Nahid Zeinali — AI Research Scientist

Multimodal AI for biomedical discovery.

I build end-to-end machine learning pipelines that integrate text and structured biomedical data — from transformer architectures and retrieval-augmented generation systems to production-ready foundation-model evaluation.

Currently: AI Research Scientist at the California Medical Innovation Institute

Download CV

Period	Position
Feb 2025 —Present	AI Research Scientist California Medical Innovation Institute · San Diego, CA Lead end-to-end AI/ML and Generative AI systems using Python, PyTorch, deep learning, and NLP for healthcare and scientific discovery. Led development of DMP Chef — a GenAI system using GPT-4, Gemini, and Llama 3.3 to generate NIH-compliant Data Management Plans with RAG and vector databases. Led and evaluated GPT-4.1 and Llama 3.3 for NIH-compliant DMP generation using automated metrics, statistical analysis, and human-centered evaluation, with GPT achieving the highest overall user satisfaction and stronger SBERT and ROUGE performance. Led and designed reproducible prompt-engineering and structured generation workflows for automated NIH Data Management Plan creation using large language models. Led development of DMP Bridge — an open-source Python-based GenAI pipeline that converts funder-specific DMP PDFs into interoperable DMPTool JSON and RDA Common Standard maDMP JSON formats. Designed a hybrid multimodal document AI architecture using pdfplumber, Qwen2-VL, Llama, prompt engineering, and schema-constrained generation to extract text, detect section headers, classify narrative content, and generate structured metadata from complex scientific PDFs. Collaborated on FAIR-aligned data standardization pipelines for ophthalmology imaging datasets-Eye ACT- improving machine learning readiness, interoperability, and future analytics/model development. Collaborated on schema evaluation and JSON-based FAIR workflows to transform scientific posters-Poster Science-into machine-readable and interoperable research objects for AI-driven discovery and reuse Prepare technical reports, manuscripts, and grant applications with investigators
Aug 2021 —Jan 2025	Graduate Research & Teaching Assistant University of Iowa, Department of Computer Science · Iowa City, IA Designed transformer-based NLP systems — Symptom-BERT, Care-BERT, Spiritual-BERT — for classification, NER, and extraction from 5M+ clinical notes. Built end-to-end NLP/LLM pipelines covering preprocessing, feature engineering, training, validation, and statistical evaluation. Symptom-GPT: Led and developed transformer-based clinical NLP and LLM models for named entity recognition and cancer symptom detection from large-scale oncology 1M EHR clinical notes, achieving F1 scores up to 0.989 for nausea/vomiting and 0.912 for anxiety detection. Symptom-BERT: Led and developed transformer-based clinical NLP and LLM models for further pretrain and fine-tune transformer models on oncology 1M EHR data to detect 13 cancer symptom groups, achieving micro-F1 = 0.933 in internal validation and micro-F1 = 0.831 in external validation. Care-BERT: Led and built transformer-based clinical NLP models to identify care priorities and life-sustaining treatment preferences from EHR narratives, achieving F1 = 0.941 and AUC = 0.978 in internal validation. Spiritual-BERT: Led and developed a Bio Clinical BERT-based NLP system to detect under-documented spiritual care information from 3.6M+ EHR notes, achieving F1 = 0.938 in internal validation and F1 = 0.832 in external validation. Implemented GPT-based workflows for synthetic clinical note generation, supporting data augmentation and external validation. OASIS – Oncology Associated Symptoms & Individualized Strategies: Collaborated with clinicians, researchers, and technical collaborators on AI- and NLP-driven oncology symptom monitoring research using EHR data to support personalized cancer care and clinical decision support. Patient-Reported Outcomes & EHR Symptom Concordance: Collaborated and analyzed concordance between patient-reported and provider-documented cancer symptoms, identifying significant documentation gaps across multiple symptom groups. Embeddings-Augmented NLP for Symptom Detection: Collaborated and developed embeddings-enhanced NLP pipelines for extracting cancer symptom information from 900K+ clinical notes, achieving F1 scores up to 0.937 for symptom classification tasks. Cancer Symptom Prediction Using Machine Learning: Collaborated and built predictive ML models using structured and unstructured EHR data to forecast cancer symptom development, with Random Forest achieving macro-AUC = 0.755 and pain prediction AUC = 0.954. Systematic Review of ML for Cancer Symptom Prediction: Led and conducted a PRISMA-guided systematic review of 42 studies evaluating machine learning approaches for cancer symptom prediction and personalized oncology care. Web Analysis: Developed an end-to-end NLP and machine learning pipeline for large-scale web scraping, sentiment classification, clustering, and topic modeling across 3,000+ online articles. Demand Forecasting for E-Commerce Sales: Built Python-based forecasting and analytics models to predict product demand and support inventory optimization for e-commerce sales data. Taught Python and machine learning courses, guiding students to develop strong skills in programming and data science. Published over 10 papers across machine learning and data science topics.
Jun 2024 —Aug 2024	NLP Data Scientist National Cancer Institute & Frederick National Laboratory, NCATS / NIH · Maryland RARe-SOURCE™: Collaborated and developed AI-powered biomedical literature intelligence workflows using LLMs, semantic search, and NLP to support rare disease knowledge discovery and scientific insight extraction. Worked closely with biomedical researchers to convert research questions into usable AI workflows and decision-support tools.
Feb 2019 —Jul 2021	Software Engineer Khorshid Hospital · Isfahan, Iran Built and enhanced EMR, PIS, and LIS systems with clinical stakeholders, improving workflow efficiency by 35%. Developed a heart-failure symptom-tracking mobile app used by 3,000+ patients in rural Isfahan.
Dec 2016 —Jan 2019	Software Engineer Parisian Institute · Tehran, Iran Designed an EHR management dashboard that reduced report turnaround time by 68%. Trained 1,500+ clinicians on EHR adoption and workflow best practices.

Period

Position

Feb 2025 —Present

AI Research Scientist

California Medical Innovation Institute · San Diego, CA

Lead end-to-end AI/ML and Generative AI systems using Python, PyTorch, deep learning, and NLP for healthcare and scientific discovery.
Led development of DMP Chef — a GenAI system using GPT-4, Gemini, and Llama 3.3 to generate NIH-compliant Data Management Plans with RAG and vector databases.
Led and evaluated GPT-4.1 and Llama 3.3 for NIH-compliant DMP generation using automated metrics, statistical analysis, and human-centered evaluation, with GPT achieving the highest overall user satisfaction and stronger SBERT and ROUGE performance.
Led and designed reproducible prompt-engineering and structured generation workflows for automated NIH Data Management Plan creation using large language models.
Led development of DMP Bridge — an open-source Python-based GenAI pipeline that converts funder-specific DMP PDFs into interoperable DMPTool JSON and RDA Common Standard maDMP JSON formats.
Designed a hybrid multimodal document AI architecture using pdfplumber, Qwen2-VL, Llama, prompt engineering, and schema-constrained generation to extract text, detect section headers, classify narrative content, and generate structured metadata from complex scientific PDFs.
Collaborated on FAIR-aligned data standardization pipelines for ophthalmology imaging datasets-Eye ACT- improving machine learning readiness, interoperability, and future analytics/model development.
Collaborated on schema evaluation and JSON-based FAIR workflows to transform scientific posters-Poster Science-into machine-readable and interoperable research objects for AI-driven discovery and reuse
Prepare technical reports, manuscripts, and grant applications with investigators

Aug 2021 —Jan 2025

Graduate Research & Teaching Assistant

University of Iowa, Department of Computer Science · Iowa City, IA

Designed transformer-based NLP systems — Symptom-BERT, Care-BERT, Spiritual-BERT — for classification, NER, and extraction from 5M+ clinical notes.
Built end-to-end NLP/LLM pipelines covering preprocessing, feature engineering, training, validation, and statistical evaluation.
Symptom-GPT: Led and developed transformer-based clinical NLP and LLM models for named entity recognition and cancer symptom detection from large-scale oncology 1M EHR clinical notes, achieving F1 scores up to 0.989 for nausea/vomiting and 0.912 for anxiety detection.
Symptom-BERT: Led and developed transformer-based clinical NLP and LLM models for further pretrain and fine-tune transformer models on oncology 1M EHR data to detect 13 cancer symptom groups, achieving micro-F1 = 0.933 in internal validation and micro-F1 = 0.831 in external validation.
Care-BERT: Led and built transformer-based clinical NLP models to identify care priorities and life-sustaining treatment preferences from EHR narratives, achieving F1 = 0.941 and AUC = 0.978 in internal validation.
Spiritual-BERT: Led and developed a Bio Clinical BERT-based NLP system to detect under-documented spiritual care information from 3.6M+ EHR notes, achieving F1 = 0.938 in internal validation and F1 = 0.832 in external validation.
Implemented GPT-based workflows for synthetic clinical note generation, supporting data augmentation and external validation.
OASIS – Oncology Associated Symptoms & Individualized Strategies: Collaborated with clinicians, researchers, and technical collaborators on AI- and NLP-driven oncology symptom monitoring research using EHR data to support personalized cancer care and clinical decision support.
Patient-Reported Outcomes & EHR Symptom Concordance: Collaborated and analyzed concordance between patient-reported and provider-documented cancer symptoms, identifying significant documentation gaps across multiple symptom groups.
Embeddings-Augmented NLP for Symptom Detection: Collaborated and developed embeddings-enhanced NLP pipelines for extracting cancer symptom information from 900K+ clinical notes, achieving F1 scores up to 0.937 for symptom classification tasks.
Cancer Symptom Prediction Using Machine Learning: Collaborated and built predictive ML models using structured and unstructured EHR data to forecast cancer symptom development, with Random Forest achieving macro-AUC = 0.755 and pain prediction AUC = 0.954.
Systematic Review of ML for Cancer Symptom Prediction: Led and conducted a PRISMA-guided systematic review of 42 studies evaluating machine learning approaches for cancer symptom prediction and personalized oncology care.
Web Analysis: Developed an end-to-end NLP and machine learning pipeline for large-scale web scraping, sentiment classification, clustering, and topic modeling across 3,000+ online articles.
Demand Forecasting for E-Commerce Sales: Built Python-based forecasting and analytics models to predict product demand and support inventory optimization for e-commerce sales data.
Taught Python and machine learning courses, guiding students to develop strong skills in programming and data science.
Published over 10 papers across machine learning and data science topics.

Jun 2024 —Aug 2024

NLP Data Scientist

National Cancer Institute & Frederick National Laboratory, NCATS / NIH · Maryland

RARe-SOURCE™: Collaborated and developed AI-powered biomedical literature intelligence workflows using LLMs, semantic search, and NLP to support rare disease knowledge discovery and scientific insight extraction.
Worked closely with biomedical researchers to convert research questions into usable AI workflows and decision-support tools.

Feb 2019 —Jul 2021

Software Engineer

Khorshid Hospital · Isfahan, Iran

Built and enhanced EMR, PIS, and LIS systems with clinical stakeholders, improving workflow efficiency by 35%.
Developed a heart-failure symptom-tracking mobile app used by 3,000+ patients in rural Isfahan.

Dec 2016 —Jan 2019

Software Engineer

Parisian Institute · Tehran, Iran

Designed an EHR management dashboard that reduced report turnaround time by 68%.
Trained 1,500+ clinicians on EHR adoption and workflow best practices.

Multimodal Document AI · Metadata Extraction

DMP Bridge

Open-source Python-based GenAI pipeline that converts funder-specific Data Management Plan PDFs into interoperable DMPTool JSON and RDA Common Standard maDMP JSON. Uses pdfplumber, Qwen2-VL, Llama, prompt engineering, and schema-constrained generation for metadata extraction and FAIR-aligned interoperability.

PythonQwen2-VL pdfplumberLLMs RDA maDMP

Generative AI · RAG

DMP Chef

Free and open-source AI-driven platform that helps researchers draft funder-compliant Data Management Plans. Uses GPT-4, Gemini, and Llama 3.3 with RAG workflows to generate structured, FAIR-aligned, and machine-actionable DMPs.

GPT-4Gemini Llama 3.3RAG Vector DB

Automatic Evaluation · Human Evaluation

NIH DMP LLM Evaluation

Evaluated Llama 3.3 and GPT-4.1 for drafting NIH-compliant Data Management Plans using automated reference-based metrics and human expert evaluation, assessing compliance, clarity, completeness, and usefulness.

GPT-4.1Llama 3.3 ROUGESBERT Human Evaluation

LLM Generation · Prompt Engineering

NIH DMP Generation

Built a reproducible workflow for automatically generating NIH-compliant DMP drafts using Llama 3.3 and GPT-4.1. The repository includes prompt engineering pipelines, generated outputs, and evaluation artifacts.

LLMsPrompt Engineering PythonFAIR

Clinical NER . Classification

Symptom-GPT / Symptom-BERT

Published clinical NLP research using BERT- and GPT-based NER models to detect anxiety and nausea/vomiting symptoms from oncology EHR notes. Symptom-BERT achieved F1 scores of 0.989 for nausea/vomiting and 0.912 for anxiety.

Bio-ClinicalBERTBio-GPT NERPyTorch

Clinical NLP . Classification

Care-BERT for Heart Failure

Transformer-based NLP model for detecting care priorities in EHR notes of older adults with heart failure, including comfort measures only and life-sustaining treatments. Achieved internal F1 = 0.941 and external F1 = 0.876.

BERTEHR Care PrioritiesAUC

Palliative Care Informatics

Care-BERT for Advance Directives

Used Care-BERT to extract life-sustaining treatment preferences from EHR narratives and study disparities in advance directive completion among 14,303 older adults with chronic conditions.

Care-BERTLogistic Regression EHR AnalyticsPalliative Care

Under-Documented Care

Spiritual-BERT

Clinical NLP model for detecting rarely documented spiritual care information in EHRs. Applied to nearly 3.6M notes from 14,729 older adults, achieving F1 = 0.938 internally and F1 = 0.832 externally.

Bio-ClinicalBERTSynthetic Notes EHRClinical NLP

Literature Intelligence · NIH

RARe-SOURCE™

AI-powered literature analysis engine built at the National Cancer Institute to surface hidden insights in rare-disease research. Combines large language models with encoder architectures for biological domain-specific Q&A.

LLMsSemantic Search Rare DiseaseBiomedical

Oncology NLP

Symptom-BERT

Published model for detecting 13 cancer symptom groups from clinical notes. Pretrained Bio-ClinicalBERT on 1M unlabeled clinical documents, fine-tuned on annotated notes, and validated using GPT-4-generated synthetic notes.

BERTPyTorch Hugging FaceClinical NLP

EHR . NimbleMiner . Statistical modeling

Patient-Reported Outcomes vs EHR Documentation

Analyzed concordance between patient-reported symptom occurrence and provider-documented symptoms in EHRs for patients receiving cancer treatment with multimorbidity, using NimbleMiner and statistical modeling.

PROsNimbleMiner Logistic RegressionEHR

Embeddings NLP . Deep Learning

Embeddings-Augmented Cancer Symptom NLP

Developed an embeddings-augmented NLP system to detect 14 cancer symptom groups and distinguish observed symptoms from negated symptoms and medication-related side effects across 902,508 clinical notes.

EmbeddingsNLP Symptom DetectionEHR

Collaborative Research · Oncology Symptom Monitoring

OASIS

Collaborative research project focused on Oncology Associated Symptoms & Individualized Strategies. Contributed to AI/NLP-supported workflows for cancer symptom monitoring, EHR-based symptom analysis,and patient-centered oncology care research in collaboration with interdisciplinary nursing informatics teams.

Clinical NLPEHR Data OncologySymptom Monitoring

Predictive Modeling . Machine Learning

ML for Cancer Symptom Prediction

Used structured and unstructured EHR data from 8,156 adults with cancer to predict 12 common symptoms. Random Forest achieved the strongest overall performance with macro AUC = 0.755 and F1 = 0.729.

Random ForestXGBoost SHAPEHR

Systematic Review . Python

ML for Cancer Symptom Prediction Review

PRISMA-guided systematic review of 42 studies using machine learning to predict cancer symptoms and identify predictors. Synthesized algorithms, cancer sites, symptoms, sample sizes, and research gaps.

PRISMAMachine Learning Cancer SymptomsReview

Web Mining

Web Analysis

Built a complete NLP pipeline for scraping 3,000+ eHow home and living articles, manually labeling sentiment data, training eight classifiers, and applying K-Means clustering and LDA topic modeling.

Web ScrapingSentiment Analysis K-MeansLDA

Forecasting . Machine Learning

Python Demand Forecasting

Applied Python-based data analysis and predictive modeling to forecast three-month product demand for an online electronics retailer using 100 weeks of weekly sales data across 44 SKUs.

PythonTime Series ForecastingE-Commerce

Big Data Healthcare

Big Data Analytics in Healthcare

Published survey categorizing Big Data analytics applications in healthcare using the WHO “6 building blocks of health systems” framework, reviewing 130 articles and books over a 10-year period.

Big DataHealthcare WHO FrameworkSurvey

Health Interoperability

Health Information Systems Interoperability

Proposed a conceptual model using a Health Service Bus and service-oriented architecture to improve interoperability across hospital information systems, EHRs, CDSS, telemedicine systems, and other healthcare platforms.

HSBSOA HL7OpenEHR

#	Title & Authors	Status
—	Evaluating the Performance of LLMs in Creating NIH Data Management Plans Zeinali N., Patel B., et al.	Review
—	Artificial Intelligence Reveals Language Disparities in Person-Centered Spiritual Care: Access and Timing Among Older Adults AlBashayreh A., Zeinali N., et al.	Review
—	Advance Directives and Dementia: How Illness Trajectories Influence Goals-of-Care AlBashayreh A., Zeinali N., et al.	Review

Title & Authors

Status

—

Evaluating the Performance of LLMs in Creating NIH Data Management Plans

Zeinali N., Patel B., et al.

Review

—

Artificial Intelligence Reveals Language Disparities in Person-Centered Spiritual Care: Access and Timing Among Older Adults

AlBashayreh A., Zeinali N., et al.

Review

—

Advance Directives and Dementia: How Illness Trajectories Influence Goals-of-Care

AlBashayreh A., Zeinali N., et al.

Review

#	Title & Authors	Venue
01	Using Large Language Models to Detect Anxiety and Nausea/Vomiting Documentation in Clinical Notes of Patients with Cancer Zeinali N., White S., et al.	CIN Journal
02	Goals-of-Care in Older Adults with Heart Failure, Cancer, and Dementia: Classifying Comfort and Life-Sustaining Preferences Using Priorities-BERT AlBashayreh A., Zeinali N., et al.	Innovation in Aging
03	An Informatics Approach to Characterizing Spiritual Care Documentation in Electronic Health Records of Older Adults AlBashayreh A., Zeinali N., Gilbertson-White S.	ACI Journal

Title & Authors

Venue

Using Large Language Models to Detect Anxiety and Nausea/Vomiting Documentation in Clinical Notes of Patients with Cancer

Zeinali N., White S., et al.

CIN Journal

Goals-of-Care in Older Adults with Heart Failure, Cancer, and Dementia: Classifying Comfort and Life-Sustaining Preferences Using Priorities-BERT

AlBashayreh A., Zeinali N., et al.

Innovation in Aging

An Informatics Approach to Characterizing Spiritual Care Documentation in Electronic Health Records of Older Adults

AlBashayreh A., Zeinali N., Gilbertson-White S.

ACI Journal

#	Title & Authors	Venue
04	Symptom-BERT: Enhancing Cancer Symptom Detection in EHR Clinical Notes ★ AMIA Zeinali N., White S., et al.	J. Pain & Symptom Mgmt.
05	Machine Learning Approaches to Predict Symptoms in People with Cancer: A Systematic Review Zeinali N., Gilbertson-White S., et al.	JMIR Cancer
06	Natural Language Processing Accurately Differentiates Cancer Symptom Information in EHR Narratives AlBashayreh A., Bandyopadhyay A., Zeinali N., et al.	JCO CCI
07	Predictors of Concordance Between Patient-Reported and Provider-Documented Symptoms in the Context of Cancer and Multimorbidity Gilbertson-White S., AlBashayreh A., Bandyopadhyay A., Zeinali N., et al.	ACI Journal
08	Using Real-World EHR Data to Predict the Development of 12 Cancer-Related Symptoms in Multimorbidity Bandyopadhyay A., AlBashayreh A., Zeinali N., et al.	Open JAMIA
09	Innovating the Detection of Care Priorities in Heart Failure Using Large Language Models AlBashayreh A., Zeinali N., Gilbertson-White S.	Innovation in Aging

Title & Authors

Venue

Symptom-BERT: Enhancing Cancer Symptom Detection in EHR Clinical Notes ★ AMIA

Zeinali N., White S., et al.

J. Pain & Symptom Mgmt.

Machine Learning Approaches to Predict Symptoms in People with Cancer: A Systematic Review

Zeinali N., Gilbertson-White S., et al.

JMIR Cancer

Natural Language Processing Accurately Differentiates Cancer Symptom Information in EHR Narratives

AlBashayreh A., Bandyopadhyay A., Zeinali N., et al.

JCO CCI

Predictors of Concordance Between Patient-Reported and Provider-Documented Symptoms in the Context of Cancer and Multimorbidity

Gilbertson-White S., AlBashayreh A., Bandyopadhyay A., Zeinali N., et al.

ACI Journal

Using Real-World EHR Data to Predict the Development of 12 Cancer-Related Symptoms in Multimorbidity

Bandyopadhyay A., AlBashayreh A., Zeinali N., et al.

Open JAMIA

Innovating the Detection of Care Priorities in Heart Failure Using Large Language Models

AlBashayreh A., Zeinali N., Gilbertson-White S.

Innovation in Aging

#	Title & Authors	Year
10	Application of Big Data Analysis in Healthcare Based on the Six Building Blocks of Health Systems' Framework: A Survey Nazari E., Zeinali N., et al. — Dokkyo Journal of Medical Sciences	2021
11	Provide Interoperability Model to Interact in Hospital Information Systems Zeinali N., Asosheh A., et al. — J. Health and Biomedical Informatics	2017
12	The Conceptual Model to Solve the Problem of Interoperability in Health Information Systems Zeinali N., Asosheh A., et al. — IEEE IST	2016
13	The Common Applications of Social Networks in Healthcare Delaram Z., Zeinali N., et al. — Health Information Management	2016

Title & Authors

Year

Application of Big Data Analysis in Healthcare Based on the Six Building Blocks of Health Systems' Framework: A Survey

Nazari E., Zeinali N., et al. — Dokkyo Journal of Medical Sciences

2021

Provide Interoperability Model to Interact in Hospital Information Systems

Zeinali N., Asosheh A., et al. — J. Health and Biomedical Informatics

2017

The Conceptual Model to Solve the Problem of Interoperability in Health Information Systems

Zeinali N., Asosheh A., et al. — IEEE IST

2016

The Common Applications of Social Networks in Healthcare

Delaram Z., Zeinali N., et al. — Health Information Management

2016

Title	Venue	Year
Evaluating the Performance of Large Language Models in Drafting Data Management Plans Zeinali N. (Presenter), Patel B., et al.	BOSC Conference	2026
Toward Envision Portal: A FAIR and AI-Ready Framework for Ophthalmic Imaging Patel B., Zeinali N., et al.	ARVO Annual Meeting	2026
Evaluating the Effectiveness of an Open-Source LLM in Drafting NIH Data Management Plans Zeinali N. (Presenter), Patel B., et al.	SciDataCon	2025
Leveraging LLMs for Named Entity Recognition of Anxiety and Nausea/Vomiting in Patients with Cancer Zeinali N. (Presenter), Gilbertson-White S., et al.	AMIA Informatics Summit	2025
Advanced Detection of Nausea/Vomiting and Anxiety in Patients with Cancer Zeinali N. (Presenter), Gilbertson-White S., et al.	AMIA Annual Symposium	2024
Comparison of BERT Implementations for Enhanced Cancer Symptom Extraction (Oral) Zeinali N. (Presenter), AlBashayreh A., Gilbertson-White S., et al.	IEEE AIMHC	2024
Leveraging Spiritual-BERT for Characterizing Spiritual Care Documentation in EHRs of Older Adults with Heart Failure AlBashayreh A., Zeinali N., et al.	AMIA Annual Symposium	2024
Innovating the Detection of Care Priorities in Heart Failure Using Large Language Models AlBashayreh A., Zeinali N., et al.	GSA Annual Meeting (Poster)	2024
Disparities in Advance Directive Completion and Life-Sustaining Treatment Preferences in Older Adults AlBashayreh A., Zeinali N., et al.	Annual Assembly Hospice & Palliative Care (Poster)	2024

Title

Venue

Year

Evaluating the Performance of Large Language Models in Drafting Data Management Plans

Zeinali N. (Presenter), Patel B., et al.

BOSC Conference

2026

Toward Envision Portal: A FAIR and AI-Ready Framework for Ophthalmic Imaging

Patel B., Zeinali N., et al.

ARVO Annual Meeting

2026

Evaluating the Effectiveness of an Open-Source LLM in Drafting NIH Data Management Plans

Zeinali N. (Presenter), Patel B., et al.

SciDataCon

2025

Leveraging LLMs for Named Entity Recognition of Anxiety and Nausea/Vomiting in Patients with Cancer

Zeinali N. (Presenter), Gilbertson-White S., et al.

AMIA Informatics Summit

2025

Advanced Detection of Nausea/Vomiting and Anxiety in Patients with Cancer

Zeinali N. (Presenter), Gilbertson-White S., et al.

AMIA Annual Symposium

2024

Comparison of BERT Implementations for Enhanced Cancer Symptom Extraction (Oral)

Zeinali N. (Presenter), AlBashayreh A., Gilbertson-White S., et al.

IEEE AIMHC

2024

Leveraging Spiritual-BERT for Characterizing Spiritual Care Documentation in EHRs of Older Adults with Heart Failure

AlBashayreh A., Zeinali N., et al.

AMIA Annual Symposium

2024

Innovating the Detection of Care Priorities in Heart Failure Using Large Language Models

AlBashayreh A., Zeinali N., et al.

GSA Annual Meeting (Poster)

2024

Disparities in Advance Directive Completion and Life-Sustaining Treatment Preferences in Older Adults

AlBashayreh A., Zeinali N., et al.

Annual Assembly Hospice & Palliative Care (Poster)

2024

Multimodal AI for biomedical discovery.

About

Impact at a glance

Professional experience

Selected projects

DMP Bridge

DMP Chef

NIH DMP LLM Evaluation

NIH DMP Generation

Symptom-GPT / Symptom-BERT

Care-BERT for Heart Failure

Care-BERT for Advance Directives

Spiritual-BERT

RARe-SOURCE™

Symptom-BERT

Patient-Reported Outcomes vs EHR Documentation

Embeddings-Augmented Cancer Symptom NLP

OASIS

ML for Cancer Symptom Prediction

ML for Cancer Symptom Prediction Review

Web Analysis

Python Demand Forecasting

Big Data Analytics in Healthcare

Health Information Systems Interoperability

Publications

Under Review

2025

2024

Earlier Work

Presentations & posters

Technical skills

LLMs & Generative AI

Deep Learning & ML

NLP Frameworks

MLOps & CI/CD

Cloud & Compute

Databases & Vector Stores

Analytics & Statistics

Programming Languages

Healthcare Standard

Education & recognition

Community & service

Get in touch

Send a message