Research

Behavioral Neuroscience Research at CSUB

Cognitive Compass is dedicated to investigating the consequences of mental distress to provide aid to those in need.

At Stockdale High School, Ishmeet saw firsthand how classmates turned to vaping for stress relief or escape, and he recognized how easily mental health challenges can fuel addiction. Determined to take action, he joined the Gancarz-Kausch Neuroscience Laboratory at California State University, Bakersfield, in 2023.

In Dr. Gancarz’s behavioral neuroscience lab, Ishmeet studied the neurobiological mechanisms underlying sucrose and drug addiction, working with rodent behavioral models to uncover how different experimental conditions shape addictive behaviors. He works with his research team to uncover novel patterns in addiction and reward pathways.

In 2024, Ishmeet co-authored and presented an abstract from this research at the Society for Neuroscience “Neuroscience 2024” conference (the world's largest annual neuroscience research conference), where he contributed to discussions advancing the broader field of behavioral neuroscience.

Computational Stroke Research with Ohio State University

Cognitive Compass is committed to leveraging technology, particularly artificial intelligence and computer science, to advance solutions for stroke, a condition caused by disrupted blood flow to the brain.

As Ishmeet expanded his expertise in artificial intelligence and Python programming, he reached out nationally for research opportunities and connected with Dr. Deepak Gulati at The Ohio State University Department of Neurology. Together, they collaborated on two projects that they presented at the International Stroke Conference:

Autonomous Stroke Scale Score Using Artificial Intelligence/Machine Learning (poster presented by Ishmeet)
Automated Detection of Facial Droop: ML/AI-Based Tool for Early Detection of Stroke (published in Stroke journal)

Independent Computational/AI Neuroradiology Research

Cognitive Compass prioritizes improving neurological healthcare, especially by leveraging the fascinating disciplines of artificial intelligence and computer science.

Cognitive Compass is dedicated to advancing neurological healthcare by applying artificial intelligence and computational methods to support clinical decision-making.

While shadowing at a local hospital, Ishmeet observed a medical imaging technician identify a missed brain tumor on an MRI. Because of policy constraints, the technician could not report the error, underscoring the human limits of radiology. Motivated to address this problem, he applied his skills in Python programming to develop a convolutional neural network capable of detecting brain tumors.

The first model initially achieved only 13 percent accuracy, but through methods such as data augmentation and L1/L2 regularization, performance improved to 94 percent across 44 tumor classes. This work was recognized at the Kern County and California State Science and Engineering Fairs and selected as a finalist project at the Regeneron International Science and Engineering Fair (ISEF).

The following year, the research advanced to 99 percent accuracy and incorporated solutions for motion artifacts, a critical barrier to real-world robustness. Explainable AI techniques were also implemented to allow physicians to visually trace the model’s diagnostic process. This project received the Regeneron Biomedical Sciences Award and qualified for ISEF, where it earned a 3rd Place Grand Award. Ishmeet also shared this work at the 2025 IEEE MIT Undergraduate Research Technology Conference and is sole author of the corresponding technical research paper published in IEEE Xplore. (IEEE is the world's leading publisher of engineering and computing research.)

Garcia Summer Scholars Research (2025)

Cognitive Compass advances healthcare by exploring how emerging materials and technologies can improve neurological treatments. At the Garcia Summer Scholars Program, Ishmeet contributed to the development of a novel shape-memory polymer (SMP) designed to make brain aneurysm procedures safer and more effective. Brain aneurysms carry a high risk of rupture, and one treatment approach involves injecting a biodegradable polymer into the aneurysm sac. A major challenge is that the polymer often adheres to the catheter tip, creating the risk of vessel damage when the catheter is withdrawn. To address this, the team synthesized a biocompatible SMP that can change shape at body temperature, enabling clean mechanical severance of the polymer from the catheter.

Ishmeet’s work focused on the design, synthesis, and testing of the SMP network. Using spectroscopy to confirm chemical composition and advanced analyses to study crystallinity and thermal behavior, the project demonstrated high efficiency and promising recovery rates in preliminary trials. Moving forward, the research aims to optimize the polymer’s performance, validate its biocompatibility, and integrate 3D-printed catheter-tip constructs enhanced with temperature-triggered activation.

This project highlights how advances in materials science can directly improve treatments for brain disorders, reducing the risks of aneurysm procedures and protecting patients’ cognitive health. By linking medical innovation with neuroscience, it reflects the core mission of Cognitive Compass: using science and technology to safeguard the brain and support healthier lives.

Computational Cognitive Systems Research at UCSC SIP (2024)

Cognitive Compass extends its mission of advancing mental and cognitive health into prevention-focused applications, including improving safety in autonomous transportation. Understanding pedestrian behavior in vehicle interactions is essential to lowering the risk of accidents, particularly in edge cases where human cognition and machine perception intersect.

At the University of California, Santa Cruz Summer Internship Program (SIP), Ishmeet worked on addressing the limitations of existing pedestrian datasets, which often suffer from inconsistent annotation methods and inadequate temporal modeling. Such shortcomings limit the ability of autonomous systems to anticipate and respond to complex human behavior.

To tackle this problem, Ishmeet co-developed PedAnalyze, a Python-based annotation platform that formalizes pedestrian–vehicle interactions into structured, queryable datasets. The tool integrates both frame-level and sequence-level annotation, reducing redundancy while capturing temporal dependencies critical for modeling human decision-making in traffic scenarios. Beyond annotation, Ishmeet curated and systematized rare but high-impact events (such as jaywalking, hesitation before crossing, or atypical avoidance maneuvers) from dashcam footage. These edge cases were integrated into an ontology of pedestrian behavior designed to enrich training corpora for autonomous vehicle research.

Through this work, Ishmeet contributed to building one of the few frameworks capable of systematically encoding nuanced human behaviors into computationally tractable datasets. This research not only strengthens the testing and development of autonomous driving systems but also reflects Cognitive Compass’s broader mission: applying cognitive and computational science to prevent harm before it occurs.

Collaborative Medical Research

1) Calcium and Muscle Contraction
Ishmeet co-authored a review article in the International Journal of Research in Medical Sciences on the role of calcium in muscle contraction and fatigue, highlighting how calcium dynamics differ in slow-twitch and fast-twitch fibers and their impact on performance and endurance.

2) Adolescent Mental Health and COVID-19
Ishmeet is collaborating on a forthcoming manuscript examining the impact of COVID-19 on adolescent mental health, synthesizing research on youth well-being, access to care, and long-term outcomes.