전체 글 썸네일형 리스트형 Predicting Breast Cancer Recurrence with AI - Customized 3D ResNet152V2 & Ansemble Model Refer to the github page to get the whole code. ResNet is a model using Residual Connections that solves the problem of accuracy saturation when layers are stacked extremely deep. (For detailed explanation, refer to the ResNet paper review or Youtube video) Customized 3D ResNet152V2 In 2D convolution, each channel performs convolution independently, and then the corresponding values are combined.. 더보기 Predicting Breast Cancer Recurrence with AI - Data Augmentation Refer to the github page to get the whole code. Data Augmentation import os import pandas as pd import nibabel as nib import numpy as np import torchio as tio from sklearn.model_selection import StratifiedShuffleSplit def trainset_aug(NumpyPath, MetadataPath, AugFileSaveDir, AugMetadataSaveDir): df = pd.read_excel(MetadataPath, header=0, index_col=0) df = df.query('RECURwithin3yrs != "NotSure"').. 더보기 Predicting Breast Cancer Recurrence with AI - Data Collection, Processing, Labelling Refer to the github page to get the whole code. Contents Data Collection 1-1. Introduction of TCIA (Cancer Image opensource DB) 1-2. Download process 1-3. Why "Duke-Breast-Cancer-MRI" Image data Processing 2-1. Image inspection 2-2. Image processing Clinical data processing 3-1. Clinical data processing 3-2. Determine "RECUR within 3 yrs" 3-3. Determine "No RECUR within 3 yrs" 3-4. Determine "RE.. 더보기 이전 1 2 3 4 다음
