Yüzdeki Dermatolojik Hastalıkların Otomatik Teşhisi İçin Bir Uygulama

Abstract

Amaç: Dermatolojik hastalıklar halk sağlığı problemleridir. Teşhisin öznel olması, yeterli dermatolog bulunmaması, yaşlılık, psikolojik sorunlar veya koronavirüs salgını gibi nedenlerle dermatoloğa gidememek gibi çeşitli faktörler dermatolojide otomatik tekniklerin kullanılmasını zorunlu hale getirmiştir. Literatürde, çeşitli dermatolojik hastalıkların teşhisinde doğru, objektif, güvenilir ve tekrarlanabilir sonuçlar sağlamak için otomatik lezyon sınıflandırması üzerine birçok teknik vardır. Fakat bu teknikler, ümit verici olmasına rağmen, uygun parametrelerin nasıl seçileceği veya nasıl ayarlanacağı ve kaynak kodlarının nasıl çalıştırılacağı konusunda önceden bilgisi olmayan birçok hasta veya kullanıcı için, bir ara yüz olmadan faydasız hale gelmektedirler. Bu nedenle, bu çalışmanın amacı, hastalar ve dermatologlar için etkili bir kullanıcı ara yüzüne sahip bir uygulama geliştirmektir. Gereç ve Yöntem: Uygulama genel veri tabanlarından sağlanan dijital fotoğraflar kullanılarak Matlab (R2019) ile geliştirilmiştir. Bulgular: Dermatolojik hastalıkları olan hastalar için etkili ve kullanıcı dostu ara yüze sahip bir uygulama tasarlanmış ve geliştirilmiştir. Sonuç: Uygulama, (i) lezyon bölütleme, (ii) görüntü sınıflandırması ve (iii) zaman periyotlarına göre analiz sonuçlarını sunmaktadır. Verilerin kolayca görüntülenmesini ve ağın parametrelendirilmesini sağlamaktadır. Dermatologlar tarafından lezyon analizinde veya karar verme aşamasında da kullanılabilir. Ayrıca uygulama, dermatolojik hastalıklar hakkında bilgi ve testler sağladığı için eğitim amaçlı da kullanılabilmektedir. Dolayısıyla bu hem hastalar hem de bu alanda çalışan araştırmacılar için faydalı olacaktır.

Keywords

• Sınıflandırma
• derin öğrenme
• teşhis
• deri lezyonu
• dermatolojik hastalıklar

An Application for Automated Diagnosis of Facial Dermatological Diseases

Abstract

Objective: Dermatological diseases are public health problems. Several factors including subjective diagnosis, lack of enough dermatologists, inability to go to a dermatologist due to old age, psychological problems or pandemic like coronavirus enforce to use automated techniques in dermatology. In the literature, there are many techniques on automated lesion classification to provide accurate, objective, reliable and reproducible results for the diagnosis of several dermatological diseases. However, although the techniques are promising, they become useless without a user interface for many patients or users who don’t have any prior knowledge on how to choose or set appropriate parameters and how to run source codes. Therefore, the objective of this work is to develop an application with an efficient user interface for patients and dermatologists. Material and Method: The application has been developed with Matlab (R2019) using digital photographs provided from public databases. Results: An application with an efficient and friendly user interface has been designed and implemented for patients with dermatological diseases. Conclusion: The application can present results of (i) lesion segmentation, (ii) image classification, and (iii) analysis according to time periods. It provides to view data easily and parametrization of the network. It can also be useful for dermatologists to analyze lesions or make decisions about diseases. Also, the application can be used for educational purpose since it provides information and tests on dermatological diseases. Therefore, it can be useful for both patients and researchers working in this area.

Keywords

• Classification
• deep learning
• dermatological diseases
• diagnosis
• skin lesion

References

1. Fritsch P, Burgdorf W. The skin and its diseases: an overview. European Journal of Dermatology. 2016 April 2;16(2):209-212.
2. Tizek L, Schielein MC, Seifert F, Biedermann T, Böhner A, Zink A. Skin diseases are more common than we think: screening results of an unreferred population at the Munich Oktoberfest. European Academy of Dermatology and Venereology. 2019 July 7;33(7):1421-1428.
3. Svensson RF, Ofenloch M, et al. 2017. Prevalence of skin disease in a population-based sample of adults from five European countries. British Journal of Dermatology. 2018 May 5;178(5):1111-1118.
4. Lim HW, Collins SAB, et al. The burden of skin disease in the United States. American Academy of Dermatology. 2017 May 17;76(5):958-972.
5. PlokaDot Mama Foundation, About Melanoma., [cited 2020 September 10]. Available from: http://polkadotmama.org/aboutmelanoma.
6. Carrera C, Marchetti MA, et al. Validity and reliability of dermoscopic criteria used to differentiate nevi from melanoma: a web-based international dermoscopy society study. Journal of The American Academy of Dermatology. 2016 July 1;152(7):798–806.
7. Masood A, Al-Jumaily AA. Computer aided diagnostic support system for skin cancer: a review of techniques and algorithms. Biomedical Imaging. 2013 December 23;1(1):1-23.
8. Thomsen K, Iversen L, Titlestad LT, Winther O. Systematic review of machine learning for diagnosis and prognosis in dermatology. Journal of Dermatological Treatment. 2019 October 31;31(5):496-510.

9. Chan S, Reddy V, Myers B, Thibodeaux Q, Brownstone N, Liao W. Machine learning in dermatology: current applications, opportunities, and limitations. Dermatology and Therapy. 2020 April 6;10(3):365-386.
10. Hogarty DT, Su JC, Pjan K, Attia M, Hossny M, Nahavandi S, Lenane P, Moloney FJ, Yazdabadi A. Artificial intelligence in dermatology – where we are and the way to the future: a review. American Journ. of Clinical Dermatology. 2020 February 21;21(1):41-47.
11. Bajwa MN, Muta K, et. al. Computer-aided diagnosis of skin diseases using deep neural networks. Applied Sciences. 2020 March 5;10(7):2-13.
12. Puri P, Comfere N, et al. Deep learning for dermatologists: part II current applications. [cited 2020 October 16] Available from: https:// doi.org/10.1016/j.jaad.2020.05.053.
13. Shanthi T, Sabeenian RS, Anand R. Automatic diagnosis of skin diseases using convolution neural network. Microprocessors and Microsystems. 2020 July 1;76(1):1-7.
14. Goceri E. Deep learning based classification of facial dermatological disorders. Computers in Biology and Medicine. 2020 November 12;128(104118):1-26.
15. Lim ZV, Akram F, Ngo CP, Winarto AA, Lee WQ, Liang K, Oon HH, Thng STG, Lee HK. Automated grading of acne vulgaris by deep learning with convolutional neural networks. Skin Research Technology. 2020 March 1;26(2):187– 192.
16. Melin P, Prado-Arechiga G, Miramontes I, Guzman JC. Hypertension diagnosis with a soft computing model using a graphical user şnterface. Hypertension. 2019 July 1;37(1):1-1.
17. Mandalika VBH, Chernoglazov A, et al. A hybrid 2D/3D user interface for radiological diagnosis. Journal of Digital Imaging. 2018 July 1;31(1):56–73.
18. Castroa DL, Tegolo D, Valenti C. A visual framework to create photorealistic retinal vessels for diagnosis purposes. Journal of Biomed Informatics. 2020 June 1;108(1):1-12.
19. Ech-Cherif A, Misbhauddin M, Ech-Cherif M. Deep neural network based mobile dermoscopy application for triaging skin cancer detection. Conf. on Comp. App. and Inf. Security, Riyadh, Saudi Arabia, 2019 May 1;1(1):1-6.
20. Sae-Lim W, Wettayaprasit W, Aiyarak P. Convolutional neural networks using mobilenet for skin lesion classification. Conf. on Comp Sc. and Software Eng., Chonburi, Thailand, 2019 July 10;1(1):242-247.
21. Velasco J, Pascion C, Alberio JW, et al. A smartphone-based skin disease classification using mobilenet cnn. Advanced Trends in Computer Science and Engineering. 2019 November 1;8(5):2632-2638.
22. Goceri E. Diagnosis of skin diseases in the era of deep learning and mobile technology. Computers in Biology and Medicine. 2021 May 4;134(104458):1-30.
23. DermNet website. [cited 2020 June 22] Available from: https://www. dermnetnz.org/image-catalogue.
24. DermQuest Image Library. [cited 2020 September 22] Available from: https://www.dermquest.com/image-library.
25. Gove R. publication on using the elbow method to determine the optimal number of clusters for k-means clustering. [cited 2021 April 1] Available from: https://bl.ocks.org/rpgove/0060ff3b656618e9136b.
26. Huang G, Liu Z, van der Maaten L, Weinberger K.Q. Densely connected convolutional networks. Conf. Comp. Vis. Pattern Rec., Honolulu, USA, 2017 July, 1;1(1):2261–2269.
27. Liu R, Sun Z, Wang A, Yang K, Wang Y, Sun Q. Lightweight efficient network for defect classification of polarizers. Concurrency Computation- Practice and Experience. 2020 January 9;32(11):1-10.
28. Ntakolia C, Diamantis DE, Papandrianos N, Moustakidis S, Papageorgiou EI. A lightweight convolutional neural network architecture applied for bone metastasis classification in nuclear medicine: A case study on prostate cancer patients. Healthcare. 2020 November 18;8(4):1-13.
29. Ayi M, El-Sharkawy M. Real-time implementation of RMNv2 classifier in NXP bluebox 2.0 and NXP i.MX RT1060. IEEE Midwest Industry Conf.. 2020 August 7;1(1):1-4.
30. Lim JS. Two-dimensional signal and image processing. Prentice Hall, Saddle River, NJ. 1990 January 1.
31. Harremoes P, Tusnady G. Information divergence is more chi squared distributed than the chi squared statistics. arXiv:1202.1125. 2012 June 17;1(1)1-7.
32. Enos C, Patel T, Patel S, França K. Seborrheic dermatitis. Stress and Skin Disorders, Springer, Cham. 2016 December 18;1(1):165-169.
33. WebMD website on seborrheic dermatitis. [cited 2020 May 12] Available from: https://www.webmd.com/skinproblemsandtreatments/ seborrheic-dermatitis-medref#1.
34. MayoClinic-Hemangioma [cited 2020 May 22] Available from: www.mayoclinic.org/diseasesconditions/hemangioma/symptoms-causes/ syc20352334.
35. Karadağ AS, Bilgili SG, Çalka Ö, Demircan YT. The retrospective evaluation of childhood psoriasis clinically and demographic features. Turk Journal of Dermatology. 2013 April 1;7(1):13-18.
36. Mıstık S, Ferahbaş A. Approach to treatment of acne vulgaris in family medicine. Turk Aile Hekimligi Dergisi. 2005 May 6;9(2):71-78.
37. Bonamigo RR, Bertolini W, Dornelles SIT. Rosacea. Dermatology in Public Health Environments, Springer, Cham. 2017 May 30;1(1):465-479.
38. National Rosacea Society official website on signs and symptoms of rosacea disease. [cited 2020 April 2] Available from: https://www. rosacea.org/patients/frequently-asked-questions#swelling.
39. American Academy of Dermatology Association (AAD) official website on rosacea treatment: eye problems. [cited 2020 April 24] Available from: https://www.aad.org/public/diseases/rosacea/ treatment/eye-problems.
40. Ada Global Health company website on acne vulgaris. [cited 2020 March 26] Available from: https://ada.com/conditions/acne-vulgaris.
41. Mayo Clinic official website on acne. [cited 2020 May 22] Available from: https://www.mayoclinic.org/diseasesconditions/acne/symptomscauses/syc-20368047.
42. Sagdeo A, Wanet K, Seykora J. Inflammatory reaction patterns and diseases of skin. Pathobiology of Human Disease. 2014 May 4;1(1):1160- 1167.
43. MedicalNewsToday information on psoriasis and rosacea diseases. [cited 2020 May 16] Available from: https://www.medicalnewstoday. com/articles/314268.
44. Harvard Medical School website. [cited 2020 May 15] Available from: https://www.health.harvard.edu/diseases-and-conditions/psoriasismore- than-skin-deep.
45. Leo Pharma Company official website on what is psoriasis. [cited 2020 May 16] Available from: https://www.schuppenflechte-im-griff.de/ psoriasisvestehen/psoriasis-alle-wichtigen-informationen-auf-einenblick.
46. MedicalNewsToday website on psoriasis. [cited 2020 May 22] Available from: https://www.medicalnewstoday.com/ articles/314268#psoriasis.
47. National Psoriasis Foundation website on assessing skin pain in psoriasis patients. [cited 2020 May 22] Available from: https://www. psoriasis.org/advance/assessing-skin-pain-psoriasis-patients.
48. Harvard Medical School website on dermatitis. [cited 2020 May 22] Available from: https://www.health.harvard.edu/diseases-andconditions/psoriasis-more-than-skin-deep.