Keywords: Diagnosis/Prediction, Cancer, rectal cancer, deep learning, pathological downstaging
Motivation: Challenges in using conventional magnetic resonance imaging to assess preoperative response to chemoradiotherapy in locally advanced rectal cancer highlight a need for more accurate predictive tools.
Goal(s): Develop and validate deep learning models combining T2-weighted magnetic resonance images with radiological and clinicopathological data to predict pathological T-downstaging after chemoradiotherapy.
Approach: The model, trained on data from 318 patients, utilized receiver operating characteristic curves and statistical tests to assess performance.
Results: Combined deep learning models achieved area-under-curve values between 0.800 and 0.817, surpassing radiologists' assessments and enhancing response evaluation in patients undergoing chemoradiotherapy for locally advanced rectal cancer.
Impact: This study’s deep learning models provide clinicians with an accurate tool for predicting pathological downstaging in locally advanced rectal cancer, improving preoperative assessment after chemoradiotherapy. It also encourages further research into integrating deep learning with multimodal data for cancer prognosis.
How to access this content:
For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.
After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.
After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.
Keywords