# Radiomics-based Risk Prediction for Therapy Selection in Crohn’s Disease via MRI

> **NIH NIH F31** · CASE WESTERN RESERVE UNIVERSITY · 2022 · $35,272

## Abstract

PROJECT SUMMARY: Of the estimated 33,000 patients who were newly diagnosed with Crohn’s Disease (CD)
in 2020, a majority of moderate to severe cases will receive biologic anti-TNF therapy to reduce bowel wall
inflammation and ensure mucosal healing. Unfortunately, 30-40% will fail to achieve clinical remission on their
first-line biologic treatment and will require surgery to treat the eventual complications (strictures, fistulas).
Stricturing disease specifically presents as a constricted region of bowel which can have extensive fibrosis which
often renders medical therapy ineffective. Unfortunately, current radiological, genetic, and endoscopic markers
are limited in their in vivo characterization of CD; making them poor prognosticators of response to therapy.
 The emerging field of radiomics (high-throughput extraction of computerized features from medical
images) offers the ability to quantify subtle textural and structural aspects of diseased tissue in vivo. I hypothesize
that radiomic techniques can be optimized to capture CD-specific facets of the diseased small bowel wall and
surrounding tissue on MRI, thus quantifying prognostic and physiologic cues that are not visually appreciable. In
addition to capturing structural distention of the inflamed bowel wall, radiomics features can be used to quantify
subtle inflammation-related heterogeneity and tissue organization in the bowel wall and the surrounding
mesenteric fat. Further, I plan to optimize radiomic descriptors to account for known sources of imaging variance
due to scanner and acquisition differences. This will be critical for clinical translation of radiomic descriptors as
robust and reliable risk predictors of treatment response in CD as well as for accurately characterizing fibrotic
and inflammatory phenotypes of stricturing CD on MRI.
 I plan to achieve these objectives with the support of an outstanding sponsor team as well as an excellent
multi-disciplinary clinical team; in conjunction with multi-faceted training experiences. In Aim 1, I will develop a
suite of radiomic features that are associated with active CD by specifically quantifying textural and morphometric
heterogeneity of inflamed bowel wall as well as bowel-adjacent mesenteric fat on MRI. I will also optimize them
within an image processing framework to ensure their reproducibility and generalizability across multiple sources
of imaging variance. Aim 2A will leverage the most relevant subset of these radiomic features that are associated
with likelihood of response to biologic therapy as well as integrate them with existing clinical biomarkers to build
a multivariate risk classifier for prognosticating likelihood of remission to biologics therapies in CD. In Aim 2B I
will investigate the ability of radiomic features to distinguish between fibrotic and inflammatory phenotypes in
stricturing disease, through detailed correlation of pathological phenotypes with radiomic features extracted on
MRI. This project builds ...

## Key facts

- **NIH application ID:** 10466352
- **Project number:** 1F31DK130587-01A1
- **Recipient organization:** CASE WESTERN RESERVE UNIVERSITY
- **Principal Investigator:** Prathyush V Chirra
- **Activity code:** F31 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2022
- **Award amount:** $35,272
- **Award type:** 1
- **Project period:** 2022-09-01 → 2023-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10466352

## Citation

> US National Institutes of Health, RePORTER application 10466352, Radiomics-based Risk Prediction for Therapy Selection in Crohn’s Disease via MRI (1F31DK130587-01A1). Retrieved via AI Analytics 2026-05-23 from https://api.ai-analytics.org/grant/nih/10466352. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*