Skip to main content

Prediction models for lung cancer therapies use image analysis software

The Moffitt Cancer Center is using image analysis software in the analysis of lung cancer CT images. The Florida-based cancer centre aims to develop more accurate prognosis and prediction models for response to specific lung cancer therapies.

Moffitt is part of an elite group of National Cancer Institute (NCI) Comprehensive Cancer Centers focusing on the development of early stage translational research models. The centre will employ a lung tumour analysis research application developed by Definiens to rapidly and accurately identify, segment and analyse lung tumours from CT and PET/CT fused images. The application will also allow the researchers to measure lung nodule volume, surface-to-volume, attenuation gradient at the edges, shape features, texture and homogeneity measures, as well as tracking tumours' volumetric changes over time. 

'At Moffitt, we are enthusiastic about deploying these cutting-edge tools to support our researchers with their critical work,' said Robert Gillies, vice-chair radiology and director of the Experimental Imaging Program at Moffitt. 'We anticipate that the Definiens platform will streamline the analysis of hundreds, if not thousands, of CT and PET images. These numbers are necessary in order to develop more robust patient stratification models.'

Definiens is developing image analysis applications for a variety of cancer types, addressing lung and liver tumours as next steps. All Definiens medical imaging applications are built upon the company's Definiens Cognition Network Technology, an image analysis technology that examines objects in relation to one another and emulates human cognitive processes to extract intelligence from images.

As part of the research engagement, Moffitt scientists will also explore Definiens TissueMap and Definiens Developer for tissue-based image analysis. Definiens versatile portfolio spans clinical imaging and tissue imaging, supporting potential development of translational cancer research models.

Topics

Media Partners