NSRI Research Groups

Find the work.
Join the people.

Only active groups with open student seats appear here. Every listed team meets on a recurring schedule. Start your own group or join work already in motion.Explore active groups

Start here

Create your own group

Choose independent research or select a PhD first. Student teams stop at 3.

Unlock group creation

Artificial Intelligence, Machine Learning & Data Science

Self-Contamination in Continuously Updated RAG Systems

1 seat open

Public Health

Digital Health literacy

2 seats open

PhD advisory

Build with a PhD

Choose an advisor in NSRI's existing order, open their profile, and start under 1 of their research directions.

Explore advisors

Loading active groups

Research groups in focus

See the people and the weekly plan.

01 / 02

Artificial Intelligence, Machine Learning & Data ScienceRecruiting

Self-Contamination in Continuously Updated RAG Systems

This project investigates self-contamination in continuously updated Retrieval-Augmented Generation (RAG) systems: the phenomenon where a RAG system's own previously generated outputs are re-ingested into its knowledge base over time, progressively degrading retrieval quality and accelerating hallucination. Grounded in production-scale infrastructure (Qdrant vector store, Ollama-served embeddings, and the bge-m3 embedding model), the work formalizes the contamination dynamics with formal math and benchmarks retrieval degradation across successive update cycles. It builds on and extends recent literature on model collapse, data-poisoning attacks on RAG (e.g., PoisonedRAG), information-cascade effects (Spiral of Silence), and hallucination detection, aiming to propose practical mitigation strategies for production RAG deployments.

Lead: Moiz Sheraz
PhD: Muhammad Tayyab Chaudhry, PhD
Weekly: Saturday · 12:00 PM PT
Team: 1/2 · 1 open

View brief

Public HealthRecruiting

Digital Health literacy

This research aims to assess the level of digital health literacy among students and the general population in Pakistan. It will explore how people access, understand, and use online health information and digital health tools. The study will also identify challenges and propose recommendations to improve digital health awareness and responsible use of health technology.

Lead: Salman Khan
PhD: Independent research
Weekly: Sunday · 21:00 Pakistan
Team: 1/3 · 2 open

View brief

Past and ongoing research

Explore the work already underway.

22 groupsThese groups are no longer accepting applications. Browse ongoing, completed, closed, and archived NSRI projects.

Health and Life SciencesApplications Closed

Intermediate Group (The Scrubs Society Team)

This intermediate research group introduces participants to the foundations of interdisciplinary research through accessible topics in healthcare. Students will explore real-world health questions while learning essential research skills such as developing research questions, reviewing literature, collecting and interpreting data, and communicating findings. As the group progresses, projects will gradually connect healthcare with fields such as psychology, technology, engineering, data science, and design, preparing participants to take on more independent and interdisciplinary research in the future.

Led by Gursharnpreet Kaur HannsView project

Health and Life SciencesApplications Closed

Research Group 1 for Beginners (The Scrubs Society Team)

This beginner research group introduces participants to the foundations of interdisciplinary research through accessible topics in healthcare. Students will explore real-world health questions while learning essential research skills such as developing research questions, reviewing literature, collecting and interpreting data, and communicating findings. As the group progresses, projects will gradually connect healthcare with fields such as psychology, technology, engineering, data science, and design, preparing participants to take on more independent and interdisciplinary research in the future.

Led by Gursharnpreet Kaur HannsView project

Health and Life SciencesApplications Closed

Induced Pluripotent Stem Cells (iPSCs) in Heart Repair

This project investigates how induced pluripotent stem cells (iPSCs) can be used to regenerate damaged heart tissue and advance future therapies for heart disease.

Led by Marvarit AbdurakhmanovaView project

Health and Life SciencesApplications Closed

Alissa Afrah Monir

Led by Alissa MonirView project

Engineering and TechnologyApplications Closed

Mathematics - Root Finding, Detection and Classification of Multiplicities

Building on my previous work on the classification and isolation of real roots in nonlinear equations, I am now exploring the next stage of the problem: root multiplicity classification. The goal is to investigate new mathematical and computational methods for determining and classifying the multiplicity of isolated roots, beyond conventional derivative-based approaches. I am looking to collaborate with researchers and students interested in numerical analysis, computational mathematics, nonlinear equations, and scientific computing. If this area aligns with your research interests, I would be happy to connect and discuss potential collaborations. The main research question is : Can root multiplicity be classified from the local behavior of a function without relying solely on higher-order derivatives?

Led by Himesh NeupaneView project

Engineering and TechnologyApplications Closed

Adaptive Geometric Intelligence & Learning for Embodied Systems (AGILE) Lab

AGILE Lab is an advanced interdisciplinary research group focused on developing adaptive geometric intelligence for autonomous embodied systems. The team investigates how geometry, morphology, and decentralized communication can enable scalable collective intelligence in robotic systems. By integrating swarm robotics, computational geometry, artificial intelligence, bio-inspired engineering, and simulation-based modeling, AGILE Lab develops novel algorithms and robotic architectures that allow autonomous agents to perceive, reason, adapt, and collaborate in complex environments. Current research focuses on geometry-induced collective inference, multi-agent coordination, autonomous reconstruction, and efficient robotic systems inspired by biological collective behaviors.

Led by Alissa MonirView project

Computer Science, AI, and Data ScienceFull

Prima Domini

Prima Domini is the principal research patronage group of CuteHeart Exosystems Ltd., dedicated to advancing deterministic execution theory, ecological computing, simulation topology, and secure protocol design. The group focuses on reducing the ecological impact of modern technology, improving safety and efficiency in aging and simulation‑based systems, and addressing extreme environmental computing conditions to enhance security and reliability. Research within Prima Domini is conducted on the CuteHeart platform using an online IDE that supports Python, Lua, and visual scripting. Contributors participate in weekly research cycles and work within deterministic and semi‑deterministic execution environments to evaluate system behavior, protocol correctness, ecological telemetry, and execution‑path stability. The group welcomes students and early‑career researchers, offering a structured environment for those interested in systems research, execution theory, and protocol design. Prima Domini maintains several core exploration tracks. These include deterministic execution, latent execution mapping, signal‑process trajectory recovery, electromagnetic hardening for computational workloads, stochastic and systolic execution modeling, protocolization and atomic execution, and manifold topology with tessellation for simulation spaces. Each track is designed to help researchers understand how execution pathways, system topology, and protocol structure influence output, safety, and ecological impact. The group operates as a federated research collective. Work progresses through a structured cycle consisting of preliminary analysis, interlocutory mid‑stage analysis, and thesis‑oriented pre‑ and post‑analysis. This methodology encourages iterative refinement, reproducibility, and clarity as execution models evolve over time.

Led by Akko SinnView project

Health and Life SciencesFull

2026 ADA Standards of Care in Diabetes

Led by Vincent YuanView project

Health and Life SciencesApplications Closed

Diabetic Retinopathy Detection Using Deep Learning

We're building a small research team to study how deep learning can help detect and grade diabetic retinopathy from eye images. The goal is to compare a few existing model architectures, understand what makes them work (or fail), and write this up for publication in the NSRI journal. No prior machine learning experience needed — we'll learn the basics together as a team. Looking for 2 MBBS students (clinical insight, literature review, write-up) and 2 tech students (data work, model training) who are curious and willing to commit weekly.

Led by Asad ChannaView project

Computational BiologyApplications Closed

Hippocampus-to-Neocortex Transfer in a Dual-Region Spiking Neural Network

This project explores how replayed memories are consolidated into long-term knowledge in biologically inspired spiking neural networks. We will develop a hippocampus–neocortex model to investigate systems consolidation, schema formation, and memory transfer using computational neuroscience methods.

Led by Ashwajit WarwatkarView project

Health and Life SciencesFull

A Scientometric Audit of the Reference Base of the 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure

Led by Vincent YuanView project

Health and Life SciencesApplications Closed

A Scientometric Audit of the Reference Base of the 2024 ESC Guidelines for the Management of Elevated Blood Pressure and Hypertension

Led by Vincent YuanView project

Health and Life SciencesFull

Creatine Supplementation in Adolescents and Youth: An Updated Systemic Review

We are conducting a systematic review examining the effects of creatine supplementation in adolescent and youth populations. Despite creatine being one of the most widely used dietary supplements among young people, the evidence base for this age group remains undercharacterized relative to adult literature. This review aims to fill that gap by comprehensively synthesizing available data across multiple outcome domains. We are recruiting motivated individuals to join the review team. Roles may include title/abstract screening, full-text review, data extraction, risk of bias assessment, and manuscript writing. Prior experience with systematic reviews and coding is helpful

Led by Vincent YuanView project

Health and Life SciencesApplications Closed

Advanced Wound Healing and Regenerative Materials Research Group

A multidisciplinary research group focused on advanced biomaterials, nanotechnology, medicinal chemistry, and healthcare innovations for wound healing, antimicrobial therapies, drug delivery, and sustainable healthcare solutions.

Led by Shahroz ZafarView project

Public HealthApplications Closed

Cardiology & Medicine : Systematic review / Literature review for USMLE

Led by Alisha LakhaniView project

Biomedical EngineeringIn Progress

Connecting computational neuroscience with mechanistic interpretability

> This study investigates whether the Replay-Gated Cascade Consolidation (RGCC) mechanism extends beyond spiking neural networks to transformer-based language models. We test the hypothesis that sequential learning in transformers exhibits the same replay-driven consolidation dynamics predicted by RGCC, including competition for shared parameters and an encoding-order consolidation gradient. By analyzing sequential fine-tuning and replay strategies, we evaluate whether RGCC provides a unified mechanistic explanation for memory consolidation and catastrophic forgetting across fundamentally different neural architectures.

Led by Ashwajit WarwatkarView project

Health and Life SciencesIn Progress

Systematic Review of Evidence database of NCCN 2026 Breast Cancer Guidelines

We are conducting a systematic review of the clinical evidence database underpinning the NCCN Guidelines for Breast Cancer Version 4.2026, with the primary objective of mapping critical evidence gaps. Objective: To identify the "Resource-Evidence Gap" by analyzing whether the guideline's preferred standards are globally representative to even lower to middle income states or just limited to high income settings with advanced medical infrastructure. Only those who are genuinely interested in dedicating their time should apply. Participants will have to analyze cited studies, identify representation gaps, evaluate demographic reporting, compare guideline evidence bases, investigate research questions regarding bias in medical evidence, and handle manuscript drafting and revisions. The goal is to get this published, so each participant must commit to staying on board through the entire process, not just the initial drafting, but also the revision stages up until final publication. Our goal is to publish within 3-4 months. None of us are experts yet, so we will learn and navigate this together. But the only acceptable behaviors are mutual respect and ethical practice.

Led by Javeria AhmedView project

Health and Life SciencesIn Progress

Artificial Intelligence in Brain Cancer: Current Applications, Challenges, and Future Prospects

Brain cancer, including primary tumors such as gliomas and glioblastomas, is one of the most complex and life-threatening neurological conditions. Diagnosis and treatment are difficult due to the brain’s sensitive structure, tumor heterogeneity, and limitations in early detection. Artificial Intelligence (AI) is increasingly being integrated into brain cancer research and clinical workflows to improve detection, classification, and treatment planning. In medical imaging, particularly MRI scans, machine learning and deep learning models can assist in identifying tumors at earlier stages, segmenting tumor boundaries more accurately, and distinguishing between different tumor types. This improves diagnostic precision and reduces the burden on radiologists. AI also plays a role in predicting patient outcomes, survival rates, and treatment responses by analyzing large-scale clinical and genomic datasets. This enables more personalized treatment strategies, supporting the shift toward precision medicine in neuro-oncology. However, challenges remain, including limited high-quality annotated brain imaging datasets, variability in imaging protocols across hospitals, and concerns about model interpretability and clinical trust. Integration into real-world healthcare systems also requires strong validation and regulatory approval. Future developments are expected to focus on explainable AI models, multi-modal data integration (MRI, genomic, and clinical data), and real-time clinical decision support systems. Ultimately, AI has the potential to significantly enhance early detection, treatment planning, and patient outcomes in brain cancer care.

Led by Ammara TariqView project

Engineering and TechnologyIn Progress

Machine Learning-Based Surrogate Modeling for Electromagnetic Damping Force Prediction and Parameter Ranking

This research investigates the influence of geometric, magnetic, and operating parameters on electromagnetic damping force in vibration damping systems. A structured dataset will be constructed from published experimental studies, and machine learning-based surrogate models will be developed to predict damping force and evaluate parameter importance. Feature importance techniques will be used to identify the most influential design variables, enabling improved interpretability and faster design insights for electromagnetic damping systems.

Led by Sripaadh Jayashree KuppusamyView project

Health and Life SciencesCompleted

Research Bias in the World

Research Bias in the World is a free NSRI research program examining how bias appears inside the evidence behind major medical guidelines. Medical guidelines shape how diseases are studied, treated, and understood. But 1 question is often missed: Does the evidence behind those guidelines actually represent the patients those guidelines affect? NSRI researchers are already working on 3 major guideline-bias projects across hypertension, cancer, and depression. These projects examine which studies are cited, which countries dominate the evidence, how much of the data comes from a small number of massive studies, whether demographic information is properly reported, and which populations are missing. Through weekly Sunday sessions, participants will work directly with the host and the NSRI community to break down real guideline-bias examples, discuss evidence gaps, compare research approaches, and learn from each other as they build their own projects. This is not a passive webinar. It is a working research program for students who want to understand how research systems shape what medicine eventually treats as truth. Participants will learn how to analyze cited studies, identify representation gaps, evaluate demographic reporting, compare guideline evidence bases, and develop research questions around bias in medical evidence. Only the first 100 students will be able to join the weekly live sessions. Selected participants will receive the session link and project details after signing up. Research should not only be trusted. It should be tested. Weekly sessions will be held every Sunday at 8:00 AM PKT Pakistan, 12:00 PM KST Korea. Equivalent times: 8:30 AM IST India, 9:00 AM Bangladesh time, 11:00 AM Philippines/Singapore time, and 11:00 PM ET Saturday for the U.S. East Coast during EDT, or 10:00 PM ET Saturday during EST.

Led by Brian KangView project

Health and Life SciencesCompleted

Ferroptosis in Alzheimer's Disease: Integrating Bibliometric Analysis and Bioinformatic Insights to Identify Emerging Research Frontiers (2012–2026)

Alzheimer's disease (AD) is the leading cause of dementia worldwide and represents a major unmet clinical challenge. Emerging evidence suggests that ferroptosis, an iron-dependent form of regulated cell death characterized by lipid peroxidation and oxidative stress, plays a pivotal role in the pathogenesis of AD by linking iron dyshomeostasis, mitochondrial dysfunction, neuroinflammation, and amyloid-β and tau pathology. Despite the rapid growth of research in this area, the global knowledge structure, evolving research priorities, and molecular mechanisms underlying ferroptosis in AD remain fragmented and insufficiently integrated. This project aims to combine bibliometric and scientometric approaches with bioinformatic analyses to comprehensively map the research landscape and identify emerging molecular targets in ferroptosis-related Alzheimer's disease research from 2012 to 2026. Bibliographic data retrieved from PubMed, Scopus, and Web of Science will be analyzed using VOSviewer, CiteSpace, and Bibliometrix to evaluate publication trends, collaborative networks, co-citation patterns, and thematic evolution. To complement these analyses, ferroptosis-associated gene sets from FerrDb will be integrated with publicly available transcriptomic datasets from GEO and AMP-AD to identify differentially expressed genes, enriched biological pathways, protein-protein interaction networks, and cell type-specific ferroptosis signatures. By synthesizing quantitative research trends with molecular evidence, this study seeks to uncover knowledge gaps, prioritize candidate biomarkers and therapeutic targets, and provide actionable insights for future translational and precision medicine approaches in Alzheimer's disease.

Led by HarshvardhanView project

Computer Science, AI, and Data ScienceCompleted

Group 130: Auditing Algorithmic Bias in Critical Healthcare AI

An official NSRI research team auditing algorithmic bias, geographic disparities, and fairness gaps in multi-modal clinical machine learning models trained on large-scale electronic health records. Target Dataset Focus: MIMIC-IV and eICU Collaborative Research Databases Clinical machine learning models are rapidly shifting from retrospective research to active bedside software. However, if these models are trained on data reflecting systemic inequalities, they risk automating and magnifying biases against historically underrepresented demographic and geographic groups. This research group will conduct a systematic meta-analysis auditing multi-modal clinical AI models (predictive health informatics, critical care survival curves, and diagnostic imaging models). Our primary objective is to map "fairness gaps" - tracking how validation metrics like AUROC, sensitivity, and false-positive rates fluctuate when applied across diverse racial, socioeconomic, and global cohorts. Students in this group will gain direct exposure to medical data science challenges and solutions, learn how to evaluate algorithmic equity, and contribute to a peer-reviewed manuscript aimed at open-access health informatics journals. We are looking for dedicated students split into two primary tracks: - Data & ML Track: Comfortable with Python, data preprocessing, and understanding statistical machine learning metrics. - Public Health Track: Strong passion for public health and experience in literature synthesis and scientific manuscript drafting.

Led by Apoorva GView project

Established before this portal

129 numbered research groups

NSRI organized these groups through its previous assignment system. They remain preserved in NSRI's internal records and are not shown as current openings.