Early Detection of Diabetic Foot Ulcers by Thermal Images of Foot Soles Using Nearest Neighbor Algorithm

Document Type : Research Paper

Authors

1 Assistant Professor, Department of Electrical Engineering, Nain Branch, Islamic Azad University, Nain, Iran

2 Assistant Professor, Department of General Surgery, Qom University of Medical Sciences, Qom, Iran

3 M.Sc., Islamic Azad University, Najafabad Branch, Isfahan, Iran

Abstract

Diabetes is a disease, which is caused by the cessation of insulin production or the dysfunction of the body. Early detection of diabetic foot ulcers thermal images of the sole of the foot is one of the new methods of diagnosing diabetic foot ulcers. In this paper, by optimizing the nearest neighbor algorithm, early detection of diabetic foot ulcers is performed by comparing the thermal similarity of the left and right soles of the feet. And the condition of the possibility of foot ulcer is diagnosed. In the proposed solution, by removing additional areas and creating a temperature image of the soles of the feet, using image matching techniques and then extracting statistical features such as standard distribution, percentage of dissimilar pixels and average temperature of the soles of the feet, early diagnosis of the ulcer condition is attempted. To evaluate the proposed method, 74 images of gray surfaces were used, in which the image of the left and right soles of the feet is specified in the image, and along with the images, there is a file in which the class of each image is specified. The information file also contains the minimum and maximum temperatures in the image to create a thermal image. Therefore, the above problem is a 3-class classification problem in which 75% of the images are used for training and 25% for testing. We have used thermal images by cross-validation method, which we have achieved with a total accuracy of 85.14%.

Keywords


  1. Luis A. Vilcahuaman C. Early diagnostic of diabetic foot using thermal images. Université d’Orléans, 2013.English. ffNNT : 2013ORLE2061ff. fftel-01022921
  2. Hossein Nejad Zahra, Nasri Mehdi. An overview of image matching methods, concepts and applications.
  3. Machine vision and image processing [Persian]. 2017;5(2):39-67. Available from:https://sid.ir/paper/393933/fa
  4.  Kashefi A, Haddadnia J. Provide an automated method for quantitive comparison of peer at the feet of
  5. diabetic foot ulcers in order to predict the thermal images and computer vision techniques. ijdld. 2015; 14(5) :325-336
  6.  Inlow Sh, Orsted H, Sibbald RG. Best practices for the prevention, diagnosis and treatment of diabetic footulcers. Ostomy/Wound Management 2000; 46: 55-68.
  7. Fryberg RG, et al. Diabetic foot disorders, a clinical practice guideline. American College of Foot and AnkleSurgeons and the American College of Foot and Ankle Orthopedics and Medicine 2000: 1-48.
  8. Boyko EJ, Ahroni JH, Stensel VL. Skin temperature in the neuropathic diabetic foot. J. DiabetesComplications 15, 2001,pp. 260–264.
  9. Armstrong DG, Lavery LA, Liswood PJ, Todd WF, and Tredwell J. Infrared dermal thermometry of thehigh-risk diabetic foot. Phys Ther, vol. 77, pp. 169–177, 1997.
  10. Lavery LA, Armstrong DG, Vela S, and Fleishli JG. Practical criteria to screen patients at risk for diabeticfoot ulceration, Arch. Int. Med., vol. 158, pp. 157–162, 1998.
  11. Chen Y, Kaabouch N, Anderson J, Forrest Ames F. Static and dynamic skin thermal distributions for earlydetection/prediction of foot ulcers. 2008 North Dakota EPSCoR State Conference, Grand Forks, September19, 2008.
  12. Pi-Chang Sun, Hong-Da Lin, Shyh-Hua Eric Jao, Yan-Chiou Ku, Rai-Chi Chan, Cheng-Kung Cheng.
  13. Relationship of skin temperature to sympathetic dysfunction in diabetic at-risk feet. 2006 Elsevier IrelandLtd.
  14. Oe M, Yotsu RR, Sanada H, Nagase T, Tamaki T. Thermographic findings in a case of type 2 diabetes withfoot ulcer and osteomyelitis.J Wound Care. 2012;21(6):274, 276-8.
  15.  Sumpio BE. Foot ulcers. New England Journal of Med. 2000; (343): 787–793.
  16. Sun PC, Jao SH, Cheng CK. Assessing foot temperature using infrared thermography. Foot Ankle Int., 2005;(26): 847–853.
  17.  Liu C, van der Heijden F, and van Netten JJ. Towards surface analysis on diabetic feet soles to predictulcerations using photometric stereo, SPIE, 2012.
  18.  Kaabouch N, Chen Y, Anderson J, Ames F, and Paulson R. Asymmetry Analysis Based on GeneticAlgorithms for the Prediction of Foot Ulcers.Visualization and Data Analysis 2009, SPIE Vol. 7243.
  19. Derakhshan Borjui, Pooya and Bahadurzadeh, Mojdeh, 2017, early detection of response to diabetes woundtreatment by genetic algorithm, National Conference of Innovative and Applied Innovation in Electrical andComputer Engineering with Knowledge-Based Approach, Nain
  20. Yadon A. Beyond the Monofilament for the Insensate Diabetic Foot A systematic review of randomized trials to prevent the occurrence of plantar foot ulcers in patients with diabetes. Diabetes Care 2011; 34:41041-1046.
  21. Ng EYK. A review of thermography as promising noninvasive detection modality for breast tumor.International Journal of Thermal Sciences 2009, 48 (5): 849-859.
  22. Ring EFJ, Ammer K. The technique of infrared imaging in medicine. Thermology International 2000; 7: 7–14.
  23. Roback K. An overview of temperature monitoring devices for early detection of diabetic foot disorders.Expert Rev Med Devices 2010; 7(5):711-718.
  24. Heydari, S., Radfar, R., Alborzi, M., Afshar Kazemi, M., Rajabzadeh Ghatari, A. Analysis of DiabeticPatients' Data for Clustering and Prescription Drug Based on Proposed Algorithm. medical journal of mashhad university of medical sciences, 2020; 63(2): 2358-2368