Perceptron Neural Network) machine learning algorithm was trained using CV data to classify L-histidine in artificial sweat for lower, physiologically relevant concentrations (e.g., 0.12 – 3.3 mM) where linearity is lost. The results achieved 90% accuracy, highlighting the potential of LS electrodes for real-time, label-free biomarker monitoring in wearable health devices. In conclusion, this study demonstrates the effectiveness of LS electrodes and data-driven classification techniques for sweat component monitoring. Future research will focus on improving the detection capabilities of LS electrodes and expanding their application to classify other sweat biomarkers, such as NaH₂PO₄, NaCl, and Na₂HPO₄. This work advances the development of high-performance and disposable wearable biosensors for non-invasive health monitoring.

Ensuring water quality is essential to safeguarding public health, as contaminated water is a leading cause of diseases such as cholera, dysentery, hepatitis, and typhoid fever. Conventional methods for monitoring bacterial contamination include membrane filtration, multiple tube fermentation, and enzyme-based assays. These methods are highly reliable but are constrained by lengthy processing times and require expensive equipment, consumables, and trained personnel. This work presents a portable UV-LED/RGB sensor system designed to address these limitations by using a multi-well self-loading microfluidic device for sample preparation-free analysis, a defined substrate assay specific to Enterococcus faecalis, a thermoelectric heater for assay incubation, UV-LEDs for sample excitation, RGB sensors for emission acquisition, a 3D-printed casing, and a microcontroller, achieving fast, low-cost, portable, and automated bacteria quantification. Wells in the microfluidic device are independent from each other and are designed to autonomously load with sample water when the device is submerged. The number of wells and volume per well are designed for bacterial quantification using Most Probable Number (MPN) analysis. Fluorogenic assay reagents are pre-loaded into each well of the microfluidic device and dissolve when the wells are loaded with sample water. Fluorescence signals captured by the RGB sensors are analyzed using machine learning (ML) algorithms including a Multilayer Perceptron Neural Network (MLPNN), which determines whether individual wells will be positive or negative by the end of a 24-hour period. The results show 100% accuracy in classifying bacterial presence within wells and a remarkably low detection time of under 30 minutes. The novel combination of ML and MPN analysis in an automated and cost-effective manner allows for near-real-time bacterial quantification and marks a significant advance in rapid bacteriological water quality analysis. The innovations presented offer a robust solution for on-site water quality monitoring, advancing public health and enabling faster responses to potential waterborne contamination.

The integration of biosensor technology with hybrid artificial intelligence (AI) algorithms has greatly advanced the field of biomedical diagnostics. Prostate cancer is one of the most common types of cancer in men, and prostate-specific antigen (PSA) is a key biomarker for early diagnosis. Using an unprecedented approach that combines the strengths of a range of machine learning and deep learning methods, this methodology shows a balance between accuracy, sensitivity, and reliability in early cancer detection. The biosensor, using an electrochemical platform, shows the dual functionalization of a specific aptamer and gold nanoparticles (AuNPs), enabling PSA detection with high specificity. The hybridization of the artificial intelligence methods used in this study includes convolutional neural networks (CNNs), support vector machines (SVMs), and gradient boosting machines (GBMs), and it allows for the processing, assessment, and classification of biosensor data. The initial step involves collecting the biosensor signal of biomolecular interactions, which are transduced and transcribed into electrical signals. To prepare the raw data set for processing, advanced denoising and normalizing techniques are applied. Then, the CNN is run on these data to encode their features and identify complex patterns. Then, SVMs classify the PSA levels into three groups, namely normal, elevated, and at risk. In contrast, GBMs use CNN and SVM output predictions as inputs for the decision process.

In this manner, this hybrid algorithm methodology balances the interpretability of SVMs, the deep feature learning capacity of CNNs, and the strong prediction power of GBMs. Its second contribution first aims at increasing the amount of data by developing an ERL (ensemble reinforcement learning) framework that dynamically modulates the parameters based on instantaneous rewards. This makes the ensemble reinforcement learning framework unfixed for different biomedical applications but also for different datasets and patient profiles. Herein, we detail the design, development, and evaluation of this transformative system with the potential to fundamentally change cancer diagnostics by enabling scalable, low-cost, and accurate cancer screening. The hybrid algorithm strategy is a paradigm shift in biosensor technology, allowing us to address the limitations of traditional approaches while overcoming issues with achievable data standardization and model optimization.

Following the summary of the acquired optically significant metrics, the output efficiency is estimated to thoroughly review the proposed AI-adapted biosensor framework. Using this approach, the system achieved high accuracy in detecting prostate-specific antigen (PSA) concentrations and classifying the risk for cancer. This can be particularly useful for early-stage detection while reducing the risk of false-negative diagnoses. The specificity will reduce the false-positive results—meaning those whose condition is misidentified—and spare some unnecessary medical interventions. The biosensor also exhibits an ultra-low limit of detection (LOD), which allows for the detection of PSA within the subclinical range. This innovative system takes advantage of machine learning and AI techniques for biosensor technology applications, thus enabling new biomedical diagnostics with high sensitivity and specificity and low cost.

Introduction: It is unanimously known that cyanide (CN⁻) is one of the most toxic ions because it can interfere with the body’s physiological phenomena, causing endocrine disorders, respiratory failure, hypoxia, vascular necrosis, and even death. Therefore, it is essential to develop cost-effective, sensitive, rapid, and efficient methods for sensing CN⁻ ions.

Methods: The chemosensor was synthesised via a multi-step reaction, and its photophysical properties were analysed using UV–visible and fluorescence analyses. Binding was confirmed through Job’s plot, ¹H-NMR, HR-MS, and FT-IR, with TD-DFT validation using Gaussian 09W software.

Results: The chemosensor exhibited high selectivity and specificity for CN⁻, with a distinct naked-eye colour change from yellow to blue. Additionally, a “turn-off” fluorescence response was observed, attributed to the inhibition of the Intramolecular Charge Transfer (ICT) process. The fluorescence quenching efficiency for CN⁻ was 73.21%, with a Stern–Volmer quenching constant of 1.22 × 10⁵ M⁻¹. The detection limit was determined to be 5.47 µM. Furthermore, FT-IR spectroscopy, ¹H-NMR titration, and HR-MS analysis were conducted to study the plausible binding site. TD-DFT calculations and topological analysis were performed to investigate molecular orbitals and their localisation in the free probe and its CN⁻ complex.

Conclusion: The optical behaviour of the chemosensor towards CN⁻ was noted in this study. The “naked-eye” changes led to the development of test kit and smartphone analysis. Moreover, the fluorescent “turn-off” behaviour could be exploited for confocal fluorescence imaging of CN^- in living cells.

Introduction. Polymer ionic liquids (PILs) are a new class of ionic liquids that have great potential in various fields of application. PILs play a role in business development, prosperity, and high business activity, as well as in the development of business in general[1,2]. These are polyelectrolytes that contain a polymer base and, at the same time, must have fragments of ionic liquid (IL) in each of the links, which are repeated. These ionic polymers can be fully insulated for a long time.

PILs in medicine. Multidimensional sensor devices based on PILs are extremely promising for the development of high-performance screening in the field of environmentally friendly chemistry and biology, while providing easier identification of a variety of analyzed substances.

PILs as electronic skin. PILs can also be considered an ionic skin. These are self-healing materials that allow partial or even complete self-healing after damage and essentially mimic natural systems. Such sensors are beneficial to monitoring human body movement.

PIL actuators in dielectric elastomers. These substances and materials are necessary for soft robotics and medical purposes and are being actively improved at the moment. Development and progress in this area are directly related to decision making. The latest research is now focused on the synthesis and testing of more advanced actuators fabricated from PIL–dielectric elastomers.

Conclusions. The unique properties present great prospects for PIL research in these areas, where progress and breakthrough technologies can be expected in the coming years.
The research was carried out under Russian National Research Project No. AAAAA-A21–122040600057–3.
References

1. Lebedeva, O.; Kultin, D.; Kustov, L. Advanced Research and Prospects on Polymer Ionic Liquids: Trends, Potential and Application. Green Chem. 2023, 10.1039.D3GC02131A, doi:10.1039/D3GC02131A.
2. Lebedeva, O.; Kultin, D.; Kustov, L. Polymeric Ionic Liquids: Here, There and Everywhere. European Polymer Journal 2024, 203, 112657, doi:10.1016/j.eurpolymj.2023.112657.

Today's population faces a major challenge in the prevention and treatment of non-communicable diseases such as cardiovascular diseases, obesity, or type 2 diabetes, which are closely related to unhealthy lifestyle habits, including an inadequate diet and sedentary patterns. Identifying physiological and pathological parameters can help prevent and treat these diseases. In this respect, biosensors have been proposed as useful tools for analytical devices with biological sensing elements capable of detecting both physiological indicators and pathological markers that aid in the diagnosis, treatment, and long-term monitoring of these diseases. These tools also play a key role in the observation and assessment of physiological activities, with wearable biosensors being of particular interest due to their ability to provide continuous, real-time physiological information. Knowing the potential applications of these devices could aid in monitoring the real effect of these products’ consumption. Given the rising interest in functional foods to improve public health, wearable biosensors hold the potential to monitor the physiological effects of their consumption. This systematic review aims to analyze the current state of wearable biosensor development, highlighting their applications, benefits, and limitations. Specifically, it focuses onimplications for individuals using these devices and the challenges associated with data collection, interpretation, and extrapolation. By addressing these aspects, this review provides insights into the real-world applicability of wearable biosensors for disease prevention and personalized health management focused on nutrient intake.

Background: Hospitalized preterm infants often require months of vital signs monitoring in the neonatal intensive care unit. To date, wired sensors are essential for survival but are associated with numerous disadvantages, including sensor dislocations, skin trauma and hygiene risks. Non-contact vital sign monitoring would therefore represent a significant improvement in the care of hospitalized neonates.

Objective: This study aimed to develop a new microwave-based sensor for non-contact monitoring of vital signs.

Methods: We developed a radar-based vital parameter monitoring system for recording the respiratory rate of premature infants in a pediatric incubator. This novel sensor is a four-channel I/Q (In-Phase and Quadrature) radar system operating at 24 GHz with adapted antennas to cover the predefined area of interest on the body surface. As a proof of principle study, the system was tested in six anesthetized newborn piglets with a body weight between 1050 and 2710 g that were located in a newborn incubator.

Results: Using the radar-based system, thorax movements were detected and the respiratory rate was calculated. We observed high accordance between the signals of respiration detected by the novel microwave sensor and the signals of the cable-bound monitor at rest.

Conclusion: The novel microwave sensor is suited for measuring respiration in the piglet model. In the future, the sensor has to be optimized in order to improve its robustness against disturbances by heartbeats and body movements.

Significance: The study results have laid the foundation for non-contact monitoring of the respiratory rate, which could be used in neonatal intensive care units.

Generally, the breathing process involves inhalation and exhalation in which the movement of air that occurs in the lungs causes an acoustic effect called lung sounds. These lung sounds can be breath sounds, adventitious sounds, and vocal resonance, which can be diagnosed based on the anatomy of the location, where the sounds are detected during physical examination. Moreover, it is essential to analyse the health condition of the respiratory system and the airway process without causing any harm to the patients using non-invasive methods. Examination through a stethoscope is a commonly followed method; however, it is quite difficult to analyse lung sounds with lower acoustic levels using existing stethoscopes. To overcome this issue, the digital electronic stethoscope is utilised nowadays. In this work, an artificial intelligence-based wearable digital stethoscope is designed and developed for the analysis of cough sounds. Furthermore, the Arduino Nicla voice-based edge computing board is utilised to acquire and analyse the cough sounds of abnormal patients. The Arduino Nicla Voice board has an inbuilt microphone and Inertial Measurement Unit (IMU) which is used to record acoustic, acceleration, and magnetometer signals from normal individuals and abnormal patients. Also, machine learning algorithms such as Random Forest and Decision Trees are adopted and deployed in the edge computing board to classify normal and abnormal cough sounds. Performance analysis parameters such as accuracy, precision, recall, and F1_Score are derived for the adopted machine learning classifiers to evaluate the efficiency and efficacy of the system. This work appears to be of high social relevance since the proposed work will assist in the early prediction of COVID-19 and similar, other diseases using cough sounds.

This research focuses on the development of a capacitive pressure sensor integrated into a contact lens for continuous monitoring of intraocular pressure (IOP), a critical parameter for the early detection and management of glaucoma.

The sensor operates on capacitive sensing principles, where fluctuations in IOP alter the distance between sensor plates, resulting in measurable changes in capacitance. Pressure variations within the physiological range of 10–25 mmHg were analyzed to evaluate the sensor's performance and ensure accurate pressure readings. The sensor was also optimized to detect changes in the range of 2–5 mmHg. COMSOL Multiphysics simulations provided detailed insights into the sensor’s durability, sensitivity, and performance by analyzing capacitance variations and electric field intensity. Laboratory experiments simulated IOP changes under realistic conditions to assess the sensor's accuracy and reliability. Unlike traditional methods, often limited to clinical environments, this approach offers a continuous, real-time solution for monitoring IOP.

This system addresses current diagnostic limitations by enabling earlier and more precise interventions in glaucoma care. Integrating the sensor into a contact lens allows for practical and patient-focused solutions, helping reduce the risk of vision loss. This project also contributes to advancements in wearable biomedical technologies and paves the way for broader applications in continuous health monitoring.