Yusuf Brima

I am a biomedical AI researcher. I build trustworthy machine learning for medicine: models that learn robust, multimodal representations of imaging and clinical data, explain their decisions, and generalise to the settings where they are used, including low-resource health systems.

Since January 2025 I have been a Biomedical AI Research Fellow in a research collaboration with Rhodes University and global academic partners, working on multimodal learning for early diagnosis and prognosis, and co-supervising students in medical imaging and clinical AI. I received my Ph.D. in Cognitive Science from Osnabrück University in 2025 (advisors: Prof. Gunther Heidemann and Prof. Simone Pika), with a thesis on disentangled representation learning in speech and vocalization. Before that I studied at AIMS Rwanda (M.Sc. Mathematical Sciences), the University of Dhaka (M.Sc. Computer Science and Engineering) and the University of Makeni (B.Sc. Computer Science).

Research interests: medical image analysis · multimodal clinical AI · trustworthy and explainable AI · robustness and generalisation · representation learning · AI for global health.

Multimodal cross-attention visualization of Grad-CAM++ for a trained multimodal deep learning model for Alzheimer's prediction on ADNI subjects.

Sankey diagram of research funding sources, disease areas, and institutional affiliations for AI in Neglected Tropical Diseases (NTDs).

Sankey diagram of research funding sources, disease areas, and institutional affiliations AI in Neglected Tropical Diseases (NTDs) between 2015 to 2025.

Kullback–Leibler divergence between two probability distributions.

A web-based DICOM viewer enabling medical image visualization and basic manipulations, such as contrast adjustment, export to png and video (for multiple uploaded files).For more details, visit the project repository: DICOMViewer

Time-domain, frequency-domain and time-frequency representations of a sine waveform across different frequencies sampled at 16KHz for 1 seconds. The top row is the original signal and the bottom row is the reconstructed version for the disentangled Variational Auto Encoder.

Time-domain, time-frequency and frequency-domain representations of a sawtooth waveform across different frequencies sampled at 16KHz for 2 seconds each.

Visualizing the convergence of clusters of classes in the MNIST validation set using both Triplet (Left) and Supervised Contrastive Representation Learning

Visualizing the convergence of latent representations over training iterations for modified Barlow Twins model: decorrelated audio representations (left) and original small training data (right). The 128-dimensional latent space reveals the progression of learned features and their alignment throughout the training process.

A visual illustration of two divergence measures: the Kullback–Leibler (KL) and Jensen–Shannon (JS) divergences between two probability distributions $P_x $ and $Q_x$.

Explainable AI in Medical Image Analysis using Visual Saliency Maps. Feature attribution was done with a trained InceptionResNetv2 model.

Learnt 2D and 3D principal component latent representations of both human and non-human primates.

Learnt 2D and 3D principal component latent representations of speeches at the United States Congress by five world leaders.

news

Oct 07, 2026	Received the Lambda AI Research Grant for supporting large-scale multimodal biomedical AI research for infectious diseases.
Sep 14, 2026	Our manuscript titled “A Multimodal Explainable Deep Learning Framework for Alzheimer’s Disease Diagnosis using 3D Magnetic Resonance Imaging and Clinical Data” got accepted for publication in Springer Nature Cognitive Computation.
Sep 02, 2026	Our manuscript titled “From Visual Saliency to Anatomical Explanation: Generating Textual Diagnostic Reports from Brain MRIs with Saliency-Grounded LLMs” got accepted for publication in Springer Nature BioData Mining.
Aug 07, 2026	I am excited to be serving as program chair of the Pan-African Artificial Intelligence and Smart Systems Conference (PA-AISS 2026) taking place at the University of Technology, Kingston, Jamaica October 13-16, 2026.
Jul 21, 2026	My research project was awarded a Google Cloud Research Grant to fully fund our high-performance academic computing, storage, and data analytics infrastructure for the next year.
May 26, 2026	Our Tabular Foundation Model manuscript titled “Few-shot Learning with Cross-country Generalization of Tabular Machine Learning and Foundation Models for Childhood Anemia Prediction under Distribution Shift” submitted to BioData Mining.
May 19, 2026	Our systematic review title “Transformers for 3D Medical Image Analysis: A Systematic Review of Architectural Innovations, Performance, and Clinical Applications,” got accepted at Artificial Intelligence Review.
Apr 20, 2026	Our LLM-Guided XAI for Medical Image Diagnostics manuscript titled “Bridging Visual Saliency and Large Language Models for Self-explainable Deep Learning in Medical Imaging” submitted to BioData Mining.
Apr 16, 2026	Our AI in NTD Consortium manuscript titled “Artificial Intelligence in Neglected Tropical Diseases: Current Applications, Challenges, and Opportunities” submitted to npj Digital Medicine.
Dec 24, 2025	Completed “ Evaluations of AI Applications in Healthcare by Stanford Online” with a focus on AI techniques and their applications in medical diagnosis. Certificate available at here.
Dec 24, 2025	Completed “ AI Applications in Healthcare Specialization by Stanford Online” with a focus on AI techniques and their applications in biomedicine. Certificate available at here.
Dec 18, 2025	Completed “Fundamentals of Machine Learning for Healthcare by Stanford Online” with a focus on machine learning techniques and their applications in healthcare. Certificate available at here.
Dec 17, 2025	Completed “Introduction to Clinical Data by Stanford Online” with a focus on clinical data types, sources, and applications in healthcare. Certificate available at here.
Dec 06, 2025	Completed “Introduction to Healthcare by Stanford Online” with a focus on healthcare systems, quality improvement, and patient safety. Certificate available at here.
Nov 19, 2025	PhD Thesis Published in the collection: FB08 - E-Dissertations: “Disentangled Representation Learning in Speech and Vocalization
Aug 28, 2025	Gave a talk on Artificial Intelligence and Its Role in Shaping Our Future at the University of Makeni Computer Science Alumni Forum.
Jun 27, 2025	Successfully defended my PhD thesis at the Institute of Cognitive Science, Osnabrück University.
Mar 30, 2025	Made a first commit to a Pytorch-based XAI for Dementia using 3D CNN and NIFTI Data
Mar 04, 2025	A systematic review of low-rank and local low-rank matrix approximation in big data medical imaging” published for publication in Springer Nature Neural Computing and Applications.
Feb 19, 2025	Talk on “Multimodal Federated Learning for Robust Lung Cancer Prognosis” at the Fraunhofer Institute for Algorithms and Scientific Computing (SCAI).
Jan 31, 2025	Completed the MUST Deep Learning Bootcamp, North-West University, South Africa MUST Deep Learning Bootcamp 2025.
Jan 22, 2025	PhD Dissertation Submitted and approved for examination and graduation: “Disentangled Representation Learning in Speech and Vocalization”.
Jan 20, 2025	The Association of Commonwealth Universities Case Study on “Advancing artificial intelligence for healthcare and developing human capital in low-resource settings” published in ACU Research.
Jan 11, 2025	Learning Disentangled Speech Representations” preprint published in arXiv:2311.03389v4.
Dec 16, 2024	Talk on “Trustworthy Healthcare AI for Mental Health Risk Prediction” at the Neurobiology Research Unit,Copenhagen University Hospital
Dec 10, 2024	Talk on “Assessing Explainability in Deep Learning for Medical Image Analysis” at the Fraunhofer Institute for Algorithms and Scientific Computing (SCAI).
Nov 11, 2024	Launched SyncSpeech Datasets: A large-scale collection of synthetic speech datasets for speech representation learning. Available at SyncSpeech Datasets.
Jul 10, 2024	Talk on “Assessing fusarium oxysporum disease severity in cotton using unmanned aerial system images and a hybrid domain adaptation deep learning time series model” at the Leibniz Institute for Agricultural Engineering and Bioeconomy
Jun 22, 2024	Saliency-driven explainable deep learning in medical imaging: bridging visual explainability and statistical quantitative analysis” published in BioData Mining 17, 18.
Mar 27, 2024	Bridging the Gap: Exploring Interpretability in Deep Learning Models for Brain Tumor Detection and Diagnosis from MRI Images” published in MDPI Information 2024, 15(4), 182.
Mar 21, 2024	Got accepted to attend in-person the Deep Learning + Reinforcement Learning (DLRL) Summer School 2024 by CIFAR and the Vector Institute in Toronto, Canada.
Mar 13, 2024	Talk on “Causal Representation Learning” at the Computer Vision Colloquium, Osnabrueck University.
Feb 16, 2024	Learning Disentangled Audio Representations through Controlled Synthesis” accepted for oral presentation at ICLR Tiny Papers 2024.
Feb 15, 2024	Understanding Self-Supervised Learning of Speech Representation via Invariance and Redundancy Reduction” accepted for oral presentation at MDPI Information 2024, 15(2), 114.
Feb 12, 2024	Bridging the Gap: Exploring Interpretability in Deep Learning Models for Brain Tumor Classification from MRI Images” submitted to MDPI Information Special Issue: Deep Learning in Medical Image Analysis: Foundations, Techniques, and Applications.
Dec 10, 2023	Attended NeurIPS 2023 in New Orleans, USA where Presented research poster on disentangled speech representation learning via self-supervision. Sharing novel techniques for learning interpretable and robust speech representations.
Nov 01, 2023	Learning Disentangled Speech Representations” accepted for poster presentation at New in ML, NeurIPS 2023.
Oct 17, 2023	Co-organizing with Ulf Krumnack and Lukas Niehaus a Seminar: Deep Representation Learning (Winter 2023/24) at Osnabrück University.
Sep 23, 2023	Self-Supervised Learning of Speech Representation via Redundancy Reduction” extended abstract published at Gesellschaft für Informatik e.V..
Aug 01, 2023	Visual Interpretable and Explainable Deep Learning Models for Brain Tumor MRI and COVID-19 Chest X-ray Images” published in arXiv preprint arXiv:2208.00953.
Jul 26, 2023	Self-Supervised Learning of Speech Representation via Redundancy Reduction” accepted for oral presentation at KI 2023 – 46th German Conference on Artificial Intelligence, Berlin, Germany.
Jun 13, 2023	submitted a manuscript to PLOS One A Mathematical Framework for Understanding Recognition Systems.
Sep 16, 2022	Launched the Graduate Assistance Initiative Network (GAIN) globally with 130+ attendees.
Sep 11, 2022	Attended Mediterranean Machine Learning school, Università Milano Bicocca, Milan, Italy.
Aug 06, 2022	Co-organized with Viktoria Zemliak the “Bridging the Gap between Biological and Artificial Neural Networks” at the (Research Training Group (RTG) in Computational Cognition).
Jul 29, 2022	Attended and presented a poster at the Bridging the technological gap – spreading technological innovations in the study of the human and non-human mind at the German Primate Center in Göttingen, Germany.
Jul 11, 2022	Participated in the Neuromatch Academy: Deep Learning intensive hands-on training.
Aug 16, 2021	Transfer Learning for the Detection and Diagnosis of Types of Pneumonia Including Pneumonia Induced by the COVID-19 from Chest X-Ray Images” published in MDPI Special Issue on Machine Learning Applications for COVID-19 and Its Complications: Screening, Diagnosis, Treatment, and Prognosis.
Jun 15, 2021	Published a dataset that was curated in collaboration between the Computer Science and Engineering Department, University of Dhaka and the National Institute of Neuroscience, Bangladesh. It comprises 5,285 T1-weighted contrast-enhanced brain MRI images belonging to 38 categories. Available here Brain MRI Dataset.

selected publications

Saliency-driven explainable deep learning in medical imaging: bridging visual explainability and statistical quantitative analysis

Yusuf Brima and Marcellin Atemkeng

BioData Mining, Jun 2024

Abs DOI Bib HTML

@article{brima2024saliencydl,
  title = {Saliency-driven explainable deep learning in medical imaging: bridging visual explainability and statistical quantitative analysis},
  volume = {17},
  issn = {1756-0381},
  url = {http://dx.doi.org/10.1186/s13040-024-00370-4},
  doi = {10.1186/s13040-024-00370-4},
  number = {1},
  journal = {BioData Mining},
  publisher = {Springer Science and Business Media LLC},
  author = {Brima, Yusuf and Atemkeng, Marcellin},
  year = {2024},
  month = jun,
}

An unsupervised XAI framework for dementia detection with context enrichment

Devesh Singh, Yusuf Brima, Fedor Levin, and 44 more authors

Scientific Reports, Nov 2025

Abs DOI Bib HTML

This work introduces a framework that evaluates XAI methods by integrating neuroanatomical morphological features with CNN-generated relevance maps for disease classification. We trained a CNN using brain MRI scans from six cohorts: ADNI, AIBL, DELCODE, DESCRIBE, EDSD, and NIFD (N = 3253), including participants that were cognitively normal, with amnestic mild cognitive impairment, dementia due to Alzheimer’s disease and frontotemporal dementia

@article{brima2025xaidementia,
  title = {An unsupervised XAI framework for dementia detection with context enrichment},
  volume = {15},
  issn = {2045-2322},
  url = {http://dx.doi.org/10.1038/s41598-025-26227-2},
  doi = {10.1038/s41598-025-26227-2},
  number = {1},
  journal = {Scientific Reports},
  publisher = {Springer Science and Business Media LLC},
  author = {Singh, Devesh and Brima, Yusuf and Levin, Fedor and Becker, Martin and Hiller, Bjarne and Hermann, Andreas and Villar-Munoz, Irene and Beichert, Lukas and Bernhardt, Alexander and Buerger, Katharina and Butryn, Michaela and Dechent, Peter and Düzel, Emrah and Ewers, Michael and Fliessbach, Klaus and Freiesleben, Silka D. and Glanz, Wenzel and Hetzer, Stefan and Janowitz, Daniel and Görß, Doreen and Kilimann, Ingo and Kimmich, Okka and Laske, Christoph and Levin, Johannes and Lohse, Andrea and Luesebrink, Falk and Munk, Matthias and Perneczky, Robert and Peters, Oliver and Preis, Lukas and Priller, Josef and Prudlo, Johannes and Prychynenko, Diana and Rauchmann, Boris S. and Rostamzadeh, Ayda and Roy-Kluth, Nina and Scheffler, Klaus and Schneider, Anja and Droste zu Senden, Louise and Schott, Björn H. and Spottke, Annika and Synofzik, Matthis and Wiltfang, Jens and Jessen, Frank and Weber, Marc-André and Teipel, Stefan J. and Dyrba, Martin},
  year = {2025},
  month = nov,
}

An empirical study of machine learning robustness and scalability for imbalanced tabular clinical data in emergency and critical care

Yusuf Brima and Marcellin Atemkeng

Scientific Reports, Jun 2026

Abs DOI Bib HTML

Machine learning can support high-stakes decision-making in emergency and intensive care settings, but severe class imbalance in clinical data limits model reliability and biases predictions toward majority outcomes. We evaluate six model families, including classical methods (Decision Tree, Random Forest, XGBoost), deep learning approaches (TabNet), and tabular foundation models (TabICL, TabPFN v2.6), on MIMIC-IV-ED and eICU datasets across multiple clinical prediction tasks. Models are assessed using Macro F1-score, robustness to increasing imbalance, and computational efficiency. Results show dataset-dependent performance: TabPFN and TabICL perform strongly on MIMIC-IV-ED, while XGBoost leads on eICU. No single model dominates across all settings, but foundation models provide a favorable efficiency–performance trade-off and are increasingly competitive in imbalanced clinical prediction scenarios.
@article{brima2026robustness, title = {An empirical study of machine learning robustness and scalability for imbalanced tabular clinical data in emergency and critical care}, volume = {16}, issn = {2045-2322}, url = {http://dx.doi.org/10.1038/s41598-026-56413-9}, doi = {10.1038/s41598-026-56413-9}, number = {1}, journal = {Scientific Reports}, publisher = {Springer Science and Business Media LLC}, author = {Brima, Yusuf and Atemkeng, Marcellin}, year = {2026}, month = jun, }

Transformers for 3D medical image analysis: a systematic review of architectural innovations, performance, and clinical applications

Wandile Nhlapho, Marcellin Atemkeng, Yusuf Brima, and 4 more authors

Artificial Intelligence Review, Jul 2026

Abs DOI Bib HTML

A systematic review of 138 studies examining Transformer-based architectures for 3D medical image analysis across MRI, CT, PET, and ultrasound. The review synthesizes architectural innovations, benchmark performance, clinical applications, and reproducibility challenges while highlighting emerging trends such as hybrid CNN–Transformer models, State Space Models, and diffusion-based Transformers.

@article{brima2026transformers,
  title = {Transformers for 3D medical image analysis: a systematic review of architectural innovations, performance, and clinical applications},
  volume = {59},
  issn = {1573-7462},
  url = {http://dx.doi.org/10.1007/s10462-026-11600-y},
  doi = {10.1007/s10462-026-11600-y},
  number = {11},
  journal = {Artificial Intelligence Review},
  publisher = {Springer Science and Business Media LLC},
  author = {Nhlapho, Wandile and Atemkeng, Marcellin and Brima, Yusuf and Ravelo, Blaise and Bouwmans, Thierry and Lalléchère, Sébastien and Ndogmo, Jean-Claude},
  year = {2026},
  month = jul,
}