You may check this paper for further information: https://www.nature.com/articles/sdata2017117.

The algorithm GLISTRboost mentioned in this paper is a winner of BRATS.

According to this paper, the definition of these tumor components are:

A. “The ET is described by areas that show hyper-intensity in T1-Gd when compared to T1, but also when compared to normal/healthy white matter (WM) in T1-Gd.”

B. “The NET represents non-enhancing tumor regions, as well as transitional/pre-necrotic and necrotic regions that belong to the non-enhancing part of the tumor core (TC) and are typically resected in addition to the ET. The appearance of the NET is typically hypo-intense in T1-Gd when compared to T1, but also when compared to normal/healthy WM in T1-Gd.”

C. “The ED is described by hyper-intense signal on the T2-FLAIR volumes.”

The BraTS dataset describes a collection of brain tumor MRI scans acquired from multiple different centers under standard clinical conditions, but with different equipment and imaging protocols, resulting in a vastly heterogeneous image quality reflecting diverse clinical practice across different institutions. However, we designed the following tumor annotation protocol, in order to make it possible to create similar ground truth delineations across various annotators. For the tasks related to BraTS, only structural MRI volumes were considered (T1, T1Gd, T2, T2-FLAIR), all of them co-registered to a common anatomical template (SRI [6]) and resampled to 1mm3 . The details of the original scans are given in Table 1. Note that different native T1 scans exist, depending on whether they were 3D acquisitions, or 2D fast spin echo, or even just localizing images, and therefore not all T1 scans can be considered suitable for the task of segmentation. To our experience the T1Gd and the T2-FLAIR volumes have been the most useful to produce the ground truth segmentations. We note that radiologic definition of tumor boundaries, especially in such infiltrative tumors as gliomas, is a well-known problem. In an attempt to offer a standardized approach to assess and evaluate various tumor sub-regions, the BraTS initiative, after consultation with internationallyrecognized expert neuroradiologists, defined the following types of tumor sub-regions. However, we note that other criteria for delineation could be set, resulting in slightly different tumor sub-regions. The BraTS tumor sub-regions do not reflect strict biologic entities, but are rather image-based. For instance, the definition of the AT could simply be the regions with hyper-intense signal on T1Gd images. However, in high grade tumors, there are non-necrotic, non-cystic regions that do not enhance, but can be separable from the surrounding vasogenic edema, and represent nonenhancing infiltrative tumor. Another problem is the definition of tumor center in low-grade gliomas. In such cases, it is difficult to differentiate tumor from vasogenic edema, particularly in the absence of enhancement. It is also noteworthy that in order to produce the ground truth labels used in the provided data, we have recommended to start delineating the sub-regions of interest from the outside tumor boundaries, i.e., one should start from the manual delineation of the abnormal signal in the T2-weighted images, primarily defining the WT, then address the TC, and finally the enhancing and non-enhancing/necrotic core, possibly using semi-automatic tools

Mr. Ujjwal, thank you too for providing such a piece of valuable information.