MYCOLab Ai Logo - gold and army green

Assessing the Validity and Impact of Remote Digital Image Reading in Fungal Diagnostics

Summary

This study tested whether trained mycologists could accurately identify fungal infections from digital images viewed remotely, similar to how radiologists review X-rays. Five experienced laboratory professionals analyzed 474 images of different fungi with accuracy rates between 78-93%. The results suggest that remote digital diagnosis could help hospitals in developing countries where expert mycologists are scarce, enabling faster and more accurate diagnosis of serious fungal infections.

Background

Mycological diagnostics are crucial for treating invasive fungal infections, but diagnostic capabilities lag in many regions, especially low- and middle-income countries. Traditional fungal diagnosis requires skilled mycologists and labor-intensive microscopy and culture methods. Remote digital image reading could provide diagnostic support across geographical regions where expertise is limited.

Objective

To develop and validate remote digital image reading for fungal diagnostics by assessing the feasibility and accuracy of preliminary fungal identification using digital images analyzed by trained mycologists.

Results

Overall correct decision rates ranged from 78-93% across image groups, with yeasts showing highest accuracy (93%), followed by direct microscopy (86%) and filamentous fungi (78%). Individual participant accuracy ranged from 76-92%, with significant differences among participants but no significant interaction between participant and image group type.

Conclusion

Remote digital image reading shows promise for fungal diagnostics, especially in settings with limited mycological expertise. The method could enable remote diagnostic support globally through networks of specialists, potentially improving timely diagnosis and treatment of fungal infections in resource-limited regions.
  • Published in:Mycopathologia,
  • Study Type:Feasibility Study,
  • Source: PMID: 41201691, DOI: 10.1007/s11046-025-01012-x
Scroll to Top