Multiparametric study for glioma grading with FLAIR, ADC map, eADC map, T1 map, and SWI images

According to the American Cancer Society report in 2021 [1], an estimated 83,570 individuals will be diagnosed with brain, and other central nervous systems (CNS) tumors in the U.S. Incidence rates for malignant tumors are declining overall, and survival remains low [1]. Around 33% of all brain tumors are gliomas, which come from glial cells [2]. Glioma is believed to be an essential CNS tumor. Gliomas can be classified as low and high grades based on histopathological evaluation, development potential, and aggressiveness [3]. Glioma grade detection plays a vital role in treatment decisions and management [4]. The current gold standard for classifying gliomas is based on tumor morphological observations and histopathological findings. Despite its efficacy and accuracy, biopsy suffers from several significant drawbacks: tissue sampling error during the biopsy, invasive procedure, and high intra-observer variation [4,5]. In recent years, there has been an increasing interest in developing an alternative method for tumor grading.

Recent developments in medical imaging, especially in magnetic resonance imaging (MRI), have led to a renewed interest in glioma grading with MRI as a non-invasive and accurate imaging modality. In the field of MRI imaging, various image weights can be found. Anatomical and functional images are two classes of MRI images.

One of the main issues we know about glioma grading with conventional anatomic MRI image weights such as T1 and T2 images is the lack of useful physiological and functional information. Several researchers have used different functional MRI image weights for glioma grading. Several studies, for example, [[6], [7], [8]], have carried out diffusion-weighted imaging (DWI) for glioma grading. These studies have reported the apparent diffusion coefficient (ADC) values of different glioma tumors. ADC map provided important data on tumor cellularity. Several attempts [[9], [10], [11]] have been made to use susceptibility-weighted imaging (SWI) for glioma grading. Most studies in glioma grading with MRI have focused on using dynamic contrast enhancement (DCE) and susceptibility contrast enhancement (SCE) image weights. Law and co-workers [12] estimated the glioma grade with the relative cerebral blood volume (rCBV) map. Despite their clinical success, DCE and SCE have several problems in use, such as lack of awareness of referring physicians, lack of standardized and optimized perfusion MRI protocols, and lack of standardized perfusion post-processing software packages [13].

To date, there has been little agreement on what MRI image weights are suitable for glioma grading. Most studies in glioma grading with MRI images have been restricted to limited comparisons of different MRI image weights [14,15]. It has been suggested [6,7] that ADC maps image contrast is based on water molecule mobility and diffusion within the tissue. The SWI image contrast, such as magnetic field distortion by micro-bleeding and blood products, has been addressed in several investigations for glioma grading [9,16,17]. Also, intrinsic tissue factors, such as tissue T2 [18,19] and T1 [20,21] relaxation times, thought to be influencing glioma grade, have been explored in several studies. One of the main issues in our knowledge of glioma grading with MRI images is a lack of comparison of different MRI image weights performance for glioma grading. This paper is a preliminary attempt to compare the glioma grading performance of different MRI image weights such as fluid attenuation inversion recovery (FLAIR), ADC map, eADC map, T1 map, and SWI. As far as we know, this is the first time to compare the glioma grading performance of FLAIR, ADC map, eADC map, T1 map, and SWI MRI image weights. Also, this paper examines the significance of the multivariate receiver operating characteristic curve (ROC curve) for combining different image information to raise diagnostic performance for glioma grading.

留言 (0)

沒有登入
gif