Andrea Buccoliero

Dottorato in Scienze Umane - 38° ciclo (1 ottobre 2022 - 30 settembre 2025)

Programma di ricerca dottorato

_________________________________________________________________

Artificial Intelligence and Speech Emotion Recognition in Digital Healthcare

Research Overview – Andrea Buccoliero

My doctoral research focuses on the intersection of Artificial Intelligence (AI), Speech Emotion Recognition (SER), and Digital Healthcare, with the goal of developing innovative solutions for the detection and monitoring of emotional and mental health conditions. Through the integration of machine learning algorithms and voice biomarkers, my work aims to enhance early diagnosis and personalized interventions for disorders such as burnout, depression, and stress-related conditions.

Research Objectives

1. Assess the Role of Digital Biomarkers in Mental Health

   • Investigating how speech features (e.g., MFCCs, pitch, speech rate, prosody) can serve as indicators of psychological well-being.
   • Comparing traditional psychological assessments with AI-based emotion recognition.

2. Develop AI-Powered Speech Emotion Recognition (SER) Models

   • Leveraging acoustic and linguistic biomarkers to detect emotional states.
   • Applying Deep Learning (DL) and Explainable AI (XAI) to improve model transparency and interpretability.

3. Implement SER in Clinical and Occupational Settings

   • Developing digital tools for healthcare professionals to enhance patient monitoring and intervention strategies.
   • Exploring real-world applications in workplace mental health, particularly in burnout detection and stress assessment.

Methodology and Technological Innovation

This research employs Speech Emotion Recognition (SER) and machine learning to assess emotional and mental health conditions. The methodology integrates:

1. Speech Feature Analysis

   • Extraction of acoustic markers (e.g., MFCCs, pitch, speech rate, prosody) linked to emotional expression.

2. AI-Driven Emotion Recognition

   • Application of machine learning and deep learning models to detect emotional states.
   • Use of Explainable AI (XAI), to interpret AI decision-making and enhance model transparency.

3. Validation and Psychological Assessment

   • Comparing AI-based predictions with self-reported emotional states and clinical evaluations.
   • Exploring applications in healthcare and occupational psychology, particularly in burnout detection and stress monitoring.

By integrating psychological theory with AI-driven tools, this research enhances emotion assessment and mental health interventions, offering scalable, interpretable, and ethical digital solutions.

Impact and Future Applications

This research contributes to the future of AI-driven mental health diagnostics, bridging the gap between technology and psychology. By refining SER models and digital biomarkers, my work has the potential to:

• Improve early screening for emotional distress and mental health disorders.
• Assist in the development of personalized digital therapies.
• Enhance workplace mental health programs by providing real-time burnout detection tools.

Through this interdisciplinary approach, my research supports the advancement of digital healthcare solutions, empowering both clinicians and individuals with innovative, AI-driven tools for mental well-being and emotional resilience.

Referente dottorato

Riccardo Sartori

Tutori dottorato

Andrea Ceschi