Browsing by Author "Gaber, M. Waleed"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Identifying Image Derived Features of Radiation Therapy Response: Tumor and Normal Tissue
    (2018-08-01) Tang, Tien T; Qutub, Amina A.; Gaber, M. Waleed
    Brain tumors constitutes the second most common malignancy in children. Management of these tumors with surgical resection, radiation therapy and chemotherapy presents significant challenges, with cure rates lagging compared to other pediatric cancers. While the introduction of radiation therapy (RT) has significantly improved patient outcome, survivors are never the less prone to cognitive impairment and other radiation-induced side effects. Therefore early detection of treatment resistance and treatment side effects are important for treatment planning and patient prognosis. Monitoring of brain tumor’s response is commonly done using medical imaging techniques such as magnetic resonance (MR) and positron emission tomography (PET). In addition to the clinical value of providing information regarding tumor location, size, and metabolism, these images can also be further analyzed to extract quantitative imaging features which can provide additional information for tumor characterization that preserves the spatial and temporal heterogeneity of the tumor. In this work, texture analysis will be utilized to establish quantitative image features that will assist in understanding and predicting RT response of tumors and detection of radiation-induced normal tissue injury. Using preclinical models, quantitative image features will be mined from MR and PET scans in radioresponsive and radioresistant tumors to establish universal and tumor-specific imaging markers of treatment response. Furthermore we will establish imaging markers that will provide immediate readout of normal tissue injury and map out the long term changes caused by RT. The outcome of our research will provide clinicians with a toolset to predict, detect, and understand RT response in both tumor and normal tissue for the personalization of treatment for affected children.In this work, texture analysis will be utilized to establish quantitative image features that will assist in understanding and predicting RT response of tumors and detection of radiation-induced normal tissue injury. Using preclinical models, quantitative image features will be mined from MR and PET scans in radioresponsive and radioresistant tumors to establish universal and tumor-specific imaging markers of treatment response. Furthermore we will establish imaging markers that will provide immediate readout of normal tissue injury and map out the long term changes caused by RT. The outcome of our research will provide clinicians with a toolset to predict, detect, and understand RT response in both tumor and normal tissue for the personalization of treatment for affected children.
  • Thumbnail Image
    Item
    Image-based Classification of Tumor Type and Growth Rate using Machine Learning: a preclinical study
    (Springer Nature, 2019) Tang, Tien T.; Zawaski, Janice A.; Francis, Kathleen N.; Qutub, Amina A.; Gaber, M. Waleed
    Medical images such as magnetic resonance (MR) imaging provide valuable information for cancer detection, diagnosis, and prognosis. In addition to the anatomical information these images provide, machine learning can identify texture features from these images to further personalize treatment. This study aims to evaluate the use of texture features derived from T1-weighted post contrast scans to classify different types of brain tumors and predict tumor growth rate in a preclinical mouse model. To optimize prediction models this study uses varying gray-level co-occurrence matrix (GLCM) sizes, tumor region selection and different machine learning models. Using a random forest classification model with a GLCM of size 512 resulted in 92%, 91%, and 92% specificity, and 89%, 85%, and 73% sensitivity for GL261 (mouse glioma), U87 (human glioma) and Daoy (human medulloblastoma), respectively. A tenfold cross-validation of the classifier resulted in 84% accuracy when using the entire tumor volume for feature extraction and 74% accuracy for the central tumor region. A two-layer feedforward neural network using the same features is able to predict tumor growth with 16% mean squared error. Broadly applicable, these predictive models can use standard medical images to classify tumor type and predict tumor growth, with model performance, varying as a function of GLCM size, tumor region, and tumor type.