Applicability of Facial De-Identification Methods for Privacy Protection in Dermatology: Systematic Review (Preprint)

BACKGROUND De-identifying facial images is critical for protecting patient anonymity in the era of increasing tools for automatic image analysis in dermatology. OBJECTIVE The purpose of this paper was to review the current literature in the field of automatic facial de-identification algorithms. METHODS We conducted a systematic search using a combination of headings and keywords to encompass the concepts of facial de-identification and privacy preservation. The databases MEDLINE (via Pubmed), Embase (via Elsevier) and Web of Science (via Clarivate) were queried from inception to 5/1/2021. Studies of wrong design and outcomes were excluded during the screening and review process. RESULTS A total of 18 studies were included in the final review reporting various methodologies of facial de-identification algorithms. The study methods were rated individually for their utility for use cases in dermatology pertaining to skin color/pigmentation and texture preservation, data utility, and human detection. Most studies notable in the literature address feature preservation while sacrificing skin color and texture. CONCLUSIONS Facial de-identification algorithms are sparse and inadequate to preserve both facial features and skin pigmentation/texture quality in facial photographs. A novel approach is needed to ensure greater patient anonymity, while increasing data access for automated image analysis in dermatology for improved patient care.

Download Full-text

Skin Color Correction for Tissue Spectroscopy: Demonstration of a Novel Approach with Tissue-Mimicking Phantoms

Applied Spectroscopy ◽

10.1366/0003702053085151 ◽

2005 ◽

Vol 59 (2) ◽

pp. 237-244 ◽

Cited By ~ 9

Author(s):

Olusola O. Soyemi ◽

Michelle R. Landry ◽

Ye Yang ◽

Patrick O. Idwasi ◽

Babs R. Soller

Keyword(s):

Skin Color ◽

Near Infrared ◽

Skin Pigmentation ◽

Nir Spectroscopy ◽

Bottom Layer ◽

Correction Method ◽

Color Variation ◽

Absorption Properties ◽

Novel Approach ◽

Components Analysis

The application of partial least squares (PLS) regression to visible–near-infrared (VIS-NIR) spectroscopy for modeling important blood and tissue parameters is generally complicated by the variation in skin pigmentation (melanin) across the human population. An orthogonal correction method for removing the influence of skin pigmentation has been demonstrated in diffuse reflectance spectra from two-layer tissue-mimicking phantoms. The absorption properties of the phantoms were defined by lyophilized human hemoglobin (bottom layer) and synthetic melanin (top layer). Tissue-like scattering was simulated in both layers with intralipid™. The approach uses principal components analysis (PCA) loading vectors from a separate set of phantom spectra that encode the unwanted melanin variation to remove the effect of melanin from the test phantoms. The preprocessing of phantom spectra using this orthogonal correction method resulted in PLS models with reduced complexity and enhanced prediction performance. Preliminary results from a separate study that evaluates the feasibility of defining skin color variation in an experiment with a single human subject are also presented.

Download Full-text

Privacy Protection With Facial Deidentification Machine Learning Methods: Can Current Methods Be Applied to Dermatology?

Iproceedings ◽

10.2196/35431 ◽

2021 ◽

Vol 6 (1) ◽

pp. e35431

Author(s):

Hyeon Ki Jeong ◽

Christine Park ◽

Ricardo Henao ◽

Meenal Kheterpal

Keyword(s):

Machine Learning ◽

Image Analysis ◽

Skin Color ◽

Image Data ◽

Facial Feature ◽

Human Detection ◽

Use Cases ◽

Facial Image ◽

High Quality ◽

Data Utility

Background In the era of increasing tools for automatic image analysis in dermatology, new machine learning models require high-quality image data sets. Facial image data are needed for developing models to evaluate attributes such as redness (acne and rosacea models), texture (wrinkles and aging models), pigmentation (melasma, seborrheic keratoses, aging, and postinflammatory hyperpigmentation), and skin lesions. Deidentifying facial images is critical for protecting patient anonymity. Traditionally, journals have required facial feature concealment typically covering the eyes, but these guidelines are largely insufficient to meet ethical and legal guidelines of the Health Insurance Portability and Accountability Act for patient privacy. Currently, facial feature deidentification is a challenging task given lack of expert consensus and lack of testing infrastructure for adequate automatic and manual facial image detection. Objective This study aimed to review the current literature on automatic facial deidentification algorithms and to assess their utility in dermatology use cases, defined by preservation of skin attributes (redness, texture, pigmentation, and lesions) and data utility. Methods We conducted a systematic search using a combination of headings and keywords to encompass the concepts of facial deidentification and privacy preservation. The MEDLINE (via PubMed), Embase (via Elsevier), and Web of Science (via Clarivate) databases were queried from inception to May 1, 2021. Studies with the incorrect design and outcomes were excluded during the screening and review process. Results A total of 18 studies, largely focusing on general adversarial network (GANs), were included in the final review reporting various methodologies of facial deidentification algorithms for still and video images. GAN-based studies were included owing to the algorithm’s capacity to generate high-quality, realistic images. Study methods were rated individually for their utility for use cases in dermatology, pertaining to skin color or pigmentation and texture preservation, data utility, and human detection, by 3 human reviewers. We found that most studies notable in the literature address facial feature and expression preservation while sacrificing skin color, texture, pigmentation, which are critical features in dermatology-related data utility. Conclusions Overall, facial deidentification algorithms have made notable advances such as disentanglement and face swapping techniques, while producing realistic faces for protecting privacy. However, they are sparse and currently not suitable for complete preservation of skin texture, color, and pigmentation quality in facial photographs. Using the current advances in artificial intelligence for facial deidentification summarized herein, a novel approach is needed to ensure greater patient anonymity, while increasing data access for automated image analysis in dermatology. Conflicts of Interest None declared.

Download Full-text

Privacy Protection With Facial Deidentification Machine Learning Methods: Can Current Methods Be Applied to Dermatology? (Preprint)

10.2196/preprints.35431 ◽

2021 ◽

Author(s):

Hyeon Ki Jeong ◽

Christine Park ◽

Ricardo Henao ◽

Meenal Kheterpal

Keyword(s):

Machine Learning ◽

Image Analysis ◽

Skin Color ◽

Image Data ◽

Facial Feature ◽

Human Detection ◽

Use Cases ◽

Facial Image ◽

High Quality ◽

Data Utility

BACKGROUND In the era of increasing tools for automatic image analysis in dermatology, new machine learning models require high-quality image data sets. Facial image data are needed for developing models to evaluate attributes such as redness (acne and rosacea models), texture (wrinkles and aging models), pigmentation (melasma, seborrheic keratoses, aging, and postinflammatory hyperpigmentation), and skin lesions. Deidentifying facial images is critical for protecting patient anonymity. Traditionally, journals have required facial feature concealment typically covering the eyes, but these guidelines are largely insufficient to meet ethical and legal guidelines of the Health Insurance Portability and Accountability Act for patient privacy. Currently, facial feature deidentification is a challenging task given lack of expert consensus and lack of testing infrastructure for adequate automatic and manual facial image detection. OBJECTIVE This study aimed to review the current literature on automatic facial deidentification algorithms and to assess their utility in dermatology use cases, defined by preservation of skin attributes (redness, texture, pigmentation, and lesions) and data utility. METHODS We conducted a systematic search using a combination of headings and keywords to encompass the concepts of facial deidentification and privacy preservation. The MEDLINE (via PubMed), Embase (via Elsevier), and Web of Science (via Clarivate) databases were queried from inception to May 1, 2021. Studies with the incorrect design and outcomes were excluded during the screening and review process. RESULTS A total of 18 studies, largely focusing on general adversarial network (GANs), were included in the final review reporting various methodologies of facial deidentification algorithms for still and video images. GAN-based studies were included owing to the algorithm’s capacity to generate high-quality, realistic images. Study methods were rated individually for their utility for use cases in dermatology, pertaining to skin color or pigmentation and texture preservation, data utility, and human detection, by 3 human reviewers. We found that most studies notable in the literature address facial feature and expression preservation while sacrificing skin color, texture, pigmentation, which are critical features in dermatology-related data utility. CONCLUSIONS Overall, facial deidentification algorithms have made notable advances such as disentanglement and face swapping techniques, while producing realistic faces for protecting privacy. However, they are sparse and currently not suitable for complete preservation of skin texture, color, and pigmentation quality in facial photographs. Using the current advances in artificial intelligence for facial deidentification summarized herein, a novel approach is needed to ensure greater patient anonymity, while increasing data access for automated image analysis in dermatology.

Download Full-text

Automated Image Analysis Of Nuclear And Cytoplasmic Changes In Exfoliated Cells From Tracheas Treated With Carcinogen

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100078638 ◽

1977 ◽

Vol 35 ◽

pp. 220-221

Author(s):

S.F. Stinson ◽

J.C. Lilga ◽

M.B. Sporn

Keyword(s):

Image Analysis ◽

Pattern Analysis ◽

Relative Proportion ◽

Automated Image Analysis ◽

Image Analyzer ◽

Cross Sectional ◽

Exfoliated Cells ◽

Neoplastic Lesions ◽

Analysis System ◽

Morphologic Criterion

Increased nuclear size, resulting in an increase in the relative proportion of nuclear to cytoplasmic sizes, is an important morphologic criterion for the evaluation of neoplastic and pre-neoplastic cells. This paper describes investigations into the suitability of automated image analysis for quantitating changes in nuclear and cytoplasmic cross-sectional areas in exfoliated cells from tracheas treated with carcinogen.Neoplastic and pre-neoplastic lesions were induced in the tracheas of Syrian hamsters with the carcinogen N-methyl-N-nitrosourea. Cytology samples were collected intra-tracheally with a specially designed catheter (1) and stained by a modified Papanicolaou technique. Three cytology specimens were selected from animals with normal tracheas, 3 from animals with dysplastic changes, and 3 from animals with epidermoid carcinoma. One hundred randomly selected cells on each slide were analyzed with a Bausch and Lomb Pattern Analysis System automated image analyzer.

Download Full-text

Progress In The Practical Application Of Automated Image Analysis To Tem Negatives

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100078614 ◽

1977 ◽

Vol 35 ◽

pp. 214-217

Author(s):

F. A. Heckman ◽

E. Redman ◽

J.E. Connolly

Keyword(s):

Image Analysis ◽

Image Quality ◽

Exposure Time ◽

Image Contrast ◽

Automated Image Analysis ◽

Practical Application ◽

Factors Affecting ◽

High Image Quality ◽

Qualitative Evidence ◽

Comprehensive Study

In our initial publication on this subject1) we reported results demonstrating that contrast is the most important factor in producing the high image quality required for reliable image analysis. We also listed the factors which enhance contrast in order of the experimentally determined magnitude of their effect. The two most powerful factors affecting image contrast attainable with sheet film are beam intensity and KV. At that time we had only qualitative evidence for the ranking of enhancing factors. Later we carried out the densitometric measurements which led to the results outlined below.Meaningful evaluations of the cause-effect relationships among the considerable number of variables in preparing EM negatives depend on doing things in a systematic way, varying only one parameter at a time. Unless otherwise noted, we adhered to the following procedure evolved during our comprehensive study:Philips EM-300; 30μ objective aperature; magnification 7000- 12000X, exposure time 1 second, anti-contamination device operating.

Download Full-text

Interfacing of a TEM/STEM System to a Computer

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100097818 ◽

1981 ◽

Vol 39 ◽

pp. 250-251

Author(s):

P. Hagemann

Keyword(s):

Image Analysis ◽

Transmission Rate ◽

Analytical Electron Microscopy ◽

Signal Beam ◽

Automated Image Analysis ◽

Beam Deflection ◽

External Control ◽

Computer Input ◽

Instrument Control ◽

Data Acquisition And Processing

The use of computers in the analytical electron microscopy today shows three different trends (1) automated image analysis with dedicated computer systems, (2) instrument control by microprocessors and (3) data acquisition and processing e.g. X-ray or EEL Spectroscopy.While image analysis in the T.E.M. usually needs a television chain to get a sequential transmission suitable as computer input, the STEM system already has this necessary facility. For the EM400T-STEM system therefore an interface was developed, that allows external control of the beam deflection in TEM as well as the control of the STEM probe and video signal/beam brightness on the STEM screen.The interface sends and receives analogue signals so that the transmission rate is determined by the convertors in the actual computer periphery.

Download Full-text

Exploring the Facial Color Representative Regions Using the Humanae Images

Journal of Imaging Science and Technology ◽

10.2352/j.imagingsci.technol.2020.64.6.060406 ◽

2020 ◽

Author(s):

Yuchun Yan ◽

Hayan Choi ◽

Hyeon-Jeong Suk

Keyword(s):

Image Analysis ◽

Future Study ◽

Skin Color ◽

Color Difference ◽

Facial Skin ◽

Female Images ◽

Gender And Race ◽

Facial Landmarks

It is difficult to describe facial skin color through a solid color as it varies from region to region. In this article, the authors utilized image analysis to identify the facial color representative region. A total of 1052 female images from Humanae project were selected as a solid color was generated for each image as their representative skin colors by the photographer. Using the open CV-based libraries, such as EOS of Surrey Face Models and DeepFace, 3448 facial landmarks together with gender and race information were detected. For an illustrative and intuitive analysis, they then re-defined 27 visually important sub-regions to cluster the landmarks. The 27 sub-region colors for each image were finally derived and recorded in L ∗ , a ∗ , and b ∗ . By estimating the color difference among representative color and 27 sub-regions, we discovered that sub-regions of below lips (low Labial) and central cheeks (upper Buccal) were the most representative regions across four major ethnicity groups. In future study, the methodology is expected to be applied for more image sources.

Download Full-text