Biology subjects

Androulakis, I. P.

Publications and source records attributed to Androulakis, I. P..

3 recordsLinked to original sources

bioRxiv2025

Deep Learning Meets Sleep Medicine: A Proof-of-Concept Clustering of Minute-Resolution CPAP Telemetry

This proof-of-concept study demonstrates that minute-resolution telemetry from continuous positive airway pressure (CPAP) devices can be effectively repurposed for large-scale chronotype and adherence phenotyping. We collated 30 consecutive nights from n=200 de-identified ResMed patients into 30 x 1440 colour images, embedded each image with a frozen ResNet-50 convolutional neural network, and clustered the embeddings with k -means. Six distinct phenotypes emerged, capturing both sleep timing (early birds, typical sleepers, night owls) and adherence patterns (high, medium-high, inconsistent, fragmented, and non-adherent). The approach leverages routinely collected clinical data without the need for additional sensors, promising significant benefits for personalized sleep medicine. External validation against actigraphy and questionnaire-based chronotype measures is planned to further strengthen these findings.

bioengineering↗

bioRxiv2024

Characterization of Chronotypes Using the Symbolic Aggregate approXimation (SAX) on Actigraphy Data

In this study, we discuss an efficient approach to characterizing chronotypes using Symbolic Aggregate approXimation (SAX) on actigraphy data. Actigraphy, a non-invasive monitoring of human rest/activity cycles, provides valuable insights into sleep-wake behaviors and circadian rhythms. However, the high dimensionality of actigraphy data poses significant challenges in storage, processing, and analysis. To address these challenges, we applied the SAX algorithm to transform continuous time-series actigraphy data into a symbolic representation, enabling dimensionality reduction while preserving essential patterns. We analyzed actigraphy data from the National Health and Nutrition Examination Survey (NHANES) database, covering over 10,000 individuals, and used unsupervised clustering to identify distinct chronotype patterns. The SAX transformation facilitated the application of machine learning techniques, revealing five chronotype clusters characterized by differences in activity onset, resolution, and intensity. Age distribution analysis showed biases towards specific age groups within the clusters, highlighting the relationship between age and chronotype. Key findings include age-related Chronotype variations with younger individuals exhibiting delayed chronotypes with significant differences in sleep onset (SOT) and wake time (WT) compared to older adults, suggesting a phase delay in sleep patterns as age decreases and activity transition dynamics where clusters showed distinct patterns in winding up and winding down periods, providing insights into the dynamics of activity transitions. This study demonstrates the efficiency and effectiveness of SAX in processing large-scale actigraphy data, enabling robust chronotype characterization that can inform personalized healthcare and public health initiatives. Further exploration of SAX integration with other biometric measures could deepen our understanding of human circadian biology and its impact on health and behavior.

systems biology↗

bioRxiv2024

Investigating the Synergistic Role of GCN2 and HPA Axis in Regulating Integrated Stress Response in the Central Circadian Timing System

The circadian timing and integrated stress response (ISR) systems are fundamental regulatory mechanisms that maintain body homeostasis. The central circadian pacemaker in the suprachiasmatic nucleus (SCN) governs daily rhythms through interactions with peripheral oscillators via the hypothalamus-pituitary-adrenal (HPA) axis. On the other hand, ISR signaling is pivotal for preserving cellular homeostasis in response to physiological changes. Notably, disrupted circadian rhythms are observed in cases of impaired ISR signaling. In this work, we examine the potential interplay between the central circadian system and the ISR, mainly through the SCN and HPA axis. We introduce a semi-mechanistic mathematical model to delineate the suprachiasmatic nucleus (SCN)s capacity for indirectly perceiving physiological stress through glucocorticoid-mediated feedback from the HPA axis, and orchestrating a cellular response via the ISR mechanism. Key components of our investigation include evaluating general control nonderepressible 2 (GCN2) expression in the SCN, the effect of physiological stress stimuli on the HPA axis, and the interconnected feedback between the HPA and SCN. Simulation reveals a critical role for GCN2 in linking ISR with circadian rhythms. Notably, a Gcn2 deletion in mice led to swift re-entrainment of the circadian clock post simulated-jetlag. This is attributed to the diminished robustness of neuronal oscillators and an extended circadian period. Our model also offers insights into phase shifts induced by acute physiological stress and the alignment/misalignment of physiological stress with external light-dark cues. Such understanding aids in strategizing responses to stressful events, such as nutritional status changes and jetlag.

systems biology↗