TY - GEN
T1 - Automatic and Comprehensive Classification of the Morphological Forms of C. Albicans
AU - Sai Spandana, C.
AU - Soans, Rijul S.
AU - Prakash, Peralam Y.
AU - Galigekere, Ramesh R.
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Symptoms of fungal diseases can be similar to those of COVID-19, and lab tests are necessary to determine the precise infections a person is suffering from. Candida albicans (C. albicans) is an opportunistic pathogenic yeast that causes nosocomial infections, and patients exposed to hospital environments are vulnerable. The presence and number of its morphological forms (morphotypes/phenotypes) i.e., ellipsoidal/ovoid cells, budding cells, and hyphae - are useful for detecting the pathogenic propensity of C. albicans. However, there are certain challenges: fatigue and human errors associated with manual screening. In this paper, we propose a novel, automatic method of detecting and counting all of the morphological phenotypes of C. albicans from simple microscope-images of stained smears. The method involves color-segmentation in the HSV space using RUS-Boosted Decision Trees, and classification based on features such as area, width, the presence of constriction, and aspect ratio. The method was tested on 11 images with the training set formed from 43 images. The method showed a specificity of 79.3, 98.1, 100 & 98.6, sensitivity of 92.8, 82.2, 100 & 88.4, and accuracy of 87.1, 94.7, 100 & 96, on ellipsoidal yeast cells, budding yeast cells, hyphae and cluster of cells, respectively.
AB - Symptoms of fungal diseases can be similar to those of COVID-19, and lab tests are necessary to determine the precise infections a person is suffering from. Candida albicans (C. albicans) is an opportunistic pathogenic yeast that causes nosocomial infections, and patients exposed to hospital environments are vulnerable. The presence and number of its morphological forms (morphotypes/phenotypes) i.e., ellipsoidal/ovoid cells, budding cells, and hyphae - are useful for detecting the pathogenic propensity of C. albicans. However, there are certain challenges: fatigue and human errors associated with manual screening. In this paper, we propose a novel, automatic method of detecting and counting all of the morphological phenotypes of C. albicans from simple microscope-images of stained smears. The method involves color-segmentation in the HSV space using RUS-Boosted Decision Trees, and classification based on features such as area, width, the presence of constriction, and aspect ratio. The method was tested on 11 images with the training set formed from 43 images. The method showed a specificity of 79.3, 98.1, 100 & 98.6, sensitivity of 92.8, 82.2, 100 & 88.4, and accuracy of 87.1, 94.7, 100 & 96, on ellipsoidal yeast cells, budding yeast cells, hyphae and cluster of cells, respectively.
UR - http://www.scopus.com/inward/record.url?scp=85163477342&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85163477342&partnerID=8YFLogxK
U2 - 10.1109/ICICT57646.2023.10134358
DO - 10.1109/ICICT57646.2023.10134358
M3 - Conference contribution
AN - SCOPUS:85163477342
T3 - 6th International Conference on Inventive Computation Technologies, ICICT 2023 - Proceedings
SP - 676
EP - 680
BT - 6th International Conference on Inventive Computation Technologies, ICICT 2023 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 6th International Conference on Inventive Computation Technologies, ICICT 2023
Y2 - 26 April 2023 through 28 April 2023
ER -