Millions of people die annually from cardiovascular diseases. Most of these deaths are due to the misjudgement of early symptoms of heart attacks – making early detection crucial for patient survival. It has been demonstrated that certain cardiac indicators in blood increase during the progression of a heart attack. To address this challenge, a test strip for such early detection is developed using commonly available paper that works with a small quantity of simulated blood. The indicators are separated from the whole blood sample using a patent-pending graphene oxide filter membrane. Machine learning is implemented to provide predictive capability of the fabricated device to monitor heart conditions. The entire process from sample collection to detection and prediction of results is less than two minutes with a final device cost of around CAD1.00, making a significant impact on United Nations’ Sustainable Development Goals in healthcare.