Advancing Medical AI,
One Discovery at a Time
I'm a student researcher passionate about leveraging artificial intelligence and computational methods to solve critical healthcare challenges. I currently conduct research in medical imaging AI and radiology under faculty guidance at the University of Pennsylvania's Perelman School of Medicine. My work spans cardiac biomarker discovery, AI-enhanced radiology, and drug discovery.
Research Impact & Focus
My research spans multiple domains in healthcare technology, with a particular focus on medical imaging AI, computational drug discovery, and precision medicine. I've developed novel approaches to enhance diagnostic accuracy, reduce AI hallucinations in radiology reports, and identify biomarkers for cardiovascular diseases.
Each project addresses real clinical challenges, with work presented at prestigious conferences including RSNA, SIIM, AACR, EuSoMII, and published in peer-reviewed journals. My goal is to create technology that directly improves patient outcomes and clinical decision-making.
Research Experience
Student Researcher
Working on developing safer, more reliable vision LLMs for medical imaging and diagnostic support.
Awards & Recognition
Research & Academic Excellence
DistinguishedResearch Projects, Conference Presentations & Publications
Conference Abbreviations:
LVSF = Lehigh Valley Science Fair |
DVSF = Delaware Valley Science Fair |
PJAS = Pennsylvania Junior Academy of Science
RSNA = Radiological Society of North America |
SIIM = Society for Imaging Informatics in Medicine |
AACR = American Association for Cancer Research
EuSoMII = European Society of Medical Imaging Informatics |
PRS = Pennsylvania Radiological Society |
CAIMI = Conference on AI in Medical Imaging
Which Bridge Design Can Support the Most Weight
Researched and investigated load-bearing capacities of various bridge designs using principles of physics and engineering. Built various miniature models to evaluate compression, tension, and torsional forces.
2021Novel MI Diagnosing System for SARS-CoV-2 Cardiac Complications
Designed an in-silico diagnostic system for detecting myocardial infarction in COVID-19 patients with cardiac asthma, improving early intervention.
2022rnaNASH: RNA-Based Drug Discovery for NASH Disease
Developed an in-silico computational approach for RNA-based therapeutics targeting Non-Alcoholic Steatohepatitis (NASH) with improved efficacy and safety profiles.
2023Discovering Novel Biomarkers in Cardiac Stent Thrombosis and Restenosis
Conducted precision intravascular imaging to identify novel biomarkers that predict post-stent complications, with potential to transform cardiovascular treatment monitoring.
2024Detecting Antibiotic Resistance in Mycobacterium Tuberculosis
Created a novel diagnostic method for rapidly identifying antibiotic-resistant TB strains, addressing a critical global health challenge.
2025Image to Insight: Vision–Language Tokens Aligned with Radiologist Reasoning
Scientific Exhibit selected for presentation at the Pennsylvania Radiological Society 110th Annual Meeting, focusing on aligning AI-generated insights with clinical radiologist reasoning.
2025RadAlign – Enhancing Diagnostic Confidence in AI-Generated Radiology Reports
Developed methods to enhance diagnostic confidence and safety while minimizing hallucinations in AI-generated radiology reports. Won Best Student Paper Award at SIIM 2025.
2025Fine-Tuning Vision-LLM with Attention-Guided Reasoning
Beyond accuracy: developed attention-guided reasoning techniques for fine-tuning Vision-LLMs to produce clinically trustworthy medical image summaries.
2025From Pixels to Prognosis: Automated Spine Cancer Image Analysis
Developed attention-based chain of thought reasoning for automated spine cancer image analysis. Published in Cancer Epidemiology, Biomarkers & Prevention (AACR CEBP).
2025Cross-Modal Discrepancy Detection in Radiology
From disagreement to diagnosis: developed methods for detecting and resolving cross-modal discrepancies in radiology imaging and reports.
2025Vector Contours for Clinically Grounded Vision-LLM Reports
Novel approach using basic shape outlines (vector contours) to keep Vision-LLM generated radiology reports clinically grounded and accurate.
2025Community Service & Leadership
Community Service & Tutoring
National Honor Society Tutor – Mentoring peers in Math, Science, and English
Remote Tutor – Volunteer as a hybrid tutor supporting underprivileged children in Delhi, initially providing online tutoring during the COVID-19 pandemic and continuing to coordinate across time zones to run ongoing sessions coaching them in math, science and english
Athletics & Extracurriculars
Well-Rounded- ★Parkland High School – Indoor Track Team (2023)
- ★SPYA Soccer (2017–2020) – South Parkland Youth Association
- ★SPYA Basketball (2017–2023) – South Parkland Youth Association
- ★Orchestra Achievement Award (2018-2020) – Excellence in performance and commitment
- ★Scholastic Scrimmage – Participant in academic quiz competitions
Media & Highlights
Photos & Awards
GalleryVisual highlights from conferences, awards ceremonies, and research presentations.
📬 Let's Connect
GitHub: github.com/krishm-12 | LinkedIn: linkedin.com/in/krishm-parkland