Biomedical Signal Processing and Control 5 (2010) 131–141 Contents lists available at ScienceDirect Biomedical Signal Processing and Control journal homepage: www.elsevier.com/locate/bspc An algorithm for FHR estimation from foetal phonocardiographic signals M. Ruffo, M. Cesarelli, M. Romano, P. Bifulco ∗ , A. Fratini Department of Biomedical, Electronic and Telecommunication Engineering, University “Federico II”, via Claudio 21, 80125 Naples, Italy article info Article history: Received 13 October 2009 Received in revised form 3 February 2010 Accepted 5 February 2010 Available online 28 March 2010 Keywords: Passive foetal monitoring Foetal phonocardiography First foetal heart sound detection Signal processing abstract The long-term foetal surveillance is often to be recommended. Hence, the fully non-invasive acoustic recording, through maternal abdomen, represents a valuable alternative to the ultrasonic cardiotocogra- phy. Unfortunately, the recorded heart sound signal is heavily loaded by noise, thus the determination of the foetal heart rate raises serious signal processing issues. In this paper, we present a new algorithm for foetal heart rate estimation from foetal phonocardiographic recordings. A filtering is employed as a first step of the algorithm to reduce the background noise. A block for first heart sounds enhancing is then used to further reduce other components of foetal heart sound signals. A complex logic block, guided by a number of rules concerning foetal heart beat regularity, is proposed as a successive block, for the detection of most probable first heart sounds from several candidates. A final block is used for exact first heart sound timing and in turn foetal heart rate estimation. Filtering and enhancing blocks are actually implemented by means of different techniques, so that different processing paths are proposed. Fur- thermore, a reliability index is introduced to quantify the consistency of the estimated foetal heart rate and, based on statistic parameters; [,] a software quality index is designed to indicate the most reliable analysis procedure (that is, combining the best processing path and the most accurate time mark of the first heart sound, provides the lowest estimation errors). The algorithm performances have been tested on phonocardiographic signals recorded in a local gynaecology private practice from a sample group of about 50 pregnant women. Phonocardiographic signals have been recorded simultaneously to ultrasonic cardiotocographic signals in order to compare the two foetal heart rate series (the one estimated by our algorithm and the other provided by cardiotocographic device). Our results show that the proposed algo- rithm, in particular some analysis procedures, provides reliable foetal heart rate signals, very close to the reference cardiotocographic recordings. © 2010 Elsevier Ltd. All rights reserved. Abbreviations: %H, % of high RI values; %L, % of low RI values; %M, % of medium RI values; ADC, analog to digital converter; AP, analysis procedure; CTG, cardiotocography/cardiotocographic signal; DS, differences series (bpm, differences of corresponding samples of FHR–FHS and FHR-CTG series); FD, fiducial degree (of the detected beats); FECG, foetal electrocardiography; FHR, foetal heart rate; FHR- CTG, FHR recorded by means of CTG; FHR–FHS, FHR estimated from FHS by the developed algorithm; FHS, foetal heart sounds; FIR, finite impulse response; HT, high amplitude threshold (50% of mean value of maximum amplitudes of the pre- vious 8 recognized beats); LP, output of B3 blocks (envelope or low-pass filtering of B3 outputs); LT, low amplitude threshold (30% of mean value of maximum ampli- tudes of the previous 8 recognized beats); M, LP mean value within the time interval T;M-, fifth order median value computed on the five previous FHR samples (with respect to the sample in analysis); M+, fifth order median value computed on the five following FHR samples (with respect to the sample in analysis); M1, time mark of S1: maximum of the selected LP peak; M2, time mark of S1: centre of gravity of the rectified signal tract of filtering block output, corresponding to the selected LP peak; M3, time mark of S1: maximum of the central peak of the rectified signal tract of filtering block output, corresponding to the selected LP peak; ME, mean value of the mean absolute values of all DS; MEAN, inter-distance mean value computed on 8 beats previous the beat in analysis; PCG, phonocardiography/phonocardiographic signal; Pi, positions of probable beats in the time interval T; PP, processing paths; PRI, linear combination of %H, %M and %L; PSD, power spectral density; QS, quality 1. Introduction In industrialized countries, all pregnant women periodically take pregnancy and foetal well-being controls. The most impor- tant aim of foetal surveillance is to avoid intrauterine death or permanent damages to the foetus. Hence, to collect reliable infor- mation about the foetus’ health, also to correctly and promptly plan successive diagnostic tests, is very important. Monitoring the variations in foetal heart rate (FHR) provides up-to-date information about the general foetus’ well-being. In the last trimester of pregnancy, the widespread diagnostic tool is signal (of FHS signal tract); RI, reliability index (associated to the FHR values); RMS, root mean square; RR, series of beat to beat inter-distances; S1, first foetal heart sound; S2, second foetal heart sound; SD, mean value of the standard deviation of all DS; SNR, signal to noise ratio; SQI, software quality index; T, current time inter- val of analysis (equal to T0 + 0.65 × MEAN, T0 + 1.35 × MEAN); T0, position of the last recognized beat; TEO, Teager energy operator; TH, local amplitude threshold (equal to 1.2 × M); UC, uterine contractions. ∗ Corresponding author. Tel.: +39 0817683794; fax: +39 0817683804. E-mail address: pabifulc@unina.it (P. Bifulco). 1746-8094/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.bspc.2010.02.002