Related Experiment Videos
A grading study of gliomas using computer aided malignancy classification and histologic morphometry
S Sharma1, A K Karak, C Sarkar
1Department of Pathology, All India Institute of Medical Sciences, New Delhi, India.
Journal of Neuro-Oncology
|January 1, 1996
Summary
The Kernohan grading system for gliomas shows high variability. Computer-aided classification (TESTAST 268) combined with morphometry significantly improves grading reproducibility.
Area of Science:
- Neuropathology
- Oncology
- Medical Imaging
Background:
- Accurate grading of astrocytic tumors and mixed gliomas is crucial for patient management and prognosis.
- Traditional grading systems, like Kernohan, suffer from significant inter- and intra-observer variability.
Purpose of the Study:
- To evaluate the reproducibility of the Kernohan grading system compared to a computer-aided classifier (TESTAST 268) and quantitative morphometric evaluation.
- To assess the potential of TESTAST 268 and morphometry to improve glioma grading objectivity.
Main Methods:
- Studied 43 cases of astrocytic tumors and mixed gliomas.
- Compared Kernohan grading with TESTAST 268 classification and morphometric analysis of histological parameters.
- Followed patients for up to 40 months.
Main Results:
- High inter- and intra-observer variability noted with the Kernohan grading system.
- TESTAST 268 demonstrated greater simplicity, speed, and reproducibility, though some subjectivity remained.
- Morphometric evaluation of TESTAST 268 parameters showed statistically significant differences.
- Combining TESTAST 268 with morphometry-derived values eliminated inter-observer variability in repeat grading.
Conclusions:
- Objectivization using TESTAST 268 and histologic morphometry is vital for reproducible glioma grading.
- This preliminary study suggests a promising approach to enhance diagnostic accuracy in neuropathology.
- Further research is needed to establish definitive cut-off values for these measurements.