A5064146030- Advanced Chemical Sensor Technologies
- Gold and Silver Nanoparticles Synthesis and Applications
- Spectroscopy and Chemometric Analyses
- Spectroscopy Techniques in Biomedical and Chemical Research
- Speech and Audio Processing
- Speech Recognition and Synthesis
- Advanced biosensing and bioanalysis techniques
- Mechanical and Optical Resonators
- Gas Sensing Nanomaterials and Sensors
- Biosensors and Analytical Detection
- Infrastructure Maintenance and Monitoring
- Time Series Analysis and Forecasting
- Microfluidic and Capillary Electrophoresis Applications
- 3D Printing in Biomedical Research
- Stochastic Gradient Optimization Techniques
- Privacy-Preserving Technologies in Data
- Analytical Chemistry and Sensors
- Blind Source Separation Techniques
- Anomaly Detection Techniques and Applications
- Traffic Prediction and Management Techniques
- Mass Spectrometry Techniques and Applications
- Identification and Quantification in Food
- Indoor and Outdoor Localization Technologies
- Analytical Chemistry and Chromatography
- Sensor Technology and Measurement Systems
Technical University of Denmark
2016-2025
Danish National Research Foundation
2020
The Velux Foundations
2020
Surface-enhanced Raman spectroscopy (SERS) has been used in a variety of biological applications due to its high sensitivity and specificity. Here, we report SERS-based biosensing approach for quantitative detection biomolecules. A SERS substrate bearing gold-decorated silicon nanopillars is functionalized with aptamers sensitive specific target molecules. In this study, TAMRA-labeled vasopressin molecules the picomolar regime (1 pM 1 nM) are specifically captured by on nanostructured...
In this work we have performed a detailed study of the influence various parameters on spray coating polymer films. Our aim is to produce films uniform thickness (500 nm 1 μm) and low roughness compared film thickness. The coatings are characterized with respect thickness, (profilometer), morphology (optical microscopy). Polyvinylpyrrolidone (PVP) used do full factorial design experiments selected process such as temperature, distance between nozzle substrate, speed nozzle. A mathematical...
Reusability of sensors is relevant when aiming to decrease variation between measurements, as well cost and time analysis. We present an electrochemically assisted surface-enhanced Raman spectroscopy (SERS) platform with the capability reverse analyte-surface interaction, without damaging SERS substrate, allowing for efficient sensor reuse. The was used in combination a sample pretreatment step, detecting melamine from milk. found that enhanced interaction results significant improvement...
Lung infections with Pseudomonas aeruginosa (PA) is the most common cause of morbidity and mortality in cystic fibrosis (CF) patients. Due to its ready adaptation dehydrated mucosa CF airways, PA tend become chronic, eventually killing patient. Hydrogen cyanide (HCN) at ppb level has been reported be a biomarker. For early detection children not yet chronically lung infected non-invasive Surface-Enhanced Raman Spectroscopy (SERS)-based breath nanosensor being developed. The triple bond...
Data heterogeneity across clients is a key challenge in federated learning. Prior works address this by either aligning client and server models or using control variates to correct model drift. Although these methods achieve fast convergence convex simple non-convex problems, the performance over-parameterized such as deep neural networks lacking. In paper, we first revisit widely used FedAvg algorithm network understand how data influences gradient updates layers. We observe that while...
Raman spectroscopy is an important, low-cost, non-intrusive technique often used for chemical identification. Typical approaches identify a spectrum by comparing it with reference database using supervised machine learning, which usually requires careful preprocessing and multiple spectra available per analyte. We propose new learning identification contrastive representation learning. Our approach no works as little single have significantly improved or are on par the existing...
The acoustic variability of noisy and reverberant speech mixtures is influenced by multiple factors, such as the spectro-temporal characteristics target speaker interfering noise, signal-to-noise ratio (SNR) room characteristics. This large poses a major challenge for learning-based enhancement systems, since mismatch between training testing conditions can substantially reduce performance system. Generalization to unseen typically assessed system with new speech, noise or binaural impulse...
Colorimetric sensing technology for the detection of explosives, drugs, and their precursor chemicals is an important effective approach. In this work, we use various machine learning models to detect these substances from colorimetric experiments conducted in controlled environments. The based on response a chip containing 26 chemo-responsive dyes indicate that homemade explosives (HMEs) such as hexamethylene triperoxide diamine (HMTD), triacetone (TATP), methyl ethyl ketone peroxide (MEKP)...
Rapidly and accurately detecting quantifying the concentrations of nitroaromatic explosives is critical for public health security.
Modern cars are equipped with many sensors that measure information about the vehicle and its surroundings. These measurements therefore related to ride-surface conditions over which is passing. The paper commences by outlining a four-component vision for performing road condition evaluation based on in-vehicle sensor readings subsequent feeding of pavement management systems (PMSs) live information. This expected enrich functionalities PMSs, ultimately lead improved maintenance repair...
The performance of neural network-based speech enhancement systems is primarily influenced by the model architecture, whereas training times and computational resource utilization are affected parameters such as batch size. Since noisy reverberant mixtures can have different duration, a batching strategy required to handle variable size inputs during training, in particular for state-of-the-art end-to-end systems. Such strategies usually strive compromise between zero-padding data...
Peak detection and localization in a noisy signal with an unknown baseline is fundamental task processing applications such as spectroscopy. A current trend to reformulate traditional pipelines (deep) neural networks that can be trained end-to-end. trainable algorithm for removal peak serve important module pipeline. In practical applications, one of the most successful approaches joint suppression based on continuous wavelet transform: We re-formulate this convolutional network (CNN)...
During the past decades, microdevices have been evaluated as a means to overcome challenges within oral drug delivery, thus improving bioavailability. Fabrication of is often limited planar or simple 3D designs. Therefore, this work explores how microscale stereolithography printing can be used fabricate radiopaque microcontainers with enhanced mucoadhesive geometries, which enhance bioavailability by increasing gastrointestinal retention. Ex vivo force measurements suggest increased...
Road roughness is a very important road condition for the infrastructure, as affects both safety and ride comfort of passengers. The roads deteriorate over time which means must be continuously monitored in order to have an accurate understand infrastructure. In this paper, we propose machine learning pipeline prediction using vertical acceleration car speed. We compared well-known supervised models such linear regression, naive Bayes, k-nearest neighbor, random forest, support vector...
In the framework of research project 'Xsense' at Technical University Denmark (DTU) we are developing a simple colorimetric sensor array which can be useful in detection explosives like DNT and TNT, identification volatile organic compounds presence water vapor air. The technology is based on an chemo-responsive dyes immobilized solid support. Upon exposure to analyte suspicion dye changes color. Each chosen reacts chemo selectively with analytes interest. A change color signature indicates...
In this work the use of a high-throughput nanomechanical detection system based on DVD-ROM optical drive and cantilever sensors is presented for urokinase plasminogen activator receptor inflammatory biomarker (uPAR). Several large scale studies have linked elevated levels soluble uPAR (suPAR) to infectious diseases, such as HIV, certain types cancer. Using hundreds cantilevers DVD-based platform, deflection response from antibody-antigen recognition investigated function suPAR concentration....
Based on single molecule surface-enhanced Raman spectroscopy (SERS) intensity distribution theory, a mathematical model is developed for highly sensitive biomolecular quantification using mapping SERS substrates with planar geometries.