Hi, I'm Vedant Malik

Driven high school senior researcher focused on AI-driven healthcare, genetic mutation analysis, and equitable STEM education. Recognized as a Top 300 Broadcom MASTERS awardee and multiple 1st Place winner at the Lehigh Valley and Delaware Valley Science Fairs.

I am actively researching in AI-driven healthcare and genetic mutation analysis: From detecting subdural hematoma disease progression and misfolded proteins to uncovering microbiome-based mechanisms of neuroinflammation.

View Research GitHub
Vedant Malik
🌟

Broadcom MASTERS

Top 300 in the U.S.

🥇

Gold Medalist

Delaware Valley Science Fair

My Fine-Tuned LLMs

Specialized language models fine-tuned on medical and biological datasets, hosted on Hugging Face.

🤗

Llama 3.2
11B Vision

Radiology Reports

View Model
🧬

Qwen 2.5
VL 7B

Genomic Analysis

View Model
🩺

Llama 3.1
8B

Patient Assistant

View Model
View all models on Hugging Face

Research & Project Demos

Selected works from international medical imaging conferences and competitive science research.

Conference Presentations & Proceedings

RSNA Logo

Cross-Modal Discrepancy

RSNA 2025 · Chicago, IL

From Disagreement to Diagnosis: advanced cross-modal discrepancy detection between radiology images and reports, using attention mechanisms to flag unsafe or inconsistent AI outputs.

View Research
SIIM Logo

RadAlign

SIIM 2025 · Portland, OR

Enhancing diagnostic confidence and safety while minimizing hallucinations in AI-generated radiology reports using a Vision-LLM system designed specifically for medical imaging.

View Research
EuSoMII Logo

Vision-LLM Medical Imaging

EuSoMII 2025 · Crete, Greece

Beyond Accuracy – fine-tuning Vision LLMs with attention-guided reasoning for clinically trustworthy medical image summaries that radiologists can actually rely on.

View Research
AACR Logo

Spine Cancer AI Analysis

AACR 2025 · Baltimore, MD

Poster presentation on attention-based chain-of-thought reasoning for automated spine cancer image analysis. Abstract published in Cancer Epidemiology, Biomarkers & Prevention.

View Research
CAIMI Logo

Vector Contours for Vision-LLM

CAIMI 2025 · Irvine, CA

Vector Contours to the Rescue: using simple shape outlines to keep Vision-LLM generated radiology reports clinically grounded and less prone to hallucination.

View Research

Competitive Research & Science Fairs

Microbiome-Neuroinflammation

2025

🥇 First Place, LVSF; 🥈 Second Place, DVSF

A Novel Microbiome-Based Mechanism to Neuroinflammation: Connecting Gut Microbial Dynamics with Neuro Inflammatory Mechanisms in Brain Disorders

View Research Watch Demo

MyeloMut

2024

🥇 First Place, LVSF; 🥈 Second Place, DVSF

A Novel Biocomputational approach for Identifying common patterns in Genetic Mutations causing Multiple Myeloma.

View Research

RTRad

2023

🥇 First Place, LVSF; 🥇 First Place, DVSF

A Novel AI-Based system to detect "Subdural Hematoma Disease Progression" that adapts to changing image data

View Research

ChaperDetect

2022

🥇 First Place & Gold Medalist, DVSF
🥇 First Place, LVSF

Detecting Misfolded Proteins That Escape Chaperones Causing Degenerative Disease Using AI-Based Deep Learning

🌟 Broadcom MASTERS Top 300
View Research

Early Research

2021

🎖️ Honorable Mention, LVSF

Early scientific research entry recognized at the Lehigh Valley Science & Engineering Research Fair.

View Research

Research Awards Recognition

Parenteral Drug Association Award

2022

Delaware Valley Science Fair (DVSF)

Awarded for excellence in health-related research and its potential impact on public health and therapeutic development.

Research Focus & Activities

AI-Driven Healthcare & Medical Imaging

SIIM, EuSoMII, AACR, RSNA & CAIMI 2025

I work on trustworthy, clinically grounded AI for radiology and medical imaging. Recent projects include: RadAlign (SIIM 2025) for enhancing diagnostic confidence and safety while minimizing hallucinations in AI-generated radiology reports; attention-guided fine-tuning of vision LLMs for clinically trustworthy medical image summaries (EuSoMII 2025); cross-modal discrepancy detection between images and reports in radiology (RSNA 2025); and vector contour–based methods to keep Vision-LLM generated reports clinically grounded (CAIMI 2025).

Genetic Mutations & Protein Misfolding

MyeloMut & ChaperDetect

MyeloMut is a novel biocomputational approach for identifying common patterns in genetic mutations causing multiple myeloma. It earned 1st Place at LVSF 2024 and 2nd Place at DVSF 2024. Earlier, I developed ChaperDetect, an AI-based deep learning system to detect misfolded proteins that escape chaperones and contribute to degenerative disease. This work received a Gold Medal and 1st Place at DVSF 2022 and 1st Place at LVSF 2022.

SAT Tutor

Schoolhouse SAT Tutoring

Tutored multiple students across 3 continents (Africa, Europe, North America) for the SAT via Schoolhouse. Administered live examinations and saw score improvements over the course of multiple weeks.

Community Volunteering

Mentorship & Education

Coached and mentored children in India during summer and winter vacations, focusing on the basics of math, science, and English. Committed to making educational resources accessible and supporting students in building strong academic foundations.