Redirect
delay 5 location https://www.cancerimagingarchive.net/collection/pdmr-833975-119-r/

Characterization of tissue using in vivo non-invasive imaging is the foundation of Radiology and is clinically used for detection and measurement of disease burden in oncology.  With the migration of imaging to digital media the possibility for advanced mathematically based imaging biomarkers was realized. As part of this endeavor, researchers have developed various algorithms using neural networks, and classification techniques to improve tissue characterization (morphological changes).  However, large datasets are a requirement in this research endeavor in part due to the genomic heterogeneity of tumors in the same histologic classification – many tumors from different patients are required to have enough with the same genomic characteristics to adequately evaluate the range of imaging variability for a specific genomic pattern.  Pre-clinical animal models of patient derived xenografts may be an important resource by providing collections with a more homogenous tumor genome across the collection with companion extensive tumor genomic characterization available, allowing determination of the variability of imaging characteristics for that pattern in different individuals. This dataset of a patient derived xenograft model adenocarcinoma pancreas PDMR: 833975-119-R can be used for training algorithms for evaluating variations in tissue texture with respect to tumor growth and regrowth after surgical resection.  

In this study we performed a detailed imaging characterization (workflow below) of this model, details are provided in the attached standard operating procedures. Tumors in half of the mice were resected in the range 200-300 mm³ size; tumors in the other half were allowed to grow until it was necessary to euthanize them because of tumor size.