Picked from the software, topic, and level of “Class Wavelet Deep Learning EKG Aritmia Classification #3: Wavelet: DWT for Features and Scalogram (CWT) as Image”.
At most one email a week. Unfollow anytime with one click.
Transcript
The full narration of this video. The current sentence is highlighted while the video plays; click any sentence to jump there.
Show transcript
Introduction
Here is the final result, two ways wave-flat view the heart rate. Dewi T breaks bits into frequency bands and summarizes them into 15 numbers, while Sewi T transforms them into colorful scalograms that clearly differ for each type of arrhythmia. Let's start. Hello, welcome to episode three of the EKG classification series with wave-flat and deep learning. The Fourier transform only tells which frequencies are present, but not when. Wave-flat provides both, making it very suitable for EKG containing sharp waves like QRS and slow waves like T. Today we are using two types of wave-flat.
Step 1: DWT Features
Create a file wave-flat underscore ek.py The Dewi T feature function decomposes each bit with wave-flat dobesi 4 up to 4 levels, producing approximation coefficients A4 for low frequencies and detail D4 to D1 for higher frequencies. From each subband, three statistics are taken: energy, standard deviation, and absolute maximum, totaling 15 features per bit.
Step 2: CWT Scalogram
The scalogram function calculates continuous wave-flat transform with wave-flat more-light on 32 logarithmically spread scales. Each bit is resampled into 64 samples, becoming an image of 32 x 64, then normalized to 0 to 1. This image will be processed by a 2D CNN like a photo.
Step 3: Energy per class in sub-bands
Create a file uji underscore wave-flat.py and run it. 50 thousand bits produce 15 Dewi T features each in a few seconds. The length of coefficients A4 to D1 is 14, 14, 21, 36, and 66, D1 represents 45 to 90 Hz and A4 0 to 5.6 Hz. Average energy shows characteristic patterns, bit V has A4 energy of 129, almost 2 times bit N67, but D1 energy is only 0.04 compared to 0.4 on N, because ventricular wide and slow QRS, poor high frequency.
Step 4: Coefficients & Scalogram
Let's look at the graph. The top line shows coefficients A4, D4, D3, and D21 of normal bits, QRS appears as a spike in detail. The bottom line is the scalogram of the four classes, bit N has narrow energy spots at small scales during QRS, bit V spreads to large scales due to its complexity, and bit S is similar to N but shifted. This visual difference will be learned by the CNN.
Done!
Done. We have extracted DWT-Dobesi 4 features and created a CWT-Morlet scalogram. In the next episode, we will train SVM and Random Forest with Wavelet features and RR. Currently working on a thesis, dissertation, or research in the field of biomedical engineering and artificial intelligence? Cerdas Inovasi Digital helps with the implementation of biomedical signal processing, from filtering and EKG signal segmentation, Wavelet transformation, classical machine learning, up to 1D and 2D CNNs along with their interpretation, as well as opening Python programming courses for electrical and biomedical engineering. Want to master the entire series? From MITBIH base data to class-wise arrhythmia evaluation? All 6 episodes, complete with step-by-step PDF guides, datasets, and source code, are available for premium tutorial subscribers. Subscribe starting at Rp69.000 per month, save 28% with the 12-month package, or pay once Rp999.000 for lifetime access. Thank you for watching, see you in the next episode. Thank you for watching!
Ask the class material
Type a question or keyword; we find the part of this episode that explains it and jump the video there.
My notes
Notes are saved with the current video time and are only visible to you.
Questions & answers
Stuck on a step? Ask here; the instructor and other learners can answer.
No questions yet.
Your instructor
Cerdas Inovasi Digital Team
Software developers, BIM & 3D engineers, and AI practitioners
52 video classes1,277 episodes
We build software, BIM models, games and AI automations for clients every day, and turn that hands-on work into step-by-step classes: every step is shown on screen and comes with a PDF guide you can keep.
