Numerical grading of astrocytomas

Improving accuracy in astrocytomas grading by integrating a robust least squares mapping driven support vector machine classifier into a two level grade classification scheme

Grading of astrocytomas is an important task for treatment planning; however, it suffers from significantly great inter-observer variability. Computer-assisted diagnosis systems have been propose to assist towards minimizing subjectivity, however, these systems present either moderate accuracy or utilize specialized staining protocols and grading systems that are difficult to apply in daily clinical practice. The present study proposes a robust mathematical formulation by integrating state-of-art technologies (support vector machines and least squares mapping) in a cascade classification scheme for separating low from high and grade III from grade IV astrocytic tumours. Results have indicated that low from high-grade tumours can be correctly separated with a certainty as high as 97.3%, whereas grade III from grade IV tumours with 97.8%. The overall performance was 95.2%. These high rates have been a result of applying the least squares mapping technique to features prior to classification. A significant byproduct of least squares mapping is that the number of support vectors of the SVM classifiers dropped dramatically from about 80% when no mapping was used to less than 5% when mapping was used. The latter is a clear indication that the SVM classifier has a greater potential to generalize well to new data. In this way, digital image analysis systems for automated grading of astrocytomas are brought closer to clinical practice.

An image-analysis system based on support vector machines for automatic grade diagnosis of brain-tumour astrocytomas in clinical routine

An image-analysis system based on the concept of Support Vector Machines (SVM) was developed to assist in grade diagnosis of brain tumour astrocytomas in clinical routine. One hundred and forty biopsies of astrocytomas were characterized according to the WHO system as grade II, III and IV. Images from biopsies were digitized, and cell nuclei regions were automatically detected by encoding texture variations in a set of wavelet, autocorrelation and parzen estimated descriptors and using an unsupervised SVM clustering methodology. Based on morphological and textural nuclear features, a decision-tree classification scheme distinguished between different grades of tumours employing an SVM classifier. The system was validated for clinical material collected from two different hospitals. On average, the SVM clustering algorithm correctly identified and accurately delineated 95% of all nuclei. Low-grade tumours were distinguished from high-grade tumours with an accuracy of 90.2% and grade III from grade IV with an accuracy of 88.3% The system was tested in a new clinical data set, and the classification rates were 87.5 and 83.8%, respectively. Segmentation and classification results are very encouraging, considering that the method was developed based on every-day clinical standards. The proposed methodology might be used in parallel with conventional grading to support the regular diagnostic procedure and reduce subjectivity in astrocytomas grading.

Histomorphometry of tumour cell nuclei in astrocytomas using shape analysis, densitometry and topometric analysis

Although tumour cell nuclei are important histological structures for grading of astrocytomas according to the WHO-classification of brain tumours, there is no reported morphometric study of astrocytomas which describes quantitatively the four main morphologic criteria of tumour cell nuclei: size, shape, texture (densitometric characteristics) and spatial relationships between the nuclei (topometric analysis). Using a set of morphometric parameters describing these criteria as well as the Ki67-proliferation index, 74 astrocytomas from 74 patients were studied by means of a digital image analysis system. The objective of the study was to test, if these morphometric parameters were sufficient for statistical discrimination between pilocytic astrocytomas WHO-grade I, astrocytomas grade II and anaplastic astrocytomas grade III. Our results showed a correct reclassification of 97.3% (72/74) of the cases with respect to the tumour grade by means of cross-validated discriminant analysis. Morphometric parameters characterizing nuclear shape (shape factors, Fourier-amplitudes) showed the most prominent differences between the three groups of cases, followed by topometric parameters (number of neighbours per nucleus, distances between the nuclei). Less pronounced differences between the tumour grades were found for parameters characterizing nuclear size, nuclear texture and the Ki67-proliferation index. In conclusion, the present morphometric procedure provided good discrimination between the tumour grades, supporting the view that histomorphometry of tumour cell nuclei could be a valuable tool for grading of astrocytomas.

Histomorphometry of brain tumours

In this review, the results of previous histomorphometric studies of brain tumours are summarized and discussed with respect to their potential value for diagnostic purposes and for tumour research. In the majority of these studies, human gliomas were investigated. In a few studies, human meningiomas and other human or experimental tumour types were investigated. A computerized image analysis system was used for the morphometric analyses in most studies. The three main histologic structures examined were tumour cell nuclei, nucleolar organizer regions and tumour vessels. The current state of knowledge provides evidence that a diagnostic benefit could be provided by histomorphometric investigations of brain tumours, especially for grading of gliomas and with respect to independent prognostic information. Additional studies are necessary to delineate the spectrum of histomorphometric parameters and the investigation of their prognostic significance for cases with the same tumour type and tumour grade. Together with many recently published observations in this field, this review shows that histomorphometry is an important approach towards the investigation of brain tumour biology.

Nuclear morphometry and DNA densitometry of human gliomas by image analysis

In 48 patients with gliomas in whom complete clinical follow-up was obtained, DNA ploidy was evaluated by using formalin-fixed paraffin-embedded tissues and by means of image analysis. The mean DNA indices, determined by averaging DNA indices of all tumor cells in a tumor, were mainly affected by mean DNA indices of the nuclei of SG2M phase tumor cell (including S phase and G2M phase cells) (SG2M DNA indices) and that mean DNA indices correlated with the SG2M phase fraction. The SG2M DNA indices and the percentage of tumor cells with S phase and G2M phase were higher in high grade gliomas including anaplastic glioma and glioblastoma multiforme than in low grade gliomas. Patients with G2M-hypertetraploid tumors demonstrated a shorter time to tumor progression than those with G2M-tetraploid in high grade glioma. Morphometrically, the nuclei of SG2M phase glioma cells were larger and more deformity than those of G0G1 phase (including G0 phase and G1 phase cells) cells. The G2M-hypertetraploid tumors were highly malignant and demonstrated large nuclei, greater nuclear deformity, and a higher proliferative potential. The G2M-tetraploid gliomas demonstrated a shorter time to tumor progression in cases whose the SG2M fraction was large. In contrast, G2M-hypotetraploid gliomas revealed an insignificant trend towards a longer time to tumor progression than those associated with tetraploid and hypertetraploid gliomas. We emphasize herein the prognostic importance of the SG2M phase cell, as well as other proliferation indices.