Related Experiment Video
Updated: May 3, 2026
07:58
Use of Image Cytometry for Quantification of Pathogenic Fungi in Association with Host Cells
Published on: June 19, 2013
12.9K
Automated grading system for quantifying KOH microscopic images in dermatophytosis.
Rajitha Kv1, Sreejith Govindan2, Prakash Py3
1Dept.of Biomedical Engg, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal, 576104, Karnataka, India.
Diagnostic Microbiology and Infectious Disease
|November 3, 2024
Summary
This study introduces an automated system to quantify fungal load in skin scrapings for dermatophytosis diagnosis. The AI model accurately grades fungal infections, aiding in prognosis and treatment monitoring.
Area of Science:
- Medical Mycology
- Computational Pathology
- Image Analysis
Background:
- Accurate quantification of fungal load in dermatophytosis is crucial for prognosis.
- Current methods for fungal load assessment can be subjective and time-consuming.
- There is a need for objective and efficient tools to grade fungal infections from microscopic images.
Purpose of the Study:
- To develop and validate an automated grading system for quantifying fungal loads in KOH microscopic images.
- To establish a pixel count-based method for grading dermatophytosis severity.
- To improve the objectivity and efficiency of fungal load assessment in clinical practice.
Main Methods:
- Fungal filaments in KOH images were segmented using a U-Net deep learning model to determine pixel counts.
- Expert dermatologists manually graded images in the absence of established thresholds.
- Cumulative receiver operating characteristic (ROC) curve analysis was employed to develop the automated grading system based on pixel counts and expert grades.
Main Results:
- The automated system achieved high performance metrics: specificity >92%, accuracy >86%, precision >82%, and sensitivity >76%.
- An 'almost perfect agreement' (Fleiss kappa = 0.847) was observed between the automated grading system and manual expert gradings.
- The developed system provides a reliable pixel count-based quantification of fungal load in dermatophytosis.
Conclusions:
- The developed automated grading system offers a novel, cost-effective, and objective method for quantifying fungal load in KOH images.
- This technology has the potential to significantly aid in the diagnosis, prognosis, and management of dermatophytosis.
- Pixel count-based analysis of microscopic images represents a promising approach for automated disease grading in clinical settings.
