Blessing I. Adeika
Research

Restoring speechwithout surgery.

Trustworthy AI and neuroscience for people who have lost the ability to speak.

Focus

Many people who lose speech remain fully aware of what they want to say. My research asks how AI can help restore that voice safely, fairly, and without surgery.

Current research · C4 Lab, Morgan State

Making EEG speech decoding trustworthy.

Selected results from my published preprints.

39participants across ds003626, KaraOne and the 2020 BCI Competition Track 3
0.743overt-speech accuracy from two lip electrodes, against 0.423 for the 128-channel scalp montage
0.54 μVmu/beta desynchronization detectable by band-power decoding
482participant pairs tested for cross-subject transfer of imagined speech
What comes next

Stay tuned.

This work is ongoing. I’ll share updates here with each new discovery and paper.

Follow on ScholarConnect
Projects & collaborations

Across health, hardware and learning.

Beckman Institute, UIUC

fMRI with vocal-tract imaging

Multimodal speech imaging in collaboration with the Sutton Lab.

Health AI

Equitable skin cancer detection

Explainable, bias-mitigated models for accurate diagnosis across Fitzpatrick skin types.

With NASA JPL

GAN-based RF circuit generation

A generator that maps S-parameter files to simulated RF circuit layouts.

STEM education

AI for classroom engagement

Computer-vision methods for measuring classroom engagement, published across several ASEE papers.

Publications

Preprints & papers.

Google Scholar
Preprint · Oct 2026arXiv:2610.20954Cross-subject decoding of imagined speech from scalp EEG: a leakage-controlled evaluation on three public corporaAdeika, B. I. & Osuagwu, O. · To be submitted to Journal of Neural EngineeringPreprint · Oct 2026arXiv:2610.20942How large an effect would we have seen? Positive controls, detection floors and dataset pitfalls in scalp-EEG speech decodingAdeika, B. I. & Osuagwu, O. · To be submitted to Journal of Neural Engineering
Latest from ORCID· syncing…
See all works on ORCID
2025
Fusion of COPUS and DEI Tools for Equitable STEM Classroom EngagementASEE Annual Conference & Exposition
2025
Perception of the Impact of Generative Artificial Intelligence on EducationASEE Annual Conference & Exposition
2025
Motivation, Challenges and Future Aspirations of Women in Engineering at an HBCUASEE Annual Conference & Exposition
2024
Breaking Barriers: Promoting Motivation, Engagement, and Learning Success among Biology Undergraduates from Minority BackgroundsASEE Annual Conference · Division Best Student Paper Award
2024
Transforming Pedagogical Assessment: AI and Computer Vision-Enhanced Classroom Observations for Experiment-Centric Learning EnvironmentsASEE Annual Conference & Exposition
2024
Impact of Experimental Centric Pedagogy on Learning Outcomes: A Comparative Trend Analysis in Industrial Engineering and BiologyASEE Annual Conference & Exposition
2023
Ensemble and Transformer Models for Infectious Disease PredictionIEEE International Conference on Bioinformatics and Bioengineering (BIBE)
2023
Increasing Student Motivation and Learning by Adopting Experiment-Centric Pedagogy: A Case of Undergraduates in BiologyASEE Annual Conference & Exposition