Skip to Main Content

An official website of the United States government

Integrating AI-Driven Analysis of Clinical and Imaging Data to Optimize Lung Cancer Treatment Strategies

Principal Investigator
Name
Ricardo Gonzales Vera
Degrees
D.Phil.
Institution
University of Oxford
Position Title
D.Phil. Candidate
Email
About this CDAS Project
Study
PLCO (Learn more about this study)
Project ID
PLCOI-1517
Initial CDAS Request Approval
Apr 2, 2024
Title
Integrating AI-Driven Analysis of Clinical and Imaging Data to Optimize Lung Cancer Treatment Strategies
Summary
This project aims to enhance lung cancer treatment decisions through the integration of artificial intelligence (AI) with both clinical and imaging data from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. Leveraging the already obtained tabular patient data, we now seek to include imaging components to apply computer vision techniques. The inclusion of imaging data will enable a more comprehensive analysis of lung cancer characteristics, such as tumor morphology and progression, which are critical for tailoring treatment plans. By combining insights from structured patient data with patterns identified in imaging, the project will develop predictive models to guide treatment decisions more effectively. The models are designed to support clinical decision-making by providing evidence-based, data-driven recommendations for personalized treatment approaches, aiming to improve patient outcomes in lung cancer care.
Aims

1. Expansion to Include Imaging Data: Acquire access to the imaging component of the PLCO lung cancer dataset to supplement the tabular patient data. Utilize computer vision techniques to analyze imaging data for features relevant to lung cancer treatment and prognosis.
2. Development of Integrated AI Models: Enhance the existing AI models to incorporate insights from imaging analysis alongside clinical data. This step involves training the models to recognize and interpret complex patterns within the imaging data, and correlate these with treatment outcomes.
3. Validation and Testing with Comprehensive Data: Conduct a validation and testing phase that includes both clinical and imaging data to evaluate the enhanced model's performance. Assess the model's ability to accurately predict treatment outcomes and its sensitivity and specificity in doing so.

Collaborators

Arjun Ulag - Veritas AI

Related Publications