The progress of research in medical-image assessment system of melasma

Ci Lingling（蔡玲玲）， Zhng Fengchun（張豐川）， Shi Jingwen（石靜紋）， Jin Yn（金艷），Wng Yo（王瑤）， Zhng Kunshun（張寬順）， nd Li Yunwen（李元文）*

a：Dermatology Department， Dongfang Hospital of Beijing University of Chinese Medicine， Beijing100078， China.b：Beijing University of Chinese Medicine， Beijing 100029，China*Corresponding author： Li Yuanwen，Doctoral supervisor，Vice President of the Dongfang Hospital of Beijing University of Chinese Medicine

?

The progress of research in medical-image assessment system of melasma

Cai Lingling（蔡玲玲）a， Zhang Fengchuan（張豐川）a， Shi Jingwen（石靜紋）b， Jin Yan（金艷）b，Wang Yao（王瑤）b， Zhang Kuanshun（張寬順）b， and Li Yuanwen（李元文）a*

a：Dermatology Department， Dongfang Hospital of Beijing University of Chinese Medicine， Beijing100078， China.
b：Beijing University of Chinese Medicine， Beijing 100029，China
*Corresponding author： Li Yuanwen，Doctoral supervisor，Vice President of the Dongfang Hospital of Beijing University of Chinese Medicine

ABSTRACT

Melasma， a kind of damaging disease， is commonly found in women during reproductive age. It causes great harm to patient's confidence and， has an effect on social contact as well. Melasma has been cured effectively by modern medicine and traditional Chinese medicine so far， in contrast， the treatment analysis system is far from flawless. Our research team is going to cooperate with Beijing University of Aeronautics and Astronautics and develop a computer-vision treatment analysis system for facial dermatosis based on present technology. It's supposed to provide clinical doctors with objective standards for melasma and other facial dermatosis.

Melasma； Medical-image assessment system.

Melasma is common in women of reproductive age， occasionally caught by male， as a common damaging disease. Even though innocent to life security， melasma causes great harm to patient's confidence1and， has an effect on social contact as well. The risk of depression in patients with melasma increases by 36%2in comparison with other damaging diseases， a survey showed. Therefore， melasma is not only a kind of skin disease， but also a psychological illness， which is in need of more public concern.

The improvements in medical technology and level of health care shadow a positive impact on the treatment of melasma with time3. It has been cured effectively by traditional Chinese medicine for long stretches of history. We got clinical effectiveness based on new theories proposed by our team and are ready to improve it. In the meantime， we apply the image analysis system to overcome the existing insufficiency. A more perfect assessment system is on the way to bloom.

According to our former research in clinical observations and relevant materials， we found that the diagnosis of melasma is clear4. In contrast， it's hard to evaluate the efficiency accurately. The doctors generally evaluate the severity of the disease in the light of subjective judgment in consulting rooms for diagnosis and tutoring （such as hue of the spots， size of the lesions，etc.）. However， the image analysis systems applied to clinical investigations as accessory equipment， at present， are far from being flawless. Multidisciplinary technology is in trend as we pursue more precise aegers. After communicating with relevant authority and looking into corresponding techniques， we hold a firm belief that an analysis of system on the basis of technology of personal 3-dimension image collection and analysis in support of assessing the treatment of melasma precisely， ruling out human interference factors， will soon emerge.

So far， it has been reported that cyber facial image system could be utilized to evaluate curative effect5， yet it compares the proportion of flat area， judging the difference of color by naked eyes. The outcomes of these methods are rich in inaccuracy，even found remarkable when the treatments bring about slight improvement. This discourages patients' faith undoubtedly.

At present， the domestic and international diagnosis of melasma is still based mainly on medical history and doctors subjective evaluation. In the case of “The Clinical Diagnosis and Curative Standard of the Melasma” （the revised edition of 2003） written by the dyschromatosis group in the Chinese Association of Integrative Medicine Dermatology & Venereology Committee. The diagnostic criteria of melasma are as followed：（1） Facial light brown to dark brown patches， well-demarcated，usually distributed symmetrically， no inflammation and scales.（2） No obvious symptoms. （3） More commonly seen in female，occurred mainly during post-adolescence. （4） The severity of disease changes with season， usually serious in winter and light in summer. （5） Exclusion of other diseases （such as nevus fuscocaeruleus zygomaticus， Riehl's melanosis， pigmentosus actiniclichen planus， etc.） which can cause pigmentation. The effective evaluation criteria include： （1） Basic cure： The disappearance rate of macroscopic patches is over 90%， the color almost disappears；the decreasing index after treatment calculated by scoring method is greater than or equal to 0.8； the ID value of pigmentation region skin image measure therapeutic evaluation unit is greater than or equal to 55. （2） Marked effect： The disappearance rate of macroscopic patches is over 60%， the color is noticeably lightened； the decreasing index after treatment calculated by scoring method is greater than or equal to 0.5； the ID value of pigmentation region skin image measure therapeutic evaluation unit is greater than or equal to 15. （3） Improvement： The disappearance rate of macroscopic patches is over 30%， the color becomes lighter； the decreasing index after treatment calculated by scoring method is greater than or equal to 0.3； the ID value of pigmentation region skin image measure therapeutic evaluation unit is greater than or equal to 5. （4） No effect： The disappearance rate of macroscopic patches is less than 30%， the color does not change obviously； the decreasing index after treatment calculated by scoring method is less than 0； the ID value of pigmentation region skin image measure therapeutic evaluation unit is greater than or equal to 1. Among them， the disappearance rate of macroscopic patches， the change of color， and the scoring index are the main evaluation standards. Therefore， the efficacy evaluation of melasma depends mainly on the doctor subjective judgment， which can easily produce error.

It has been a long time since we made use of face recognition as an assessment method of melasma in accordance with the 2003 edition Diagnosis and Treatment Criteria of Melasma. It indicates ID， as a major indicator， calculated from OPT， which is used to evaluate curative effect. Xiaohong Wu6， from Guang'anmen Hospital， estimated the treatment outcome depending on the data about the area and color of the spot which was processed from the information observing through dermoscope. It applies multi-spectral dermal polarizing microscopy and digital imageprocessing techniques to turn the pattern signals into digital signals. They have the capacity for providing more accurate and objective results7instead of inexact conclusions made by different people through naked eyes. Besides， Wei Luo， Wei Chen8and other domestic scholars assessed the clinical effect of melasma based on mean optical density and area calculated from digital image-processing system cured by Banke. Study proves the mean optical density and area decreased dramatically after 5-week treatment by Banke. Research suggests that the results calculated from digital image-processing system are objective. Also， Banke has effect on treating melasma. In E.Y.Tay and E.Y.Gan's5research paper， “Pilot study of an automated method to determine Melasma Area and Severity Index”， image processing digital scoring system was manifested as a reusable method for judging the melasma area and severity precisely. What's more， doctors can tell the distinction of different treatment plans. Coincidentally，Tsilika K and Levy JL7evaluated the treatment outcome by the combination of reflectance confocal microscopy and ultraviolet image technique.

At the same time， recently， with the improvement of big data and cloud computing， the development of mobile Internet，computer vision calculated camera， etc. related to the technology in face image analysis had made great progress in 3D imaging9，intrinsic image decomposition， the material analysis， and the color light according to the consistency. The few representative research works on this area are：

Joan Rubin認(rèn)為“學(xué)習(xí)策略是語言學(xué)習(xí)者用以獲取知識的技術(shù)或手段?！盵14]她指出,有意識地采用學(xué)習(xí)策略的學(xué)習(xí)者能夠幫助自己習(xí)得第二語言。按照她的研究,優(yōu)秀語言學(xué)習(xí)者具備的條件之一就是要在犯錯中提高自己的語言糾錯意識，不斷調(diào)整自己的學(xué)習(xí)策略。

In 2014， Chen Li， Kun Zhou10from Zhejiang University put forward a prior face images based on eigen decomposition methods with Microsoft Asia Research Institute cooperation. It is able to analyze mirror surface light from a single face image，light diffuse reflection， face material reflectance， 3D geometric information， environmental light， as shown in figure.

Figure 1： face image sign graph decomposition. （a） the method of single input image， （b） the method to decompose the specular light， （c） diffuse light， （d） face material reflectance， （e） 3D face geometry information， （f） acquisition of ambient light.

In 2014， the Chinese University of Hong Kong Zhanpeng Zhang11put forward a face image feature point detection based on the depth of learning， the method can more accurately detect theface in 68 individuals face key feature points from a single face image， as shown in Figure 2， the method can overcome a variety of influencing factors such as illumination， pose， expression，occlusion， etc. and， accurately detect the key feature points from the face image.

By 2015， Claudio Ferrari9， University of Florence， Italy，proposed dictionary based learning and deformable model for 3D face model reconstruction method， as shown in Figure 3， the method can recover the corresponding 3D face model from a single image.

We can make a conclusion from the studies above-mentioned that： image analysis system is playing a critically important role in treatment assessment. However， the method is still at an early age. It is expected to develop by absorbing latest achievements in computer vision technique and computer-assisted tomography. The present image analysis system still has some shortcomings. It lacks in consideration of ：（1） chromatic aberration induced by equipment and ray intensity， （2）the original complexion of patients with melasma， （3）the distinguishing features of 3-dimension in human face.

Problems listed above affect the accuracy of objective evaluation using computer images of Melasma， however，its diagnostic score combined with computer analysis of standardized procedures will no doubt greatly reduce human interference and， provide clinical and imaging diagnosis of melasma the objective， and accurate reference standards. This will with develop the whole digitalization evaluation system of diagnosis and treatment of skin diseases.

To sum up， melasma curative effect evaluation system using the digital precision of therapeutic effect evaluation system has become the development trend， melasma color and size are important indicators in the evaluation system.Therefore， we hope that by using multidisciplinary methods， we can create a more objective and accurate evaluation method. In addition， we also try to establish a database of normal skin in patients with melasma prototype， build personalized， standardized combination of pigment disorder tetter evaluation database services in clinic. Also， we value the 2003 version melasma clinic standard ID value evaluation system and score method in the used sucking photometric the plane area calculation and actual area exist insufficient differences. Through the cooperative development of Beijing University of Aviation Space of people face skin effect evaluation system based on computer Visual， we hope to evaluate the effect of Melasma in a single research based on reducing subjective factors interference， building a more precise three dimensional digital effect evaluation system for facial skin for Melasma and therapeutic effect of facial skin in the future provides an objective reference criteria.

1 Seite S， Deshayes P， Dreno B， et al.Interest of corrective makeup in the management of patients in dermatology. Clin Osmetic and Investigative Dermatol， 2012（5）：123-128.

2 TJLieu， AG Pandya.Melasma Quality of Life Measures Dermatol Clin， 2012， 30（2）：269-280.

3 Sarkar R，Arora P，Garg VK，Sonthalia S.Melasma update Indian dermatology online journal， 2014， 11，5（4）：426-435.

4 Bolognia JL， Jorrizo JL， Rapini RP. Dermatology. Second Edition. Mosby Elsevier， 2008.

5 Tay EY，Gan EY，Tan VW，et al. Pilot study of an automated method to determine Melasma Area and Severity Index. The British Journal of Dermatology， 2015， 72（6）：1535-1540.

6 Wu XH. Clinical Analysis of Famale Melasma and Assessment of Zishen Huaban Herbal Decoction by Dermoscopy，2010.

7 Tsilika K，Levy JL，Kang HY.A pilot study using reflectance confocal microscopy （RCM） in the assessment of a novel formulation for the treatment of melasma.Journal of drugs in dermatology. Journal of drugs in dermatology， 2011，10（11）：1260.

8 Luo W， Chen W， Wu RS，et al. Evaluation of Melasma Treatment by Computer Image Analysis System. Junyi Jinxiu Xueyuan Xuebao， 2008（04）：310-311.

9 Claudio Ferrari， Giuseppe Lisanti， Stefano Berretti， et al. Dictionary Learning based 3D Morphable Model Construction for Face Recognition with Varying Expressionand Pose. International Conference on 3D Vision （3DV），Lyon France， pp. 509-517， 19-22 October 2015.

10 Li C， Zhou K， Stephen Lin. Intrinsic Face Image Decomposition with Human Face Priors. In Proceedings of European Conference on Computer Vision （ECCV）， Zurich， Switzerland. Vol. 8693， pp. 218-233， September 6-12， 2014.

11 Zhang ZP， Luo P， Chen Change Loy， et al. Facial Landmark Detection by Deep Multi-task Learning， in Proceedings of European Conference on Computer Vision （ECCV）， Zurich，Switzerland. Vol. 8694， pp. 94-108， September 6-12， 2014.