Khajista Nizam | Multimedia University (original) (raw)
Graduate Research Assistant at Multimedia University, Malaysia.
Project title: Automated Diagnosis & Prognosis of Chronic Wound for E-Health Applications
Collaboration organizations:
- Hospital Kuala Lumpur (HKL), Malaysia
- The Malaysian Society for Wound Care Professional (MSWCP)
- Ohio State University, (OSU), U.S.A.
Funded by:
- Ministry of Science, Technology, & Innovation (MOSTI), Malaysia
Publications:
- Characterization of Tissues in Chronic Wound Images
"Best Paper Award" in IEEE SCOReD Conference 2018
(https://doi.org/10.1109/SCORED.2018.8710941)
- Enhancement in the Identification of Slough Tissue in Chronic Wound Assessment
Accepted in IEEE ICSIPA Conference 2019
less
Uploads
Papers by Khajista Nizam
IEEE, 2019
Diabetes is a serious health issue faced by the society. Classification of tissues in a diabetic ... more Diabetes is a serious health issue faced by the society. Classification of tissues in a diabetic wound is highly essential to monitor the healing progress of the wound. The standard manual tissue classification methods utilized by the clinicians are prone to inaccuracy and is highly dependant on the knowledge and experience of the expert. Moreover, since the wounds may form at a body part that is difficult to access, assessing of the wound becomes time-consuming and causes distress to the patient. Hence, an automatic diabetic wound classification system is required for accurate assessment of the wound. In this paper, we offer tissue classification system for diabetic ulcers that utilizes Fuzzy C-Means and colour analysis using colour features to cluster the digital images. The clusters are initially labelled by calculating the minimum Euclidean Distance of the cluster center from the reference points. Each cluster was further analyzed by using spatial relation to label the smaller blobs, and tissue relation to label larger blobs. Granulation and Epithelial were further evaluated by computing the shiny regions caused by the reflection of light on wetter regions of the wound. The proposed algorithm will segment wound tissues into granulation, slough, eschar, and epithelial tissues and display the segmented regions along with the percentage area of each type of tissue. This system can be used by the diagnostician to effectively and easily monitor the healing progress of the wound.
IEEE, 2019
Diabetes is a serious health issue faced by the society. Classification of tissues in a diabetic ... more Diabetes is a serious health issue faced by the society. Classification of tissues in a diabetic wound is highly essential to monitor the healing progress of the wound. The standard manual tissue classification methods utilized by the clinicians are prone to inaccuracy and is highly dependant on the knowledge and experience of the expert. Moreover, since the wounds may form at a body part that is difficult to access, assessing of the wound becomes time-consuming and causes distress to the patient. Hence, an automatic diabetic wound classification system is required for accurate assessment of the wound. In this paper, we offer tissue classification system for diabetic ulcers that utilizes Fuzzy C-Means and colour analysis using colour features to cluster the digital images. The clusters are initially labelled by calculating the minimum Euclidean Distance of the cluster center from the reference points. Each cluster was further analyzed by using spatial relation to label the smaller blobs, and tissue relation to label larger blobs. Granulation and Epithelial were further evaluated by computing the shiny regions caused by the reflection of light on wetter regions of the wound. The proposed algorithm will segment wound tissues into granulation, slough, eschar, and epithelial tissues and display the segmented regions along with the percentage area of each type of tissue. This system can be used by the diagnostician to effectively and easily monitor the healing progress of the wound.