Page 18 - PR 2014 2016 02 Lasers Technology

32 Lasers Technology | Progress Report

to breathing movements. In attempt to cor-
rect the baseline of a signal with peaks, the
background signal of each register was esti-
mated using the MATLAB function msbackadj.
Baseline tracking and correction is a common
problem in DNA sequencing. Automated DNA
sequencing is a well established technique.
In this case, the accurate identification of a
DNA sequence depends on the correction of
the baseline of the chromatographic signal.
The raw chromatogram generally presents a
Figure 14: Typical mapping of blood flow from a C57BL/6
mouse produced via Laser speckle contrast imaging. Isch- slowly varying baseline. For DNA sequencing,
emic hindlimb in region (1) and control in region (2). Color the correction of the baseline is necessary in
code: blue meaning low flow to red meaning high flow.
order to establish a trustable reference of the
background signal to further processing. In
cular function. Briefly, a low power (typically this case, the corrected data is the raw data
50 mW), non-collimated laser radiation, fre- minus the estimated baseline signal. Here
quently in the range from 635 nm to 780 nm, (blood flow measurement, the estimated base-
illuminates the skin. Scattered photons from line is the useful signal. The raw data are used
static and moving (mainly red) blood cells only to estimate the background signal. The
in the microvascular plexus are collected by measured SBF signal from mice hindlimbs
a CCD or a CMOS camera during a selected via LSCI, Fig. 14, is corrupted by respiratory
exposure time. The captured image presents movements. The recovery of a SBF signal cor-
a granular pattern, known as speckle pattern, rupted by artefacts from breathing is feasible,
due to interferences of the scattered radia- allowing more accurate measurements, Fig. 15.
tion. The pattern changes with the movement
of the red blood cells. The differences in the 80
statistics of speckle patterns from static and 70
moving structures are explored to compute a
60
quantity related to the skin blood flow (SBF). B
A sequence of images is captured , and each 50 A
frame is processed producing a mapping of Flux (AU) 40
blood flow in the investigated region. A region
30
of interest (ROI) can be selected to compute a
regional (into the ROI) blood flow. The result 20
is a time series of SBF values. This spatial pro- 10
cessing methodology allows high sampling 99 99.5 100 100.5 101 101.5 102
rate (typically in the order of 25 frames/s), Time (s)
detecting rapid changes of flow. As a result Figure 15: Raw register of blood flow from a control hindlimb during
of the non-contact feature of the LSCI system, five minutes (blue), the corresponding mean value (red) and the
estimated background signal (green). The measured flow is corrupt-
any body movement of the target generally ed by respiratory movements: (A) inspiratory and (B) expiratory.
interferes in the measurement. This fact may
restrict the use of the LSCI system to monitor As a facility that has access to radioactive and
the blood flow from anesthetized mice due nuclear materials, research and development

Instituto de Pesquisas Energéticas e Nucleares