Part I — Completed TÜBİTAK 1002-A

Project iShe

An AI-Supported Communication and Cognitive Monitoring System for Older Adults

A MAARS Platform Project

Combining large language models, natural language processing, and voice interaction to create a companion system that monitors cognitive health and reduces loneliness in adults aged 65+.

About the Project

The Challenge

Cognitive decline and social isolation are among the most pressing health concerns for older adults worldwide. Traditional screening methods are often infrequent, clinic-dependent, and unable to capture subtle longitudinal changes in cognitive and emotional well-being.

Our Approach

Project iShe develops an AI-powered voice companion that engages older adults in natural daily conversations. Through dialogue, the system performs continuous cognitive screening, monitors emotional states via NLP-based sentiment analysis, and reduces loneliness through meaningful social engagement.

Key Features

Voice-First Design
Natural spoken dialogue optimized for elderly users — no complex UI required
Continuous Monitoring
Daily cognitive and emotional assessment embedded within natural conversation
Loneliness Reduction
Empathetic AI companion providing social engagement and emotional support
Mobile App (BYOD)
iOS & Android — participants use their own smartphones

Part I — Pilot Study

Completed

The pilot study (TÜBİTAK 1002-A, Project No: 324S300) established the foundational framework for the iShe system. Part I demonstrated the feasibility and potential of AI-powered voice interaction with older adults and validated the core technological approach.

Key Outcomes

System architecture designed — LLM-based conversational AI with voice interface
NLP-based sentiment analysis pipeline developed and tested
Cognitive screening methodology embedded within natural dialogue
Feasibility and user acceptance validated with older adult participants
Research paper published
Publication

PROJECT iShe: An AI-Supported Communication and Cognitive Monitoring System for Older Adults

Part I — System Design and Pilot Validation

324S300 TÜBİTAK 1002-A Published
Supported by

The Scientific and Technological Research Council of Türkiye (TÜBİTAK)

Technology

LLM Voice Dialogue

Large Language Model-powered conversational AI that engages older adults in natural, empathetic daily dialogues through voice interaction on iOS and Android.

Cognitive Screening

Automated cognitive assessment embedded within natural conversation, enabling continuous longitudinal monitoring without clinical visit dependency.

NLP Sentiment Analysis

Real-time emotion and mood detection through natural language processing, tracking psychological well-being and detecting early signs of depression or distress.

Research Team

A multidisciplinary team combining expertise in neurosurgery, anesthesiology, geriatrics, computer engineering, and data science.

ÖA

Assoc. Prof. Özgür Akşan, MD

Project Lead & Architect

Department of Neurosurgery

İstanbul Aydın University

AK

Assoc. Prof. Aslı Kılavuz, MD

Researcher — Geriatrics

Department of Geriatrics

Ege University

FA

Feryal Akşan, MD, PhD

Principal Investigator

Department of Anesthesiology and Reanimation

İstanbul Aydın University

ŞB

Assoc. Prof. Şebnem Bora

Researcher — Computer Engineering

Department of Computer Engineering

Ege University

BB

Batu Bora, MSc

Researcher — Data Science

Department of Computing Science

University of Alberta, Canada

Participating Institutions

İstanbul Aydın University

Faculty of Medicine, İstanbul

Ege University

Faculty of Medicine & Engineering, İzmir

University of Alberta

Computing Science, Edmonton, Canada

Part II — Coming Soon

Building on the success of Part I, we are preparing a large-scale randomized controlled trial to evaluate the clinical effectiveness of the iShe system. Part II will expand the scope with a multicenter RCT design, validated outcome measures, and a broader participant cohort.

TÜBİTAK 1001 Application in Progress

Contact

For inquiries about the project, collaboration opportunities, or participation, please reach out to us.

Affiliated Institutions

İstanbul Aydın University, Faculty of Medicine
Department of Neurosurgery & Anesthesiology, İstanbul, Türkiye

Ege University, Faculty of Medicine
Department of Geriatrics, İzmir, Türkiye

University of Alberta
Department of Computing Science, Edmonton, Canada