Skin Research and Technology 2010; 16: 397–400
Printed in Singapore  All rights reserved
doi: 10.1111/j.1600-0846.2010.00450.x
Diffuse reflectance spectrophotometry for skin
phototype determination
F. J. González1, M. Martı́nez-Escanamé2, R. I. Muñoz1, B. Torres-Álvarez2 and B. Moncada2
1
Instituto de Investigación en Comunicación Óptica, Universidad Autónoma de San Luis Potosı́, SLP, México and 2Dermatology Department, Hospital
Central ‘Dr Ignacio Morones Prieto’, Universidad Autónoma de San Luis Potosı́, SLP, México
Background: The Fitzpatrick skin phototype classification
scheme has become the standard method for assessing the
reaction of the skin to solar stimuli; this method can be
easily biased by different factors, such as ethnicity or
chronic sun exposure.
Methods: Diffuse reflectance spectrophotometry (DRS) is
an objective and non-invasive method used in this work to
determine constitutive skin color from the upper volar arm as
an objective way of measuring skin pigmentation. A DRS-
determined melanin index that accounts for skin pigmenta-
tion was obtained for 35 subjects of Hispanic origin, this
melanin index was compared with the physician-diagnosed
and self-reported skin phototypes.
T HE FITZPATRICK classification scheme is a sub-
jective way of assessing the reaction of the
skin to solar stimuli, this scheme consists of a six-
integer Roman numeral scales I–VI and has
become the standard for determining human
skin phototypes (1). This scale has been useful
in assessing risk factors for both melanoma and
non-melanoma skin cancer as well as for estimat-
ing UV, PUVA, and laser treatment doses (2).
The Fitzpatrick classification scheme can be
easily biased by different factors, such as ethni-
city or chronic sun exposure, which might induce
errors in the subject’s susceptibility to skin cancer
or in determining the required dose for laser
treatment (3).
Diffuse reflectance spectrophotometry (DRS),
which is used for the detection of light that has
been scattered multiple times within the sample
of interest, has been widely employed for the
evaluation of melanin and hemoglobin concen-
trations in skin tissue and for the evaluation of
pigmented lesions (4).
In this work, a melanin index obtained from
DRS is used to determine constitutive skin pig-
mentation from the upper volar arm (3), these
r 2010 John Wiley & Sons A/S
Skin Research and Technology
Results: The results show that at least for Hispanic indivi-
duals, there is a clear clinical distinction between subjects
with skin phototype I and their DRS-determined melanin
index; however, subjects with skin phototypes II–VI have a
large melanin index overlap.
Conclusion: Clinical assessment of skin phototype can be
complemented by using DRS.
Key words: diffuse reflectance spectrophotometry – skin
phototype – melanin content – melanin index
& 2010 John Wiley & Sons A/S
Accepted for publication 23 January 2010
results were compared with the clinically as-
sessed skin phototype for each subject. Using
this melanin index, an objective phototype clas-
sification scheme based on DRS is proposed.
Methods
Thirty-five subjects of Hispanic origin, aged 18–
40 years, with clinically assessed skin phototypes
ranging from I to VI were included in the study
(six subjects with phototype I, six with phototype
II, nine with phototype III, nine with phototype
IV, three with phototype V and one with photo-
type VI). Informed consent was obtained from all
participants, and the study was approved by the
local ethics committee.
The evaluation of the skin phototype was
realized by board-certified dermatologists using
the Fitzpatrick scale, and the participants also
completed a self-reported skin phototype.
DRS was performed on the upper volar arm
region (5 cm below the armpit) in order to obtain
the constitutive skin color of the subjects, measure-
ments at this site have proven to be as effective
as unexposed buttock skin for constitutive skin
397
González et al.
determination purposes (3). These measurements
were made using a USB4000-VIS-NIR spectro-
meter (Ocean Optics, Dunedin, FL, USA) with an
optical resolution of approximately 1.5 nm (full-
width at half-maximum), an LS-1 tungsten-halo-
gen light source (Ocean Optics) and an R200-7-
VIS-NIR reflection probe (Ocean Optics).
The raw reflectance spectra was corrected for
detector dark current and normalized to the
spectrum obtained from the light source reflected
on a white reference standard according to the
following equation:
Smeas ðlÞ  DðlÞ
RðlÞ ¼
Sref ðlÞ  DðlÞ
where Smeas(l) is the raw reflection data, D(l) is
the detector dark current and Sref(l) is the spec-
trum obtained using a white reflectance standard
(5). The white reference standard used was a
certified reflectance standard WS-1-SL (Lab-
sphere, North Sutton, NH, USA) made of Spec-
tralont. These dark current and reference
measurements were performed before each skin
reflectance measurement.
Figure 1 shows the typical reflectance spectra
for three subjects with clinically assessed photo-
types I, III and VI. These spectra show how
darker skin reflect less light than lighter skin,
and how the absorption bands of hemoglobin in
the spectral regions of 530–590 nm (q-bands) and
410–440 nm (Soret bands, at 450 nm for oxyhemo-
globin and at 430 nm for deoxyhemoglobin) are
more evident in the spectra of less pigmented
skin (6).
In order to obtain a melanin concentration
index, the apparent absorbance spectrum, de-
fined as the negative of the logarithm base 10 of
Fig. 1. Raw reflectance spectra of three subjects with clinically
assessed phototypes I, III and VI.
398
the reflectance, was calculated. The melanin con-
tribution to the diffuse reflectance spectrum
can be approximated by fitting a straight line
through the points from 620 to 720 nm of the
absorbance spectrum (7); however, the contribu-
tion of deoxyhemoglobin (deoxy-Hb) can be sig-
nificant at those wavelengths and should be
subtracted (8).
In order to account for the deoxy-Hb contribu-
tion, the method presented by Stamatas et al. (6)
was used. In this method, a corrected absorbance
spectrum is calculated by subtracting an initial
estimation for the melanin concentration. The
deoxy-Hb and oxy-Hb concentrations were ob-
ð1Þ
tained by solving a system of two equations and
two unknowns. This system of equations was
obtained calculating the apparent concentrations
of deoxy-Hb and oxy-Hb from the DRS-spectra
and the extinction coefficients of deoxy-Hb and
oxy-Hb in the 560–580 nm range (6).
Once the deoxy-Hb contribution has been cal-
culated, it can be subtracted from the original
absorption spectrum, and the corrected melanin
contribution can then be calculated by fitting a
straight line through the points from 620 to
720 nm. A melanin index was obtained by inte-
grating this linear equation through the 620–
720 nm interval and dividing by 100 nm, the
spectral range, the index was later multiplied
by 100 in order to make it easier to quantify.
Figure 2 shows the absorbance spectra, cor-
rected for the deoxy-Hb contribution, for three
subjects with clinically assessed phototypes I, III
and VI, along with the interval where the inte-
gration was made in order to obtain the melanin
index. It can be seen from Fig. 2 how darker skin
will result in a higher melanin index than lighter
skin.
Fig. 2. Absorbance spectra of three subjects with clinically assessed
phototypes I, III and VI and the interval used to obtain the melanin
index.
 Reflectance spectrophotometry for skin phototyping

Fig. 4. Melanin index obtained for all subjects and their respective
diffuse reflectance spectrophotometry (DRS)-determined phototype.
Fig. 3. Melanin index obtained for 35 Hispanic subjects as a function
of their clinically assigned phototypes.

TABLE 1. Percentage of overlap of the clinically assessed phototype and

the diffuse reflectance spectrophotometry (DRS)-determined phototype
Results
1 2 3 4 5 6
Figure 3 shows the averaged melanin index and
I 100% 0 0 0 0 0
the standard deviation obtained for 35 Hispanic
II 0 18% 64% 18% 0 0
subjects as a function of their clinically assigned III 0 10% 50% 40% 0 0
phototypes. From the figure it can be seen how IV 0 0 33% 44% 23% 0
there is a significant overlap in the melanin index V 0 0 0 33% 33% 33%
VI 0 0 0 0 100% 0
obtained from DRS of subjects with clinically
assigned phototypes II–VI. It is worth noting
that at least for Hispanic individuals, there is a
clear clinical distinction between subjects with
skin phototype I and their DRS determined mel-
Table 1 shows the percentage of overlap of
anin index; however, subjects with skin photo-
the clinically assessed phototype based on the
types II–VI have a large melanin index overlap.
Fitzpatrick classification and an arbitrary classi-
Figure 4 shows the melanin index obtained for
fication based on the DRS-determined melanin
every subject as a function of the clinically
index. From Table 1 it can be seen how a single
assessed phototype. An arbitrary phototype clas-
clinically assessed phototype can fall into one
sification was developed based on the melanin
of three DRS-determined phototypes. This indi-
index; this classification was made assigning a
cates that skin phototyping can be improved by
phototype I to subjects with a melanin index
using a melanin index based on DRS measure-
lower than 40, a phototype VI to subjects with a
ments that are objective and free from observer
melanin index higher than 70 and dividing
bias.
equally the 30 melanin index units in four photo-
It is worth noting that a reflectance setup and a
types. From Fig. 4 it can be seen how for a DRS-
laptop or personal computer can process the
determined phototype IV, for example, resulted
diffuse reflectance instantly and provide with a
in clinically assessed phototypes II–V.
melanin index in a matter of seconds; this could
be used in clinical practice in the assessment of
risk factors for both melanoma and non-mela-
Conclusions noma skin cancer as well as for estimating UV,
A melanin index was determined using DRS by PUVA and laser treatment doses. Although mel-
integrating the absorbance spectra, corrected for anin content is one of the main factors that
deoxy-Hb, over the 620–720 nm interval. This determines one person’s ability to tan, further
melanin index was calculated for 35 subjects of work should be done to relate the DRS-deter-
Hispanic origin. mined phototype with the ability to tan.

399
González et al.

Acknowledgements 6. Stamatas GN, Zmudzka BZ, Kollias N, Beer JZ. In vivo

measurement of skin erythema and pigmentation: new
F. J. González acknowledges support by CONACyT means of implementation of diffuse reflectance spectro-
and FOMIX-SLP through grants CB-2006-60349 and scopy with a commercial instrument. Br J Dermatol
2008; 159: 683–690.
FMSLP-2008-C01-87127, respectively. 7. Kollias N, Baqer A. Spectroscopic characteristics of
human melanin in vivo. J Invest Dermatol 1985; 85:
38–42.
References 8. Stamatas GN, Kollias N. Blood stasis contributions to the
perception of skin pigmentation. J Biomed Opt 2004; 9:
1. Fitzpatrick TB. The validity and practicality of sun- 315–322.
reactive skin type-I through type-VI. Arch Dermatol
1988; 124: 869–871.
2. Astner S, Anderson RR. Skin phototypes 2003. J Invest Address:
Dermatol 2004; 122: xxx–xxxi. F. J. González
3. Pershing LK, Tirumala VP, Nelson JL, Corlett JL, Lin AG, Instituto de Investigación en Comunicación Óptica
Meyer LJ, Leachman SA. Reflectance spectrophotometer: Universidad Autónoma de San Luis Potosı́
the dermatologists’ sphygmomanometer for skin photo- Álvaro Obregón #64
typing? J Invest Dermatol 2008; 128: 1633–1640. Col. Centro, CP 78000
4. Nishidate I, Sasaoka K, Yuasa T, Niizeki K, Maeda T, San Luis Potosi, SLP
Aizu Y. Visualizing of skin chromophore concentrations Mexico
by use of RGB images. Opt Lett 2008; 33: 2263–2265. Tel:152 444 825 0183 ext 232
5. González FJ. Reflectance spectrophotometry in clinical Fax:152 444 825 0198
dermatology. J Invest Dermatol 2009; 129: 1582–1583. e-mail: javier.gonzalez@uaslp.mx

400