Alkmini Nega

Dr. Alkmini Nega


Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 June 2023 in Science Advances
Verified
Journal Article
Open access
PJPolina Jacoutot
Polina Jacoutot
ASAlberto D. Scaccabarozzi
Alberto D. Scaccabarozzi
DNDavide Nodari
Davide Nodari
JPJulianna Panidi
Julianna Panidi
ZQZhuoran Qiao
Zhuoran Qiao
NGNicola Gasparini
Nicola Gasparini
One of the key challenges facing organic photodiodes (OPDs) is increasing the detection into the infrared region. Organic semiconductor polymers provide a platform for tuning the bandgap and optoelectronic response to go beyond the traditional 1000-nanometer benchmark. In this work, we present a near-infrared (NIR) polymer with absorption up to 1500 nanometers. The polymer-based OPD delivers a high specific detectivity D * of 1.03 × 10 10 Jones (−2 volts) at 1200 nanometers and a dark current J d of just 2.3 × 10 −6 ampere per square centimeter at −2 volts. We demonstrate a strong improvement of all OPD metrics in the NIR region compared to previously reported NIR OPD due to the enhanced crystallinity and optimized energy alignment, which leads to reduced charge recombination. The high D * value in the 1100-to-1300-nanometer region is particularly promising for biosensing applications. We demonstrate the OPD as a pulse oximeter under NIR illumination, delivering heart rate and blood oxygen saturation readings in real time without signal amplification.

28 January 2024 in The Journal of Chemical Physics
Verified
Journal Article
Open access
PDF file
PKPanagiotis Kardasis
Panagiotis Kardasis
ITIoannis Tzourtzouklis
Ioannis Tzourtzouklis
ANAlkmini Nega
Alkmini Nega
GSGeorgios Sakellariou
Georgios Sakellariou
MSMartin Steinhart
Martin Steinhart
GFGeorge Floudas
George Floudas
We report the imbibition and adsorption kinetics of a series of symmetric linear/star cis-1,4-polyisoprene blends within the long channels of self-ordered nanoporous anodic aluminum oxide (abbreviated: AAO). Using in situ nanodielectric spectroscopy, we followed the evolution of the longest chain modes in the blends with a judicious selection of molar masses for the constituent components. We demonstrated differences in the imbibition kinetics of linear and star components based on the relative viscosities (e.g., polymers with lower zero-shear viscosity penetrated first the nanopores). Following the complete imbibition of the pores, the adsorption time, τads, of each component was evaluated from the reduction in the dielectric strength of the respective chain modes. In the majority of blends, both components exhibited slower adsorption kinetics with respect to the homopolymers. The only exception was the case of entangled stars mixed with shorter linear chains, the latter acting as a diluent for the star component. This gives rise to what is known as topology sorting, e.g., the separation of linear/star blend components in the absence of solvent. Moreover, a simple relation (τads ∼ 10 × tpeak; tpeak is the time needed for the complete filling of pores) was found for linear polymers and stars. This suggested that the characteristic timescale of imbibition (tpeak) governs the adsorption process of polymers. It further implied the possibility of predicting the adsorption times of high molar mass polymers of various architectures by the shorter imbibition times.

1 July 2025 in Eng
Verified
Journal Article
Open access
PDF file
DKDimitris Kalatzis
Dimitris Kalatzis
ANAlkmini Nega
Alkmini Nega
YKYiannis Kiouvrekis
Yiannis Kiouvrekis
Raman spectroscopy has become an indispensable analytical technique in pharmaceutical research, offering non-invasive, rapid, and chemically specific insights into pharmaceutical compounds. In this study, we present a comprehensive benchmark of machine learning models for classifying 32 pharmaceutical compounds based on their Raman spectral signatures. A diverse array of algorithms—including Support Vector Machines (SVMs), Random Forests, k-Nearest Neighbors (k-NN), Gradient Boosting (XGBoost, LightGBM), and 1D Convolutional Neural Networks (CNNs)—were evaluated on a publicly available dataset. The results demonstrate outstanding classification performance across models, with linear SVM achieving the highest accuracy of 99.88%, followed closely by CNN (99.26%). Ensemble methods such as Random Forest and XGBoost also yielded high accuracies above 98.3%. In addition to strong predictive performance, SHAP (SHapley Additive exPlanations) analysis was employed to interpret model decisions. CNN models, in particular, revealed well-localized and chemically meaningful spectral regions critical to classification. This combination of high accuracy and interpretability highlights the promise of explainable AI in pharmaceutical analysis and quality control, offering robust, transparent, and scalable solutions for real-world applications.

18 January 2021 in Electronic Materials
Verified
Journal Article
Open access
PDF file
BSBenedetta Maria Squeo
Benedetta Maria Squeo
Aggelos AvramopoulosAggelos Avramopoulos
Aggelos Avramopoulos
ANAlkmini Nega
Alkmini Nega
APAristea Pavlou
Aristea Pavlou
MSMichael G. Siskos
Michael G. Siskos
Christos ChochosChristos Chochos
Christos Chochos
A new organic material with three 4,4-difluoro-4-borata-3a-azonia-4a-aza-s-indacene dyes (BODIPYs) at the periphery of the central core is successfully synthesized (3BDP3T) and its corresponding aqueous nanoparticles are prepared via the encapsulation approach and characterized in detail both experimentally and theoretically with the aid of the Density Functional Theory (DFT). The linear and non-linear optical properties of the synthesized material are also studied. Until now, the development of organic materials with three BODIPYs as substituents is limited and their properties are not fully resolved. The obtained 3BDP3T-based nanoparticles exhibit far-red and near infrared (NIR) emission with photoluminescence quantum yields of 0.021, which is promising as a new fluorescent contrast agent in the far-red and NIR spectral regions.

18 September 2026 in AI in Medicine
Verified
Journal Article
Open access
PDF file
DKDimitris Kalatzis
Dimitris Kalatzis
ANAlkmini Nega
Alkmini Nega
Biomedical spectroscopy, including Raman, surface-enhanced Raman spectroscopy (SERS), infrared spectroscopy, and hyperspectral imaging, is increasingly combined with machine learning for disease classification, sample characterization, and biomarker-oriented analysis. However, high predictive performance does not establish whether model-relevant spectral features are biologically meaningful or whether highlighted regions can support reliable biochemical interpretation. Explainable artificial intelligence (XAI) methods, particularly SHAP and LIME, are increasingly used to identify influential wavenumbers, spectral bands, and wavelength intervals; yet feature importance is often interpreted too directly as biochemical or clinical evidence. This focused narrative review synthesizes SHAP, LIME, and related XAI methods across biomedical spectroscopy applications in cancer diagnostics, microbial identification, pharmaceutical analysis, and tissue or biofluid characterization. We examine key challenges, including correlated variables, peak overlap, preprocessing dependence, background choice, model dependence, and explanation instability. Beyond spectral attribution, we propose a five-step validation framework linking valid model development, explanation stability, region-level interpretation, biochemical plausibility, and independent analytical, biological, or clinical validation. The framework is intended to distinguish candidate spectral evidence from unstable or technically confounded explanations and to support reproducible, transparent, and clinically meaningful use of XAI in biomedical spectroscopy.