Fourier analysis is well known to provide complete information of the frequencies present in a signal. But in the process, time information is lost. Therefore, its time-frequency representation is required for depicting both time and frequency information simultaneously. Therefore in this paper, fractional wavelet transform (FrWT) is proposed to be used for the first time for extracting the features of various datasets in a standard ECG database by combining the advantages of both fractional domain techniques and wavelets as case-II. Afterwards, Probabilistic Principal Component Analysis (PPCA) is used for detecting R-peaks for diagnosing heart abnormalities in various morphologies of the ECG signal. The proposed technique has been evaluated on the basis of sensitivity (SEN), detection error rate (DER), and positive predictivity (PPR) (of the detected ECG beats) for MIT-BIH Arrhythmia database (M/B Ar DB). Even though both FrFT and FrWT techniques exhibit a high degree of robustness, but SEN of 99.99%, DER of 0.026%, & PPR of 99.99% obtained by latter in case-II are better than SEN of 99.97%, DER of 0.053%, & PPR of 99.98% obtained by the former in case-I for M/B Ar DB. In this paper, average time error (ATE) is also obtained for the considered datasets establishing the effectiveness of the proposed technique further. These encouraging results suggest that the proposed methodology will go a long way in assisting the cardiologists to detect temporal patterns in a wide variety of electrophysiological cases, which is important for improved management of healthcare system.