A5028755508- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Brain Tumor Detection and Classification
- Advanced Radiotherapy Techniques
- Lung Cancer Diagnosis and Treatment
Norwegian University of Life Sciences
2019-2021
Target volume delineation is a vital but time-consuming and challenging part of radiotherapy, where the goal to deliver sufficient dose target while reducing risks side effects. For head neck cancer (HNC) this complicated by complex anatomy region proximity volumes organs at risk. The purpose study was compare evaluate conventional PET thresholding methods, six classical machine learning algorithms 2D U-Net convolutional neural network (CNN) for automatic gross tumor (GTV) segmentation HNC...
An automatic segmentation algorithm for delineation of the gross tumour volume and pathologic lymph nodes head neck cancers in PET/CT images is described. The proposed based on a convolutional neural network using U-Net architecture. Several model hyperparameters were explored performance terms Dice similarity coefficient was validated from 15 patients. A separate test set consisting 40 patients used to assess generalisability algorithm. showed close-to-oncologist level delineations as...