BIOSTEC_DC 2025 Abstracts

Full Papers
Full Papers
Paper Nr: 5
Title:

Development of Metal Based Piezoelectric Micropumps for Targeted Drug Dosing Applications Incorporating Integrated Failure Monitoring Techniques
Authors:

Nivedha Surendran
Abstract: The demand for innovative drug dosing devices has surged in recent years, driven by the rapid development of novel biological drugs that require customized and precise administration methods. With the introduction of complex biologics, such as monoclonal antibodies and gene therapies, there is a growing need for dosing systems that can efficiently handle high-viscosity drugs and deliver accurate dosages. This has raised significant challenges for conventional drug delivery mechanisms, particularly in the context of two widely used methods: short-term subcutaneous drug infusion and long-term implantable drug infusion systems. However, these drugs, often embedded in viscous matrices up to 25 times the viscosity of water. Conventional dosing systems quickly reach their limits, as these viscous liquids cause an extremely large pressure drop on the thin injection needles, leading to inconsistent dosing, device wear, and even system failure. This research focuses on the development of miniaturized metal-based piezoelectric micropumps capable of overcoming the challenges associated with high-viscosity drug administration along with high force generating actuators to generate the precise, controlled movement required to pump high-viscosity fluids. In addition to the inherent challenges of dosing high-viscosity drugs, external disturbances such as presence of air bubbles, catheter blockages, kinking of tubes, and delamination of piezoceramic actuators can lead to dosing errors. Therefore, this research also focuses on developing techniques to monitor the potential failure states by incorporating sensing capabilities into the dosing device to identify deviations from normal function, such as changes in pump chamber pressure or net flow rate and activating corrective measures in real-time. This approach facilitates the creation of a closed-loop control system for the microdosing device, enhancing both its reliability and safety.
Paper Nr: 6
Title:

Low-Rank Tensor Decomposition for Cross-Bispectral Analysis of Neural Data
Authors:

Dionysia Kaziki
Abstract: Understanding the complex, multi-dimensional nature of neural dynamics requires advanced computational methods capable of analyzing high-dimensional data with non-linear interactions. Traditional approaches, such as those applied to EEG and MEG datasets, often fail to capture critical cross-frequency coupling and spatial dependencies, while also struggling with the curse of dimensionality. This work introduces a novel low-rank tensor decomposition framework for cross-bispectral analysis, enabling the extraction of meaningful patterns from high-dimensional data. By modeling the bispectrum as a function of latent sources, the proposed method significantly reduces computational complexity while preserving critical structural information. Validation on EEG datasets demonstrates the method’s ability to retain interpretability and enhance insights into neural activity.
Paper Nr: 7
Title:

Strategies for Monitoring non-Invasive of Patient Status During Cardiopulmonary Resuscitation and Its Uses as Indicator of Cardiopulmonary Resuscitation Prognosis
Authors:

Camilo Sandoval-Rodriguez
Abstract: Effective treatment of cardiorespiratory arrest includes chest compressions and ventilations during cardiopulmonary resuscitation (CPR) to maintain minimal blood flow to increase the likelihood of survival . Resuscitation guidelines recommend monitoring exhaled air carbon dioxide concentrations using capnography signals as a noninvasive indicator of blood flow during CPR . Interference artifacts have been reported during chest compressions, which complicate interpretation of the capnography signal. However, the presence of these artifacts could help to better understand the role of capnography in interpreting the patient's hemodynamic status. Additionally, the behavior of the thorax at the time of the maneuver may differ according to the characteristics of each individual. After consulting the background, some guidelines are presented regarding the relationship that the capnography signal may have as an indicator of the early detection of the return of spontaneous circulation (ROSC) and the prognosis of the patient during CPR. The work has made significant progress, the origins of the artifact in capnography were studied, different models are presented with respect to the dynamics of the chest in order to establish a dynamic response of the patient during CPR. The standardization of end tidal CO2 (ETCO2) levels has also been assessed as a tool for predicting the return to spontaneous circulation ROSC. Different machine learning techniques are proposed.
Paper Nr: 8
Title:

Polarization-Resolved Third Harmonic Generation Imaging in Starch
Authors:

Maria Kefalogianni
Abstract: In this study, we present a deeper insight into the polarization dependent THG signals produced by a hydrated corn starch granule. By varying the direction of the excitation linear polarization, we observe two different THG signal modulations, one-peaked and double-peaked which correspond to different regions of the granule, i.e. outer shell, and bulk inside. We employ a theoretical model based on nonlinear optics to describe this response and extract susceptibility ratios and molecular angle orientation. We define the contribution of x orientation compared to y in susceptibility for an orthorhombic starch structure symmetry on the derived THG signal as anisotropy ratio .The outer surface of the granule reveals anisotropy ratio approximately 1,2 which means is bigger than , while in bulk of starch is almost equal to 1. This technique provides a non-invasive way to study and discriminate outer shell from inner regions of granule, with potential applications in agricultural science, food technology, and biopolymer research.