Tumor Connectomics: Mapping the intra-tumoral complex interaction network
Tumors are extremely heterogeneous and comprise of a number of intratumor microenvironments or sub-regions. These tumor microenvironments may interact with eac based on complex high-level relationships, which could provide important insight into the organizational structure of the tumor network. To that end, we developed a tumor connectomics framework (TCF) to understand and model the complex functional and morphological interactions within the tumor. Then, we demonstrate the TCF's potential in predicting treatment response in breast cancer patients being treated with neoadjuvant chemotherapy. The TCF was implemented on a breast cancer patient cohort of thirty-four patients with dynamic contrast enhanced (DCE) magnetic resonance imaging (MRI) undergoing neodjuvant chemotherapy treatment. The intra-tumor network connections (tumor connectome) before and after treatment were modeled using advanced graph theoretic centrality, path length and clustering metrics from the DCE-MRI. The percentage change of the graph metrics between two time-points (Baseline and 1st cycle) was computed to predict the patient's final response to treatment. The TCF visualized the inter-voxel network connections across multiple time-points and was able to evaluate specific changes in the tumor connectome with treatment. Degree centrality was identified as the most significant predictor of treatment response with an AUC of 0.83 for classifying responders from non-responders. In conclusion, the TCF graph metrics produced excellent biomarkers for prediction of breast cancer treatment response with improved visualization and interpretability of changes both locally and globally in the tumor.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Incorporating intratumoral heterogeneity into weakly-supervised deep learning models via variance pooling
Supervised learning tasks such as cancer survival prediction from gigapixel whole slide images (WSIs) are a critical challenge in computational pathology that requires modeling complex features of the tumor microenvironm…
Survival Predictionwhole slide imagesQuantifying intra-tumoral genetic heterogeneity of glioblastoma toward precision medicine using MRI and a data-inclusive machine learning algorithm
Glioblastoma (GBM) is one of the most aggressive and lethal human cancers. Intra-tumoral genetic heterogeneity poses a significant challenge for treatment. Biopsy is invasive, which motivates the development of non-invas…
GANDA: A deep generative adversarial network predicts the spatial distribution of nanoparticles in tumor pixelly
Intratumoral nanoparticles (NPs) distribution is critical for the success of nanomedicine in imaging and treatment, but computational models to describe the NPs distribution remain unavailable due to the complex tumor-na…
Generative Adversarial Networkwhole slide imagesGlioma Grade Prediction Using Wavelet Scattering-Based Radiomics
Glioma grading before surgery is very critical for the prognosis prediction and treatment plan making. We present a novel wavelet scattering-based radiomic method to predict noninvasively and accurately the glioma grades…
Dimensionality ReductionPredictionPrognosisregressionArtificial intelligence-based locoregional markers of brain peritumoral microenvironment
In malignant primary brain tumors, cancer cells infiltrate into the peritumoral brain structures which results in inevitable recurrence. Quantitative assessment of infiltrative heterogeneity in the peritumoral region, th…
Decision MakingDescriptivePrognosis