This research poster from Technological University Dublin (TU Dublin) describes a mobile AI application designed for early cognitive monitoring. ## Title and Overview The main heading at the top states: "Mobile AI uses the natural way you speak about your photos to identify the hidden acoustic markers of cognitive change." ## Content Sections * **Background:** Explains the shift from episodic testing to continuous, unobtrusive monitoring in domestic settings to identify neurodegenerative trajectories. * **Methods:** Details a pipeline that extracts language-agnostic acoustic and prosodic characteristics from audio recordings to serve as inputs for machine learning classification models. * **Visuals:** * **Figure 1:** An illustration of the user workflow: taking a photo, receiving a prompt at a later time, and recording a verbal description. * **Figure 2:** Displays speech features including waveforms, Mel spectrograms, MFCCs, and fundamental frequency. * **Figure 3 & 4:** Bar charts showing model performance (Mean Balanced Accuracy) across different classification models like Logistic Regression and Random Forest, including cross-linguistic performance in English, Greek, and Mandarin. ## Footer Information The bottom of the poster lists the research team, led by Giulio Gabrieli, and provides a QR code for supplemental materials. The TU Dublin logo is prominently displayed in the bottom right corner.
On my way to beautiful Galway for #Dementia Research Network #Ireland (DRNI) Early Career Researcher Day. Presenting a poster on how to use biomarkers for the identification of cognitive impairment using a gamification approach. Come meet me :)
#artificialIntelligence #Neuroscience #Dementia
