Quantitative assessments of burn degree by high-frequency ultrasonic backscattering and statistical model

Phys Med Biol. 2011 Feb 7;56(3):757-73. doi: 10.1088/0031-9155/56/3/014. Epub 2011 Jan 14.

Abstract

An accurate and quantitative modality to assess the burn degree is crucial for determining further treatments to be properly applied to burn injury patients. Ultrasounds with frequencies higher than 20 MHz have been applied to dermatological diagnosis due to its high resolution and noninvasive capability. Yet, it is still lacking a substantial means to sensitively correlate the burn degree and ultrasonic measurements quantitatively. Thus, a 50 MHz ultrasound system was developed and implemented to measure ultrasonic signals backscattered from the burned skin tissues. Various burn degrees were achieved by placing a 100 °C brass plate onto the dorsal skins of anesthetized rats for various durations ranged from 5 to 20 s. The burn degrees were correlated with ultrasonic parameters, including integrated backscatter (IB) and Nakagami parameter (m) calculated from ultrasonic signals acquired from the burned tissues of a 5 × 1.4 mm (width × depth) area. Results demonstrated that both IB and m decreased exponentially with the increase of burn degree. Specifically, an IB of -79.0 ± 2.4 (mean ± standard deviation) dB for normal skin tissues tended to decrease to -94.0 ± 1.3 dB for those burned for 20 s, while the corresponding Nakagami parameters tended to decrease from 0.76 ± 0.08 to 0.45 ± 0.04. The variation of both IB and m was partially associated with the change of properties of collagen fibers from the burned tissues verified by samples of tissue histological sections. Particularly, the m parameter may be more sensitive to differentiate burned skin due to the fact that it has a greater rate of change with respect to different burn durations. These ultrasonic parameters in conjunction with high-frequency B-mode and Nakagami images could have the potential to assess the burn degree quantitatively.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Burns / diagnostic imaging*
  • Burns / pathology*
  • Female
  • Models, Statistical*
  • Rats
  • Rats, Wistar
  • Skin / diagnostic imaging
  • Skin / pathology
  • Ultrasonography / methods*