MSc Thesis presentation of Glykeria Spyrou, Friday 13/11/2026, 2pm
On Friday 13/11//2026 at 2 pm, Glykeria Spyrou, a graduate student in the program “Data Science and Information Technologies, Biomedical Data Science – Bioinformatics” will present their MSc thesis titled:
Investigating biological mediators between circadian rhythm disruption and Mild Cognitive Impairment – A causal inference approach
Mild Cognitive Impairment (MCI) is the first clinically recognized stage of cognitive decline: deficits are measurable, yet everyday independence is preserved. Because prevention remains possible at this stage, the processes that lead to MCI are a natural target for study. Disturbed rest–activity rhythms have been associated with cognitive decline, and experimental work suggests that sleep and circadian regulation influence the production and clearance of proteins central to neurodegeneration, including tau. Recent studies have examined tau PET and white matter microstructure as statistical mediators of cross-sectional associations between rest–activity rhythms and cognitive scores. To our knowledge, none has asked whether cerebrospinal fluid (CSF) biomarkers mediate the association between rest–activity rhythms and a later clinical diagnosis of MCI. This thesis addresses that question using counterfactual mediation analysis.
We studied adults without dementia from the Aiginition Longitudinal Biomarker Investigation of Neurodegeneration (ALBION) cohort. We defined baseline (T0) using a six-day wrist actigraphy recording, from which we derived seven exposures describing the strength and fragmentation of the rest–activity rhythm and daytime and nighttime activity levels. Two proved central: cosinor amplitude, the size of the fitted 24-hour activity cycle, and the M10–RA composite, the mean of standardized daytime activity (M10, the most active ten hours) and standardized relative amplitude (RA, the day–night activity contrast). The candidate mediators were eight CSF measures: total tau (tTau), phosphorylated tau (pTau) and amyloid-β42 (Aβ42), their biomarker status (normal or abnormal), and the tTau/Aβ42 and pTau/Aβ42 ratios. The outcome was the consensus clinical diagnosis, normal cognition (NC) or MCI, at the first follow-up (T1) within two years of baseline (T0), made by neurologists after a full neurological and neuropsychological examination and without reference to actigraphy. The primary analysis used the 68 participants (36 NC, 32 MCI) who had CSF obtained within one year of T0. Two expanded datasets, each containing the previous one, served as sensitivity analyses of this choice: the first also admitted CSF outside that window when the participant’s biomarker status could be inferred not to have changed (82 participants), and the second widened the window to two years under the same rule (93). With DoWhy’s two-stage linear regression, effects are differences in the probability of MCI at T1, and the Natural Indirect Effect (NIE) is the part of an exposure’s effect transmitted through the mediator. Each exposure–mediator pair was fitted without adjustment and with adjustment for age, sex, or both: 224 models per dataset, 672 in all. We further examined change in memory and composite cognitive scores between T0 and T1 in 80 participants. We checked total effects (TE) with three refutation tests. We repeated the bootstrap (5,000 resamples) with five random seeds, and called an indirect effect robust when its 95% confidence interval excluded zero in all five.
Of the 672 models, 13 showed robust indirect effects, all mediated by CSF tau; none went through Aβ42, biomarker status, or the ratios, and none involved rhythm fragmentation or nighttime activity. Two pathways were robust across all three datasets: participants with higher amplitude or a higher M10–RA composite had a lower probability of MCI at T1, and part of each association ran through tTau. Per standard deviation of amplitude, the probability of MCI was 3.1 to 3.6 percentage points lower through tTau, out of an estimated total difference of 8 to 10 points. Across the two pathways, tTau accounted for 36–49% of the total effect; the remaining direct effects were imprecise, with intervals that included zero. Estimates were slightly smaller with the two-year window. The other robust models were variants of these: with pTau as the mediator, with M10 alone or a circadian disruption index (which contains amplitude and RA) as the exposure, or with adjustment for sex. pTau yielded indirect effects of similar magnitude but with wider intervals. Because the two tau measures were strongly correlated (r = 0.81–0.84), the results point to CSF tau in general, with tTau the most consistent. No indirect effect was detected for change in memory or composite cognitive scores.
The main qualification is confounding by age, which is associated with both rhythm amplitude and CSF tau. No age-adjusted model was robust: adjustment for age reduced the indirect effects through tTau by roughly 15–30% and widened their intervals to include zero. Age may therefore account for part of the pathway but, on these estimates, not most of it; samples of this size cannot settle the question.
These findings identify rhythm amplitude and the M10–RA composite, with CSF tau as a mediator, as candidate pathways linking weaker rest–activity rhythms to MCI. They should be tested first in a larger, independent cohort in which actigraphy precedes CSF sampling and clinical follow-up.
Examination Committee:
Dr. Elias S. Manolakos (advisor), Professor Emeritus, Dept. of Informatics and Telecommunications, NKUA
Dr. Nikolaos Scarmeas, Professor of Neurology, Department of Medicine, NKUA
Dr. Theodore Dalamagas, Research Director, Information Management Systems Institute, ATHENA Research Center
Elias Manolakos is inviting you to a scheduled Zoom meeting.
Join Zoom Meeting
https://us02web.zoom.us/j/82811794963?pwd=jCLGsqRtWuectyPRDLweaaltDcEGpI.1
Meeting ID: 828 1179 4963
Passcode: 097072
One tap mobile
+302311180599,,82811794963#,,,,*097072# Greece
+302111984488,,82811794963#,,,,*097072# Greece
Join by SIP
• 82811794963@zoomcrc.com
Join instructions
https://us02web.zoom.us/meetings/82811794963/invitations?signature=Mz1W02tQBqRPVudmn7_n_stGwOh_8GXl0VAjrbYgaFU
Course and Master’s Thesis Registrations for the Winter Semester 2026-27
Αθήνα 29.10.2026
Δηλώσεις μεταπτυχιακών μαθημάτων και Διπλωματικών Εργασιών Χειμερινού εξαμήνου 2026-27
Καλούνται οι μεταπτυχιακοί φοιτητές ΜΕ ΕΤΗ ΕΙΣΑΓΩΓΗΣ 2018, 2019 ΚΑΙ 2020 τoυ ΔΠΜΣ:
Data Science and Information Technologies
να δηλώσουν έως 5 μαθήματα χειμερινού εξαμήνου ή/και Διπλωματική Εργασία, από τις 29.10.2026 ως τις 26.11.2026 στο Πληροφοριακό Σύστημα Διαχείρισης Μεταπτυχιακών Σπουδών:
https://pgs.di.uoa.gr
Καλούνται οι μεταπτυχιακοί φοιτητές ΜΕ ΕΤΗ ΕΙΣΑΓΩΓΗΣ 2021, 2022, 2023, 2024,2025 και 2026 του ΔΠΜΣ να δηλώσουν έως 5 μαθήματα χειμερινού εξαμήνου ή/και Διπλωματική Εργασία, από τις από τις 29.10.2026 ως τις 26.11.2026 στην Πύλη Φοιτητολογίου του ΕΚΠΑ:
https://my-uni.uoa.gr/
Η δήλωση μαθημάτων είναι υποχρεωτική για όλους τους μεταπτυχιακούς φοιτητές.
———————————————————
TRANSLATED – FOR INTERNATIONAL STUDENTS
Athens, 29.10.2026
Course and Master's Thesis Registrations for the Winter Semester 2026-27
• Postgraduate students YEARS OF ADMISSION 2021, 2022, 2023, 2024, 2025, and 2026 of the following MSc and Interinstitutional MSc program:
Data Science and Information Technologies
are invited to register for up to 5 winter semester courses and/or a Master's Thesis, from 29.10.2026 to 26.11.2026, in the Postgraduate Studies Management Information System: https://pgs.di.uoa.gr
• Course registration is mandatory for all postgraduate students.
Registration is submitted through the NKUA Student Portal (My Uni):
https://my-uni.uoa.gr/
From Secretariat of NKUA Department of Informatics and Telecommunications. Email: secretpm@di.uoa.gr
MSc Thesis presentation of Eleni Panagiotopoulou Thursday 22/10/2026, 3pm
MSc Thesis presentation of Eleni Panagiotopoulou Thursday 22/10/2026, 3pm
On Thursday 22/a0/2025 at 3 pm, Eleni Panagiotopoulou, a graduate student in the program “Data Science and Information Technologies, Biomedical Data Science – Bioinformatics” will present their MSc thesis titled:
Actigraphy Feature Representations for Mild Cognitive Impairment Classification: Comparing Summary, Circadian and Wavelet Features across Sensitivity–Specificity Operating Points
Mild cognitive impairment (MCI) is characterized by cognitive deficits that exceed those expected with normal aging without substantially compromising independence in everyday activities. Actigraphy provides a non-invasive means of monitoring rest-activity patterns and may offer additional information for identifying cognitive impairment. This thesis investigated how different actigraphy-derived feature representations distinguish participants with MCI from those with normal cognition.
We compared three feature families: summary statistics, combined cosinor and non-parametric rest-activity descriptors, and discrete wavelet transform (DWT) features. An integrated software pipeline supported signal preprocessing, feature extraction, feature selection, and evaluation of multiple classical machine-learning algorithms within a repeated nested cross-validation framework. We compared configurations that met predefined performance criteria using recall-specificity Pareto fronts and parsimony-based selection. Additional analyses examined age-only models, matched signal lengths, time-of-day windows, model complexity, and hyperparameter tuning.
The feature families supported different recall-specificity trade-offs. Summary statistics achieved the highest median recall, reaching 0.818 at a median specificity of 0.700. Combined cosinor and non-parametric descriptors achieved recall of 0.800 and specificity of 0.750, while DWT features achieved the highest specificity (0.850) with recall of 0.727. Age-only models provided a competitive reference, but incorporating actigraphy features improved the attainable trade-offs. Compact actigraphy representations for wavelet feature extraction retained high specificity, although with lower recall than the leading full-series configuration. Simpler models and configurations without hyperparameter tuning also achieved competitive performance.
These findings indicate that the preferred actigraphy representation depends on the sensitivity-specificity balance sought, rather than a single feature family being universally superior. The thesis contributes an extended analysis pipeline and a systematic comparison of complementary activity representations for MCI classification. The results encourage further investigation of actigraphy as a source of information for clinical decision support, while the limited cohort and exploratory configuration-selection process necessitate independent validation before clinical application.
Examination Committee:
Prof. Em. Elias S. Manolakos (advisor), Dept. of Informatics and Telecommunications, NKUA
Dr. Nikolaos Scarmeas, Professor of Neurology, Department of Medicine, NKUA
Prof. Dimitrios J. Stravopodis, Assoc. Professor of Biology, Department of Biology, NKUA
Meeting Details
Elias Manolakos is inviting you to a scheduled Zoom meeting.
Join Zoom Meeting
https://us02web.zoom.us/j/82515571992?pwd=g0abksV0wrVJmptP5QngbNWQP4DyeR.1
Meeting ID: 825 1557 1992
Passcode: 082617
—
One tap mobile
+302111984488,,82515571992#,,,,*082617# Greece
+302311180599,,82515571992#,,,,*082617# Greece
—
Join by SIP
• 82515571992@zoomcrc.com
Join instructions
https://us02web.zoom.us/meetings/82515571992/invitations?signature=kBwy6M_-PrW9XDPf2ODz5KTSMbQp4zd4TpEBkfH4Ixo
2026-10-19_12-05-58_251da878e80b_4_thesis_presentation_announcement_Eleni_Panagiotopoulou.docx
MSc Thesis presentation of Dimitrios Krikonis, Wednesday, October 21, 2026
MSc Thesis presentation of Dimitrios Krikonis, Wednesday, October 21, 2026
Student: Dimitrios Krikonis
Program: Data Science and Information Technologies / Big Data and
Artificial Intelligence
Title:
Machine Learning for Archaeology: A Human-in-the-Loop Chronology
Prediction System for Pottery from the Athenian Agora
Abstract:
This thesis investigates the use of machine learning as a decision-support
tool for archaeological pottery dating, aiming to provide initial
estimates without replacing expert judgement. First, it conducts a
comparative evaluation of text- and image-processing methods and different
classification and regression algorithms, using 1,910 pottery items from
the The Athenian Agora publication series. Five feature extraction methods
are compared across two modalities: TF-IDF and BERT for text, Canny with
HOG, ResNet and ViT for images, along with their six multimodal
combinations, giving eleven feature sets in total. The results show that
text-based methods outperform image-based approaches, while combining the
two does not provide a reliable improvement, possibly due to the limited
availability of photographs. TF-IDF performs better than BERT on the
dataset, and Vision Transformer (ViT) achieves the strongest results among
image representations. Additionally, a web application was designed and
developed (https://dimitris-kr.github.io/pottery-chronology-predictor/)
that integrates the selected models into a predict–validate–retrain
workflow, allowing archaeologists to review predictions, examine their
uncertainty, validate or correct chronologies, and contribute data for
future model retraining. The system was evaluated by its intended users,
who successfully completed the full workflow. However, whether successive
retraining cycles improve predictive performance remains an open question,
requiring additional validated data collected over time and a dedicated
evaluation methodology.
Date/Time: October 21, 2026 – 13:05 PM EEST
Examination Committee:
Maria Roussou (supervisor), Associate Professor, University of Athens
Alexandros Ntoulas, Assistant Professor, University of Athens
Ioannis Panagakis, Associate Professor, University of Athens
Presentation Link:
https://uoa.webex.com/meet/mariar
MSc Thesis presentation of Ms. Myrto Potamiti Friday, October 16, 2026
MSc Thesis presentation of Ms. Myrto Potamiti Friday, October 16, 2026
Student: “Myrto Potamiti”
Program: “Data Science and Information Technologies”
Title: “MSTOK: Adapting Discrete Visual Tokenization to Multispectral Earth Observation Imagery”
Abstract:
Recent advances in generative modelling for Earth observation highlight the importance of compact image representations that preserve spatial and spectral information. This thesis presents MSTok, an adaptation of the Open-MAGVIT2 visual tokenizer using lookup-free quantization (LFQ) for 12-band Sentinel-2 Level-2A imagery from TerraMesh. Experiments investigate reflectance preprocessing, discrete code-space capacity, ImageNet-pretrained initialization, architectural adaptations, and joint ImageNet–TerraMesh training. Reconstruction quality is assessed
on 89,088 validation samples at 224×224 resolution using multispectral and RGB perceptual metrics. After one epoch, the standard model trained from scratch outperformed its
ImageNet-initialized counterpart across the reported reconstruction metrics, achieving an FID of 61.38 versus 69.53. A recurring trade-off emerged between multispectral accuracy and RGB perceptual quality: extending linear-adapter training reduced all-band MAE from 0.0480 to 0.0469, while increasing FID from 29.26 to 39.18. High code utilization and accurate standalone adapter reconstruction did not ensure effective end-to-end tokenization. The from-scratch configuration was retained as the preferred overall model within the available training budget.
These findings demonstrate the feasibility of LFQ-based multispectral tokenization and the need to assess multispectral reconstruction fidelity and visible-band perceptual quality jointly.
Date/Time: October 16, 2026 – 14:00 PM.
Examination Committee:
”
Dr. Bill Psomas
Dr. Stavros Perantonis
Dr. Ioannis Kakogeorgiou
“
Presentation link:
https://meet.google.com/dvs-gzfy-nkq