Running left_ventricular_mass_from_ecg_student_and_mri_teacher models results in very small LVM values

[ezeid]

Hi,

In the README for left_ventricular_mass_from_ecg_student_and_mri_teacher models, it is stated the raw ECG once read from the .xml file (UK biobank files are in uV by default) should be are normalized by dividing by 2000 prior to being input to the model. I testing these models on raw ECG data from UK Biobank. I am following these instructions, but when I predict with the model, I am getting a very low LVM values: in the range [-2, 2].

Am I missing anything here? Is there an example on how I should properly process the data before feeding it to the models to get the proper LVM value?

@rxrathod

Thank you

ezeid