Publications by authors named "Kashvi Ankitbhai Sheth"
BMC Med Imaging
December 2024
Article Synopsis
- Myocardial infarction (MI) is a critical health issue that requires quick and accurate diagnostic methods, and this paper introduces a new technique using the Discrete Wavelet Transform (DWT) to improve detection of MI from ECG signals.
- The DWT helps isolate and reduce noise in ECG recordings, allowing for cleaner signals, which are then analyzed using lightweight one-dimensional Convolutional Neural Networks (CNN) to classify them as either MI or normal.
- The proposed model demonstrates strong performance, achieving a 96% accuracy with one dataset and 91.18% with another, while also addressing class imbalances through under-sampling and up-sampling techniques.
