This study aims to compare conventionally acquired Left Ventricle Ejection Fraction (LVEF) and Global Longitudinal Strain (GLS) data to Artificial Intelligence (AI) driven automated processing of 2 dimensional contrast and 2 dimensional non-contrast resting transthoracic echocardiograms for application in the assessment of patients undergoing chemotherapy with cardiotoxic drugs. This is a single-centre retrospective study which utilizes echocardiographic DICOM image and meta-data datasets received from a Canadian site. Data processed using the AI driven automated processing will be compared to conventionally acquired LVEF and GLS measurements and results will be analysed to determine accuracy and precision.

Inclusion Criteria:

• Normal EF and no regional wall motion abnormalities prior to starting chemotherapy treatment.
• Follow-up EF measurements available for at least 1 year during the treatment period.
• Follow-up EF measurements for at least 1 year during the treatment period.

Exclusion Criteria:

• Age < 18 years
• Inadequate image quality (as determined by the Ultromics Operators Quality Control process)

- Data included for analysis in this study will be retrospective data from male and female patients aged 18 years or above who received an echocardiogram for baseline and follow-up of cardiotoxic agent management, who agreed to participate in the echocardiography registry of the MAHI, and who have already provided a written consent that their anonymized imaging data can be used to advance echocardiographic assessment of LV function.