The focus of this task is the classification of echocardiography video data, a very noisy but inexpensive and vastly available tool for the diagnostic of heart diseases. In this task we will perform the differential diagnosis of the Acute Coronary Syndrome (ACS) and the Takostubo Syndrome (TTS). The task is highly challenging for the high dimensionality of the data, low number of samples and high noise level.
The data for this task contains the following folders and files:
train - a folder with 158 training videos with name id.avi test - a folder with 69 test videos with name id.avi train_target.csv - the labels of the training data set sample.csv - a sample submission file in the correct format
checkout the following link: https://gitlab.vis.ethz.ch/lucac/aml_task4