A method for speedy binarized approximate entropy estimation from long data series belongs to biomedical data processing field. The novelty it proposes is an improved estimation of ApEn as the most cited measure for the complexity and cross-complexity evaluation, in cardiovascular, respiratory, neural and other physiological signals. The novelty is that it operates over the binary differentially encoded signals, so the vectors are binary comprising m bits, and there is a total of 2m different vectors. The novelty is the discrete threshold. The novelty is a vector distance measured as a Hamming distance, forming a Hamming distance matrix just once, to be implemented many times. The novelty is that template matching probabilities are estimated from the histogram representatives with already stored distances in a Hamming distance matrix, decreasing the number of processor operations several orders of magnitude. The novelty is that the final summand Φ(m) is estimated jointly using total of 2m vector histograms ant the total of 2m template matching probabilities. The method is intended for wearable devices for self-measuring the vital parameters during the fitness, recreation, sport, walking, sleeping, work, where the processor operations should be diminished because of the battery consumption, and because of the necessity for real time operation.Postupak za ubrzanu binarizovanu procenu aproksimativne entropije iz dugačkih nizova podataka pripada oblasti obrade biomedicinskih signala i ima za novost to što predlaže poboljšano određivanje ApEn - najcitiranije mere za procenu kompleksnosti i unakrsne kompleksnosti kardiovaskularnih, respiratornih, neuroloških i sličnih fizioloških signala. Novost je što operiše nad binarnim diferencijalno kodovanim signalom što omogućuje da vektori dužine m budu binarni, tako da postoji ukupno 2m različitih vektora. Novost je što je skup dozvoljenih pragova za poređenje diskretan. Novost je što se rastojanje između vektora definiše kao Hem
