The Objective Identification of Bruising in Children using Diffuse Reflectance Spectroscopy

PROJECT SUMMARY The objective of this study is to examine the validity of reflectance spectroscopy as a diagnostic tool for objectively diagnosing bruising in infants and children. In cases of suspected child physical abuse, bruising is a key clinical finding. The identification of bruising is extremely important, as it 1) triggers standard of care child

physical abuse medical evaluations in infants and young children; and 2) often forms the “reasonable suspicion” basis for making reports to child protective services. Yet, the current clinical assessment of bruising in children is subjective and imprecise. One difficulty is that bruising can be difficult to detect in children with darker skin tones, as melanin is a chromophore that can

obscure or distort characteristic color appearance of extravasated blood in the skin. Unfortunately, this clinical imprecision has contributed to racial disproportionality of child physical abuse evaluations and reporting. The current gold-standard of clinical bruising confirmation is time-dependent, i.e., serial evaluations of the lesion

that demonstrate color changes and disappearance of the lesion in a matter of weeks. In the suspected child physical abuse scenario, the time-dependent confirmation of bruising is problematic, as decisions on whether to undertake medical evaluations and/or make reports to child protective services require immediate, point-of-

care determinations. We propose that reflectance spectroscopy is an ideal candidate for this pediatric point-of-care tool because it is painless, non-invasive, quickly utilized, and potentially scalable for widespread use in smartphone cameras. The use of reflectance spectroscopy in other lines of clinical research—melanoma, neonatal

hyperbilirubinemia, and forensic dating of bruising--provide the scientific substrate for our novel and innovative application of this technology. We propose that, irrespective of skin tone, reflectance spectroscopy will demonstrate a reflectance pattern that is characteristic of traumatic bruising and is different from common

bruising mimics (i.e, birthmarks, vascular malformations, etc). Finally, we aim to utilize the patterns of hemoglobin breakdown products to create a predictive model that identifies a bruise with high sensitivity and specificity. We will utilize a multi-center approach to diversify our patient population and ensure sufficient patient

volumes, recruiting 260 children (ages 0-17-years) who have suspected bruising and common bruising mimics. Using a laboratory-grade spectrometer, we will obtain spectral readings of both affected and adjacent unaffected skin areas on all patients, controlling for skin tone. Spectral analysis will be performed using a least-

squares fitting method and Monte Carlo simulations will be used to help generate a predictive model. We believe that, ultimately, the point-of-care use of reflectance spectroscopy will diminish unnecessary healthcare utilization and racial disparities in child physical abuse evaluations and reporting.