Foot ulceration remains a serious health concern for diabetic patients and has a major impact on the cost of diabetes treatment. Early detection and preventive care, such as offloading or improved hygiene, can greatly reduce the risk of further complications. We aim to assess the use of hyperspectral tissue oximetry in predicting the risk of diabetic foot ulcer formation. Tissue oximetry measurements are performed during several visits with hyperspectral imaging of the feet in type 1 and 2 diabetes mellitus subjects that are at risk for foot ulceration. The data are retrospectively analyzed at 21 sites that ulcerated during the course of our study and an ulceration prediction index is developed. Then, an image processing algorithm based on this index is implemented. This algorithm is able to predict tissue at risk of ulceration with a sensitivity and specificity of 95 and 80%, respectively, for images taken, on average, 58 days before tissue damage is apparent to the naked eye. Receiver operating characteristic analysis is also performed to give a range of sensitivity/specificity values resulting in a Q-value of 89%.
Foot ulceration is a debilitating comorbidity of diabetes that may result in loss of mobility and amputation. Optical
detection of cutaneous tissue changes due to inflammation and necrosis at the preulcer site could constitute a preventative
strategy. A commercial hyperspectral oximetry system was used to measure tissue oxygenation on the feet of diabetic
patients. A previously developed predictive index was used to differentiate preulcer tissue from surrounding healthy tissue
with a sensitivity of 92% and specificity of 80%. To improve prediction accuracy, an optical skin model was developed
treating skin as a two-layer medium and explicitly accounting for (i) melanin content and thickness of the epidermis,
(ii) blood content and hemoglobin saturation of the dermis, and (iii) tissue scattering in both layers. Using this forward
model, an iterative inverse method was used to determine the skin properties from hyperspectral images of preulcerative
areas. The use of this information in lowering the false positive rate was discussed.
We present an evaluation of time-resolved fluorescence measurements on human skin for screening type 2 diabetes. In vivo human skin is excited with a pulse diode at 375 nm and pulse width of 700 ps. Fluorescence decays are recorded at four different emission wavelengths: 442, 460, 478, and 496 nm. Experiments are performed at various locations, including the palms, arms, legs, and cheeks of a healthy Caucasian subject to test single-subject variability. The fluorescence decays obtained are modeled using a three-exponential decay. The variations in the lifetimes and amplitudes from one location to another are minimal, except on the cheek. We compare the fluorescent decays of 38 diabetic subjects and 37 nondiabetic subjects, with different skin complexions and of ages ranging from 6 to 85 yr. The average lifetimes for nondiabetic subjects were 0.5, 2.6, and 9.2 ns with fractional amplitudes of 0.78, 0.18, and 0.03, respectively. The effects of average hemoglobin A1c (HbA1c) from the previous 4 yr and diabetes duration are evaluated. While no significant differences between the fluorescence lifetimes of nondiabetic and diabetic subjects are observed, two of the fractional amplitudes are statistically different. Additionally, none of the six fluorescence parameters correlated with diabetes duration or HbA1c. One of the lifetimes as well as two of the fractional amplitudes differ between diabetic subjects with foot ulcers and nondiabetic subjects.
This paper presents numerical simulations predicting the time-resolved reflectance and autofluorescence of human skin exposed to a pulse of collimated light at 337 nm and pulse width of 1 ns. Moreover, the feasibility of using an embedded time-resolved fluorescence sensor for monitoring glucose concentration is also studied. Skin is modeled as a multilayer medium with each layer having its own optical properties and fluorophore absorption coefficients, lifetimes and quantum yields. The intensity distributions of excitation and fluorescent light in skin are then determined by solving the transient radiative transfer equation using the modified method of characteristics. In both cases, the fluorophore lifetimes are recovered from the simulated fluorescence decays and compared with the actual lifetimes used in the simulations. It was found that the fluorescence lifetime of the fluorophore contributing the least to the fluorescence signal could not be recovered while the other lifetimes could be recovered within 2.5% of input values. Such simulations could be valuable in interpreting data from time-resolved fluorescence experiments on healthy and diseased tissue as well as in designing and testing the feasibility of various optical sensors for biomedical diagnostics.
In this paper we present preliminary results obtained from fluorescence lifetime measurements on human skin using time-correlated single photon counting (TCSPC) techniques. Human skin was exposed to light from a pulsed LED of 700 ps pulse width at a wavelength of 375 nm and fluorescence decays were recorded at four
different emission wavelengths (442, 460, 478 and 496 nm) using a photomultiplier tube (PMT) coupled to a monochromator. Measurements were carried out on the left and right palms of subjects recruited for the study after obtaining consent using a UCLA IRB approved consent form. The subjects recruited consisted of 18 males and 17 females with different skin complexions and ages ranging from 10 to 70 years. In addition, a set of experiments were also performed on various locations including the palm, the arm and the cheek of a Caucasian subject. The fluorescence decays thus obtained were fit to a three-exponential decay model in all cases and were approximately 0.4, 2.7 and 9.4 ns, respectively. The variations in these lifetimes with location, gender, skin complexion and age are studied. It is speculated that the shorter lifetimes correspond to free and bound NADH while the longer lifetime is due to AGE crosslinks.