At its core, my research seeks to answer: Can we create scalable, accessible biomarkers that capture what only PET scans and autopsy could?

Khalid Saifullah

Google Scholar Researchgate LinkedIn Resume Github Twitter

What drives me is simple: turning complexity into clarity. I approach research as I do people — with patience, curiosity, and a collaborative spirit. Whether debugging an algorithm, building a pipeline, or mapping atrophy across hundreds of brains, my goal is always the same: to make the complex understandable, the hidden visible, and the impossible just a little more possible.

About Me

I am a researcher at the intersection of neuroimaging, machine learning, and neuropathology, with a mission to uncover how Alzheimer’s disease and related dementias reshape the human brain — and to transform these discoveries into scalable tools for detection and precision care.

At the Medical Imaging Research Center (MRIIT), Illinois Institute of Technology, under the mentorship of Dr. Konstantinos Arfanakis, and in close collaboration with the Rush Alzheimer’s Disease Center, I work with one of the world’s largest autopsy-confirmed dementia cohorts. This unique setting allows me to link MRI scans, cognition, plasma biomarkers, and detailed pathology to reveal how hidden disease processes manifest in brain structure.

🔑 Key Contributions

Deep Gray Matter Biomarkers
My first major work, published in Human Brain Mapping (Neuropathological Correlates of Volume and Shape of Deep Gray Matter Structures in Community-Based Older Adults), showed how multiple neuropathologies — including Alzheimer’s, LATE, vascular disease, and Lewy bodies — independently alter the shape and volume of deep brain structures. This was the first study to disentangle the effects of co-existing pathologies at scale, advancing biomarker development for early diagnosis and clinical trial design.
AD vs. LATE Signatures
My submitted paper in Alzheimer’s & Dementia (Differential Patterns of Brain Atrophy in Alzheimer’s Disease and LATE Neuropathology: A Deformation-Based Morphometry Study) demonstrates that AD and LATE produce distinct atrophy signatures: AD primarily affects posterior medial temporal regions, while LATE causes more severe anterior hippocampal shrinkage — even in non-demented individuals. When both occur together, they act synergistically, producing widespread cortical and subcortical degeneration. These findings refine how we differentially diagnose and study overlapping dementias.
TANGLE: MRI-Based Tau Biomarker
I am now developing TANGLE: an MRI-based, fully automated, in-vivo marker of neurofibrillary tangles, one of the central hallmarks of Alzheimer’s disease. Like ARTS for arteriolosclerosis and MARBLE for LATE, TANGLE is trained on ex-vivo MRI linked with autopsy-confirmed pathology, then translated into in-vivo MRI applications. The pipeline is containerized, enabling reproducible deployment across research centers worldwide.

Preliminary results show that TANGLE achieves strong predictive performance in distinguishing brains with advanced tau pathology from those with minimal tangles, using features such as deformation-based morphometry, lobar volumes, and diffusion metrics. Once validated in-vivo, TANGLE will provide clinicians and researchers with a non-invasive imaging biomarker of tau pathology, filling a critical gap in Alzheimer’s research where tau PET is costly and inaccessible.

🌱 What I Bring & What I’m Looking For

I bring a strong track record of working with large, autopsy-confirmed cohorts, developing scalable MRI pipelines, and applying machine learning to neuropathology-linked imaging. My work has already produced new insights into how Alzheimer’s disease and LATE leave distinct structural fingerprints on the brain and is advancing toward in-vivo biomarkers of tau pathology.

At this stage of my career, I am seeking a postdoctoral position that allows me to deepen my expertise in multimodal data — including plasma biomarkers, multi-omics integration, and PET imaging — while continuing to build translational tools for dementia research.

In the long term, I aim to bring these skills into industry, where I can contribute to the development of next-generation diagnostic and therapeutic platforms for Alzheimer’s disease and related dementias.

Research Interests

Alzheimer's Disease and Neurodegenerative Disorders

Machine Learning and Deep Learning: Development of AI models for disease classification and prediction, including MRI-based classifiers and biomarker integration

Neuroimaging Techniques: Advanced analysis of MRI data (e.g., ex-vivo and in-vivo), deformation-based morphometry (DBM), shape analysis of brain structures

Biomarker Analysis: Exploring plasma biomarkers (e.g., Ptau, amyloid-beta, NfL) and their correlation with neuropathological markers for early diagnosis

Diagnostic and Therapeutic Applications: Leveraging AI and neuroimaging for improved diagnosis, patient stratification, and therapeutic interventions in Alzheimer’s and Limbic-predominant Age-related TDP-43 Encephalopathy Neuropathological Change (LATE-NC)

Experience

Graduate Research Assistant at MIRC, Dept. of BME, IIT, USA Aug 2020 – Present
1. 1. 1. 1. Applied advanced statistical models (e.g., linear regression, mixed-effects modeling) and machine learning algorithms to analyze over 900 ex-vivo MRI datasets, improving the understanding of Alzheimer's Disease and LATE pathology contributing to early detection and better treatment planning for Alzheimer’s patients
        Developed classifiers incorporating biomarker data (e.g., tau, amyloid-beta, neurofilament light chain) and imaging modalities, enhancing neuroimaging-based diagnosis by 15% and managing 1.5 TB of multi-modality neuroimaging and pathology data, reducing patient misdiagnosis and improving clinical trial participant selection
        Collaborated with cross-functional research teams from the Rush Alzheimer’s Disease Center (RADC), contributing to 2 peer-reviewed publications (under review) and presenting research at 11 international conferences, demonstrating expertise in biomedical data analysis and neurodegenerative disease research
Graduate Teaching Assistant - Dept. of Biomedical Engineering, IIT, USA Jan 2022 – May 2022
1. 1. 1. 1. Instructed 30+ students on application of AI techniques in neuroimaging, supervising labs on data preprocessing, modeling, and analysis.
        Guided students through end-to-end ML pipelines for image segmentation and statistical modeling, elevating project quality and enhancing reproducibility across 10+ biomedical research projects.
Lecturer, Dept. of EEE, Eastern University, Bangladesh Sept 2018 – Aug 2020
1. 1. 1. 1. Taught courses in Electrical and Electronic Engineering, focusing on signal processing and medical physics

Research

1. MRI-Based Classifier for Neurofibrillary Tangles in Alzheimer’s Disease [Ongoing]

Developed a fully automated in-vivo MRI classifier for neurofibrillary tangles using stacked ensemble models (SVM, XGBoost, Random Forest) and engineered multimodal imaging features derived from brain structure, tissue shape, and imaging signals.
Boosted diagnostic accuracy (AUC = 0.84) by 10% over existing clinical methods, enhancing reliable, non-invasive identification of Alzheimer's pathology.
Validated classifier on multi-center MRI datasets (N=1000+), achieving high robustness (ICC > 0.99) and reproducibility across GE, Siemens, and Philips 3T scanners.
Established clinical utility by showing biomarker scores predict a 25% faster cognitive decline, enabling earlier intervention in patient care.
Packaged into a portable Singularity container for scalable, preprocessing-free deployment on Linux/Mac and HPC systems.

2. Plasma NfL as a Predictive Biomarker for Neurodegeneration and Vascular Brain Injury

Engineered a translational biomarker analysis framework integrating plasma Neurofilament Light (NfL), white matter hyperintensities (WMH), and whole-brain morphometry to characterize neurodegeneration in 200+ aging individuals.
Revealed WMH mediate link between plasma NfL and brain atrophy, uncovering vascular pathology as a critical mechanismadvancing biomarker-driven approaches for early-stage neurodegenerative disease detection.

3. Imaging Biomarkers of Alzheimer’s and LATE Pathology

Constructed a high-throughput MRI analysis pipeline and voxel-wise linear regression framework to identify structural brain changes in 912 autopsy-confirmed cases.
Discovered LATE, a common age-related condition, causes 2x more hippocampal shrinkage than Alzheimer’s, challenging existing diagnostic assumptions.
Demonstrated individuals with both Alzheimer’s and LATE show over 3x more widespread brain atrophy, improving understanding of disease progression.

4. Neuropathological Signatures in Subcortical Brain Structures

Built an automated MRI analysis pipeline to extract volumetric and shape metrics from deep gray matter structures in 842 autopsy-confirmed cases, uncovering distinct signatures of 8 different brain diseases.
Applied non-parametric permutation-based linear regression to identify independent effects of brain pathologies on structure, controlling for key demographic and clinicopathologic factors to ensure clinically meaningful insights.
Supported creation of clinically relevant imaging biomarkers for early disease detection and patient stratification in neurodegenerative disease trials.

Projects

1. Sentence Classification Using Densely Connected Bi-LSTM

Implemented and benchmarked a deep NLP classifier using Densely Connected Bidirectional LSTM (DC-Bi-LSTM) for sentence classification across 7 benchmark datasets.
Developed comparative models including Bi-LSTM and Deep Stacked Bi-LSTM (DS-Bi-LSTM) to evaluate gradient issues and scalability.
Integrated 300-dimensional GloVe embeddings and designed robust data preprocessing and tokenization pipelines.
Led model evaluation and performance visualization, showcasing superior generalization across diverse linguistic tasks.

2. 3D CNN-Based Deep Learning for Brain MRI Segmentation

Designed 3D U-Net and ResUNet models to automate segmentation of subcortical brain regions from Quantitative Susceptibility Mapping (QSM) MRI, replacing manual annotation.
Built end-to-end pipeline including data normalization, atlas registration (ANTs), patch extraction, and label preprocessing for 100+ subjects.
Tuned hyperparameters and optimized model performance using Dice coefficient and Hausdorff distance; achieved Dice scores up to 0.87.

Journal Papers

Saifullah, Khalid et al. “Neuropathological Correlates of Volume and Shape of Deep Gray Matter Structures in Community-Based Older Adults.” Human brain mapping vol. 46,10 (2025): e70273. doi:10.1002/hbm.70273

Conference Papers

Khalid Saifullah et al. Bridgeless AC-DC Buck-Boost Converter with Switched Capacitor for Low Power Applications. TENCON 2017 - 2017 IEEE Region 10 Conference, pp. 1761-1765. 5-8 Nov, 2017, Penang, Malaysia. Doi: 10.1109/TENCON.2017.8228143
Md. Abdullah Al Hysam, Md. Zihad Ul Haque, Khalid Saifullah et al. New Topologies of Cuk PFC Converter with Switched Capacitor for Low Power Applications. IEEE Region 10 Humanitarian Technology Conference (R10-HTC), pp. 620-623. 21-23 Dec, 2017, Dhaka, Bangladesh. Doi: 10.1109/R10-HTC.2017.8289036

Poster Presentation

Saifullah, K., et al. Plasma Neurofilament Light is associated with less brain tissue, larger ventricles, and higher white matter hyperintensities volume. Presented at the 2025 ISMRM & ISMRT Annual Meeting & Exhibition, May 10–15, 2025, Honolulu, Hawaiʻi, USA.
Saifullah, K., et al. Plasma Neurofilament Light Links to Brain Tissue Loss, Ventricular Enlargement, and Elevated White Matter Hyperintensities. Presented at the Alzheimer’s Association International Conference (AAIC) Annual Meeting, July 27–31, 2025, Toronto, Canada.
Saifullah, K., et al. Plasma Neurofilament Light Reflects White Matter Hyperintensity-Driven Neurodegeneration in Aging. Presented at the Organization for Human Brain Mapping (OHBM) Annual Meeting, June 24–28, 2025, Brisbane, Australia.
Saifullah, K., et al. Plasma Neurofilament Light Reflects Brain Tissue Loss, Ventricular Enlargement, and Vascular-Mediated Neurodegeneration. Will be presented at the Radiological Society of North America (RSNA) 111th Scientific Assembly and Annual Meeting, November 30 – December 4, 2025, Chicago, Illinois, USA.
Chowdhury, G. M., Saifullah, K., et al. Elevated ARTS Scores Correlate with Increased NfL Levels but Not p-Tau217. Presented at the Organization for Human Brain Mapping (OHBM) Annual Meeting, June 24–28, 2025, Brisbane, Australia.
Chowdhury, G. M., Saifullah, K., et al. Higher ARTS Score Is Associated with Higher Levels of NfL but Not with p-Tau217. Will be presented at the Radiological Society of North America (RSNA) 111th Scientific Assembly and Annual Meeting, November 30 – December 4, 2025, Chicago, Illinois, USA.
Khalid Saifullah et al. Brain atrophy in Alzheimer’s and LATE neuropathology. Poster Presentation at Radiological Society of North America (RSNA), 109th scientific assembly and annual meeting. 1-5 Dec, 2024, Chicago, Illinois, USA
Saifullah, K., et al. Spatial pattern of brain atrophy in Alzheimer’s and LATE neuropathology. Poster Presentation at Organization for Human Brain Mapping (OHBM) annual meeting. 23-27 June, 2024, Seoul, Korea
Saifullah, K., Nazanin Makkinejad et al. Associations of subcortical shapes with age-related neuropathologies in community-based older adults. Poster Presentation at Alzheimer’s Association International Conference (AAIC) annual meeting. 16-20 July, 2023, Amsterdam, Netherlands
Saifullah, K., Makkinejad, N et al. Subcortical shapes and age-related neuropathologies. Poster Presentation at Organization for Human Brain Mapping (OHBM) annual meeting. 22-26 July, 2023, Montréal, Canada
Saifullah, K., Makkinejad, N et al. Associations of the shape of subcortical brain structures with age-related neuropathologist in community-based older adults. Poster Presentation at ISMRM & ISMRT Annual Meeting & Exhibition. 3-8 June 2023, Toronto, ON, Canada
Saifullah, K., et al. MRI-based Neurofibrillary Tangles Prediction. Poster Presentation at Organization for Human Brain Mapping (OHBM) annual meeting. 19-23 June, 2022, Glasgow, Scotland

Oral Presentation

Khalid Saifullah et al. Investigating spatial characteristics of brain atrophy in Alzheimer’s and LATE neuropathology. Oral Presentation at Alzheimer’s Association International Conference (AAIC) annual meeting. 28 July-1 Aug, 2024, Philadelphia, Pennsylvania, USA
Khalid Saifullah et al. Difference in the spatial pattern of brain atrophy associated with Alzheimer's and LATE neuropathology. Oral Presentation at International Society for Magnetic Resonance in Medicine. 4-9 May, 2024, Singapore
Khalid Saifullah et al. Towards Classification of Neurofibrillary Tangles with MRI Based Biomarker. Oral Presentation at Radiological Society of North America (RSNA), 108th scientific assembly and annual meeting. 27 Nov-1 Dec, 2022, Chicago, Illinois, USA
Khalid Saifullah et al. Neurofibrillary Tangles Prediction Based On MRI. Oral Presentation at Alzheimer’s Association International Conference (AAIC) annual meeting. 31 July-4 Aug, 2022, San Diego, California, USA
Khalid Saifullah et al. Towards an MRI-based Prediction of Neurofibrillary Tangles. Oral Presentation at International Society for Magnetic Resonance in Medicine, ISMRM-ESMRMB & ISMRT joint annual meeting. 07-12 May, 2022, London, England, UK

Awards and Achievements

ISMRM Magna Cum Laude Merit Award 2024 - Awarded to the top 15% of presenters
AAIC Conference Fellowship Award - Awarded full conference registration and four nights of hotel stay for Alzheimer’s Association International Conference (AAIC) 2024
International Society for Magnetic Resonance in Medicine (ISMRM) Trainee (Educational) Stipend - Awarded for the maximum of three consecutive years ($825 in 2024, $575 in 2023, $665 in 2022)
Graduate Research Assistantship at Medical Imaging Research Center (MIRC) in Magnetic Resonance Imaging at Illinois Institute of Technology (MRIIT) Lab
Champion in Project Showcasing in Cennovation 2015 – Inter University National Civil Fest “Matlab and Arduino Geo-Tech Device to Prevent Failure of Retention Wall”, Bangladesh
2nd Runners-Up in Poster Presentation in Esonance 2017: Inter University National Electrical Fest - “Modeling and Control of a New Grid Connected PV System”, Bangladesh
Awarded OIC scholarship of USD $9000 for the 4 year Undergraduate Program for outstanding result in entrance exam of Islamic University of Technology (IUT), Bangladesh
Daily Star Award - outstanding performance in O'Level & A'Level
17th Position in National Physics Olympiad (NPhO) in 2011, Dhaka Regional Round, Bangladesh
Top 10 public speakers, 2009, High School Category, National Debate Federation (NDF), Bangladesh

Technical Skills

Programming Languages: Python, R, MATLAB, Bash scripting, Perl, Linux/UNIX, C, C++
Deep Learning Frameworks: TensorFlow, Keras, PyTorch
Machine Learning & AI: Neural Networks (CNN, RNN), Transformers, Natural Language Processing (NLP), Generative Adversarial Networks (GANs), Autoencoders, Dimensionality Reduction (PCA), Support Vector Machines (SVM), Ensemble Methods (XGBoost, Random Forest), Supervised & Unsupervised Learning, Clustering (K-Means, DBSCAN), A/B Testing, Hyperparameter Tuning, Cross-Validation
Medical Imaging: DICOM, NIfTI, NRRD, ITK-SNAP, FSL, ANTs, SPM
Statistical Analysis: Linear & Logistic Regression, Mixed-Effects Models, Hypothesis Testing, T-tests, ANOVA (Analysis of Variance), PALM (Permutation Analysis of Linear Models), Partial F-tests
Data Processing & Analysis: OpenCV, NumPy, pandas, scikit-learn, SciPy, nibabel, nilearn
Database Management: SQL, MongoDB (NoSQL), PostgreSQL
Cloud & High-Performance Computing (HPC): AWS (EC2, S3, ECR), Google Cloud Platform (GCP), Microsoft Azure, Cluster Computing, Parallel & Distributed Processing, Job Scheduling (SLURM), Scalable Deployment & Virtualization
Software Development & DevOps: CI/CD Pipelines, Docker, Kubernetes, Agile/Scrum Methodologies, Git, GitHub, Flask
Expertise: Communication, Teamwork, Problem Solving, Time Management, Creativity, Leadership, Scientific Writing

Services and Workshops

Lead Organizer, Hackathon & Robotics Workshop - MUNA, USA Jan 2023 - Present
- - Led an annual Python Hackathon and Robotics Workshop for 100 high school students, introducing programming concepts.
  - Designed 5 educational modules, resulting in 100% completion rate among participants with no prior coding experience.
  - Coordinated event logistics and mentored students, improving problem-solving skills by 20% through hands-on activities.
Coordinator, Engineering Club - Eastern University, Bangladesh Sept 2018 – Aug 2020
- - Led 12+ workshops on machine learning, AI, & data science, engaging 100+ students including healthcare applications.
  - Supervised 8 student teams, leading to top 3 finishes in 3 inter-university competitions with AI focused projects.
Robotics Team Mentor - Islamic University of Technology, Bangladesh Jan 2015 – Nov 2017
- - Led a 15+ member robotics team to top 3 finishes in 5 national and 2 international competitions
  - Organized 8+ workshops on robotics and programming, improving team skills and performance

Page updated

Google Sites

Report abuse