Medical Engineering & Physics 33 (2011) 720–729 Contents lists available at ScienceDirect Medical Engineering & Physics j o ur nal homep age : www.elsevier.com/locate/medengphy Low resource processing algorithms for laser Doppler blood flow imaging Hoang C. Nguyen, Barrie R. Hayes-Gill, Yiqun Zhu, John A. Crowe, Diwei He, Stephen P. Morgan ∗ Electrical Systems and Optics Research Division, Faculty of Engineering, The University of Nottingham, University Park, Nottingham NG7 2RD, United Kingdom a r t i c l e i n f o Article history: Received 14 April 2010 Received in revised form 18 November 2010 Accepted 16 January 2011 Keywords: Laser Doppler blood flowmetry Signal processing Full field a b s t r a c t The emergence of full field laser Doppler blood flow imaging systems based on CMOS camera technology means that a large amount of data from each pixel in the image needs to be processed rapidly and system resources need to be used efficiently. Conventional processing algorithms that are utilized in single point or scanning systems are therefore not an ideal solution as they will consume too much system resource. Two processing algorithms that address this problem are described and efficiently implemented in a field programmable gate array. The algorithms are simple enough to use low system resource but effec- tive enough to produce accurate flow measurements. This enables the processing unit to be integrated entirely in an embedded system, such as in an application-specific integrated circuit. The first algorithm uses a short Fourier transformation length (typically 8) but averages the output multiple times (typically 128). The second method utilizes an infinite impulse response filter with a low number of filter coeffi- cients that operates in the time domain and has a frequency-weighted response. The algorithms compare favorably with the reference standard 1024 point fast Fourier transform in terms of both resource usage and accuracy. The number of data words per pixel that need to be stored for the algorithms is 1024 for the reference standard, 8 for the short length Fourier transform algorithm and 5 for the algorithm based on the infinite impulse response filter. Compared to the reference standard the error in the flow calculation is 1.3% for the short length Fourier transform algorithm and 0.7% for the algorithm based on the infinite impulse response filter. © 2011 IPEM. Published by Elsevier Ltd. All rights reserved. 1. Introduction The first reported laser Doppler blood flow (LDBF) measure- ments were made more than 30 years ago [1] establishing a field that has remained of interest [2–5]. Whilst early systems used single point measurements, and these are still valuable in many applications, the spatial variation of blood flow is also of interest [6]. Spatial solutions often use a single laser beam scanning over an area of interest to build up a flow image [7,8]; however, the acquisition time is relatively long due to mechanical scanning. An alternative approach, laser speckle contrast analysis (LASCA) uses a CCD camera to acquire speckle images [9]. There are some draw- backs as it does not provide a direct measurement of the temporal fluctuations that occur in blood flow imaging and there is a reduc- tion in spatial resolution as spatial averaging is performed over a sub-array (often ∼7 × 7 pixels). In recent years, with the development of complementary metal- oxide-semiconductor (CMOS) technology, several full-field LDBF cameras have been introduced [10–13]. These full-field cameras acquire data at several thousand raw data frames per second and ∗ Corresponding author. Tel.: +44 115 9515570. E-mail address: steve.morgan@nottingham.ac.uk (S.P. Morgan). display processed blood flow maps at typically 1–10 frames/s. This high data rate acquisition inevitably introduces data bottlenecks in either the data transmission or the processing. Therefore, an efficient data acquisition and processing scheme needs to be intro- duced to make both high resolution and high frame rate LDBF imaging feasible. Most LDBF systems estimate blood flow via the first moment of the Doppler shifted signal captured by a photodetector as [14]: Flow =  ω 2 ω 1 ωP(ω) dω, (1) where P(ω) is the power spectrum of the Doppler-shifted signal, and ω 1 and ω 2 are the limits of the Doppler frequency range, which are usually chosen greater than 0 Hz (typically ∼200 Hz for both scanning and full-field systems) and up to 20 kHz respectively [2]. The algorithm that is defined as the reference standard in this paper uses Eq. (1) as its basis. The data is transformed to the frequency domain and then the weighted frequency components are accumulated. For a full-field application, this calculation is performed digitally using processors such as a digital signal pro- cessor, a field-programmable gate array (FPGA) [15] or a computer [10,12,13]. The implementation that uses a computer is convenient in terms of system setup and high accuracy, but it takes a consid- erable time for data transmission and processing. For example, the 1350-4533/$ – see front matter © 2011 IPEM. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.medengphy.2011.01.009