Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( 🔒 ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Need Help

Novel CT-based Biomarkers to Improve Acute Stroke Triage

Award Number: K23NS144706
ORGANIZATION: NATIONAL INSTITUTE OF NEUROLOGICAL DISORDERS AND STROKE
OPDIV: NIH
AWARD CLASS: DISCRETIONARY
AWARD ACTIVITY TYPE: TRAINING/TRAINEESHIP
PERIOD OF PERFORMANCE START DATE: 09/07/2026
PERIOD OF PERFORMANCE END DATE: 07/31/2031

Group Awards By:

View Award Description

Novel CT-based Biomarkers to Improve Acute Stroke Triage - PROJECT SUMMARY Acute ischemic stroke (AIS) is a leading cause of death and disability around the world, and despite the advancements in reperfusion treatments, many are left with substantial functional disability. Furthermore, cognitive disability is an increasingly recognized, but still under appreciated, consequence of AIS. Our current imaging methods of evaluating early infarct on non-contrast head CT and CT perfusion are incomplete, and patients have a wide range of outcomes regardless of infarct estimations from these standard tools. Preliminary studies have shown that radiomics and deep learning have the ability to extract far more information from CT imaging. A recently developed radiomics pipeline performs automated pre-processing and calculation of radiomic features including net water uptake (NWU), quantifying brain injury and edema based on density differences between the stroke area and contralateral normal brain. The hypothesis based on the preliminary data is that NWU and other quantitative CT markers can accurately predict long-term functional outcomes. As part of the future of precision stroke medicine, highly accurate tools will be required to predict not just short-term outcomes, but long-term brain health. Beyond radiomic markers, deep learning (DL) techniques have shown excellent reliability and may perform even better than individual quantitative measurements, especially when trained with a diverse clinical and imaging dataset. In Aim 1, the performance of NWU and CT radiomic markers will be determined when predicting 90-day functional outcomes for patients with AIS. The existing cohorts include internal testing and fully external validation, and they contain biological variables, social factors, and CT imaging. The primary outcome will be the predictive performance compared to standard CT scores, adjusted for confounding factors, and quantified by area under the ROC curve. In Aim 2, DL models will be developed and validated to predict 90-day functional outcome after AIS. Convolutional neural networks will be evaluated to perform the prediction task using whole brain non-contrast CT and clinical variables, using internal cohort training/testing and fully external validation. In Aim 3, DL models will be explored to determine performance when predicting long-term cognitive outcome, which has not been done previously in a large cohort study of AIS. Inputs will include comprehensive clinical and imaging data, and performance will be benchmarked, with a particular focus on interpretability methods to identify the key features that are most important for prediction. This K23 Award will provide crucial training and mentorship for Dr. Jerome Jeevarajan to develop into a clinician- scientist, including coursework at the School of Biomedical Informatics. The data from this project will be the catalyst for independent R01 funding. This study is anticipated to provide new understanding of ischemia and reperfusion as we learn which features are most impactful for prediction of post-AIS outcomes. Using widely available non-contrast CT, these new tools could be used to identify ideal candidates for reperfusion and adjunctive treatments during upcoming clinical trials.


Issue Date FY	Funding FY	Legal Entity Name	Legal Entity Address	Legal Entity City	Legal Entity State	Legal Entity Zip Code	Legal Entity COUNTY	Legal Entity COUNTRY	Assistance Listing	Award Code	Budget Year	Action Date	Action Type	Action Amount

Issue Date FY: 2026 ( Subtotal = $217,080 )
2026	2026	UNIVERSITY OF TEXAS HEALTH SCIENCE CENTER AT HOUSTON	7000 FANNIN ST FL 9	HOUSTON	TX	77030	HARRIS	USA	Extramural Research Programs in the Neurosciences and Neurological Disorders	000	1	9/7/2026	NEW	$217,080
														Subtotal = $217,080

Grand Total All Awards = $217,080

Top

All Categories

About

Search

Reports

Data Submission

Award Information

Novel CT-based Biomarkers to Improve Acute Stroke Triage

Award Number: K23NS144706

ORGANIZATION: NATIONAL INSTITUTE OF NEUROLOGICAL DISORDERS AND STROKE

OPDIV: NIH

AWARD CLASS: DISCRETIONARY

AWARD ACTIVITY TYPE: TRAINING/TRAINEESHIP

PERIOD OF PERFORMANCE START DATE: 09/07/2026

PERIOD OF PERFORMANCE END DATE: 07/31/2031

Federal Websites

Department of Health & Human Services

HHS Operating Divisions

HHS Staff Divisions

Download A Document Viewer