Deep convolutional neural networks for onychomycosis detection using microscopic images with KOH examination

Abstract

Background The diagnosis of superficial fungal infections is still mostly based on direct microscopic examination with potassium hydroxide solution. However, this method can be time consuming, and its diagnostic accuracy rates vary widely depending on the clinician's experience. Objectives This study presents a deep neural network structure that enables the rapid solutions for these problems and can perform automatic fungi detection in grayscale images without dyes. Methods One hundred sixty microscopic full field photographs containing the fungal element, obtained from patients with onychomycosis, and 297 microscopic full field photographs containing dissolved keratin obtained from normal nails were collected. Smaller patches containing fungi (n = 1835) and keratin (n = 5238) were extracted from these full field images. In order to detect fungus and keratin, VGG16 and InceptionV3 models were developed by the use of these patches. The diagnostic performance of models was compared with 16 dermatologists by using 200 test patches. Results For the VGG16 model, the InceptionV3 model and 16 dermatologists, mean accuracy rates were 88.10 ± 0.8%, 88.78 ± 0.35% and 74.53 ± 8.57%, respectively; mean sensitivity rates were 75.04 ± 2.73%, 74.93 ± 4.52% and 74.81 ± 19.51%, respectively; and mean specificity rates were 92.67 ± 1.17%, 93.78 ± 1.74% and 74.25 ± 18.03%, respectively. The models were statistically superior to dermatologists according to rates of accuracy and specificity but not to sensitivity (p < .0001, p < .005 and p > .05, respectively). Area under curve values of the VGG16 and InceptionV3 models were 0.9339 and 0.9292, respectively. Conclusion Our research demonstrates that it is possible to build an automated system capable of detecting fungi present in microscopic images employing the proposed deep learning models. It has great potential for fungal detection applications based on AI.

Artificial Intelligence in microbiomes analysis: A review of applications in dermatology

“…The differences in diagnosing onychomycosis using CNN vs. manual microscopic examinations have been compared and analyzed. Results from one report found that the accuracy of CNN diagnosis was 10% greater than that of traditional diagnostic methods ( Yilmaz et al, 2022 ). In that study, 60 nail samples from patients with onychomycosis and 297 nail samples from healthy controls were treated with KOH.…”

Diagnosing onychomycosis: A step forward?

“…It was found that the CNN had an AUC of 1.00 (99.84% accuracy). 18 In comparison, the average AUC value of clinicians is 0.87 (72.8% accuracy, determined through a meta-analysis of 11 studies 18 ). These results indicate that AI can accurately perform an automatic diagnosis of onychomycosis using fungi detection from grayscale microscopic images, and at a level that outperforms clinicians.…”

“…Yilmaz et al (2021) used a CNN to examine grayscale microscopic images of nail samples (as colorants can be expensive and time consuming) and determine diagnoses of onychomycosis. Microscopic images of fungi and keratin were taken from patients with distal and lateral onychomycosis 18. The samples were kept in 20% KOH for 15-20 min and then examined directly under a microscope by two experienced dermatologists and determined to be positive or not.…”

“…Yilmaz et al (2021) used a CNN to examine grayscale microscopic images of nail samples (as colorants can be expensive and time consuming) and determine diagnoses of onychomycosis. Microscopic images of fungi and keratin were taken from patients with distal and lateral onychomycosis 18.…”

MobileSkin: Classification of Skin Lesion Images Acquired Using Mobile Phone-Attached Hand-Held Dermoscopes

“…Recently, deep learning systems that will work as a decision support mechanism for the diagnosis of skin diseases have begun to be developed for people such as inexperienced dermatologists, nurses, and primary care doctors who are not experts in the field of dermatology [ 22 , 23 ]. Teledermoscopy was studied based on feature extraction with a small binary dataset [ 24 ].…”