Objectification of Facial Color Inspection to Differentiate Obstructive/nonobstructive Jaundice in Neonates by Spectrophotometer

Zhen Shen,Shan Zheng,Rui Dong,Gong Chen
DOI: https://doi.org/10.1016/j.jpedsurg.2017.08.069
IF: 2.549
2017-01-01
Journal of Pediatric Surgery
Abstract:Purpose: The purpose of this study was to study whether color difference in facial color truly exists between neonates with obstructive and nonobstructive jaundice, and whether the color difference could be objectified by spectrophotometer. Methods: Twelve biliary atresia patients were enrolled in an obstructive jaundice group and 15 neonates admitted for non-conjugated hyperbilirubinemia in a nonobstructive group. Nine patients with syphilis (n = 6) and sacrococcygeal teratoma (n = 3) were studied as control. Transcutaneous total bilirubin (TB) and hemoglobin were recorded. Face color was measured by spectrophotometer. Spectral reflection curve and L*a*b* model parameters were studied. Results: Facial color of jaundiced neonates were characteristic in waveform that reflectivity at wavelength of 550 nm was significantly decreased compared with control by 16.4 +/- 3.4%, while not significantly different between obstructive and nonobstructive jaundice (p = 0.124). At 650 nm, reflection in nonobstructive jaundice was decreased by 8.4 +/- 2.3% (p < 0.01), and reflection in obstructive jaundice was (9.6 +/- 3.2) % lower compared with nonobstructive jaundice (p < 0.01). In L*a*b* model, mean color difference between obstructive and nonobstructive jaundice was 9.60. L* was significantly different: control (71.84 +/- 3.75%) > obstructive jaundice (58.09 +/- 1.25%) N nonobstructive jaundice (54.25 +/- 7.27%). Value b* was higher in jaundiced patients compared to normal control (11.88 +/- 2.16, p < 0.001), but not significantly different in obstructive and nonobstructive jaundice (20.12 +/- 2.17vs18.25 +/- 4.58). Value a* was not significantly different between normal control (5.57 +/- 2.38) and obstructive jaundice (5.25 +/- 1.19), but was lower than nonobstructive jaundice (14.03 +/- 3.29, p < 0.001). TB was significantly correlated with b* (R=0.526, p=0.014), while hemoglobin was correlated with a* (R=0.791, p < .001) and L* (R=-0.707, p < .001). Conclusions: Obstructive and nonobstructive jaundice could be objectively differentiated through facial color inspection by spectrophotometer. (C) 2017 Elsevier Inc. All rights reserved.
What problem does this paper attempt to address?