Apparent diffusion coefficient values and non-homogeneity of diffusion in brain tumors in diffusion-weighted MRI

Abstract

Background

The values that have been received from apparent diffusion coefficient (ADC) maps of diffusion-weighted magnetic resonance imaging (DW-MRI) might play a vital role in evaluating tumors and their grading scale.

Purpose

To investigate the predictive role of this heterogeneity in brain tumor pathologies and its correlation with Ki-67.

Material and Methods

A total of 124 patients with brain tumors underwent brain MRI with gadolinium injection. ADC and standard deviation of each lesion have been obtained from manual localization of the region of interest on the ADC map. A receiver operating characteristic analysis was conducted to determine the minimum cut-off values of the mean ADC and mean standard deviation of ADC maps having the highest sensitivity and specificity to differentiate high-grade and low-grade tumors.

Results

Mean ADC values in the region of interest were significantly lower for malignant tumors (grade IV and metastasis) than grade I brain tumors, while a higher mean standard deviation was observed. In a more detailed comparison of tumor groups, the mean standard deviation of the ADC for glioblastoma multiform was significantly higher than meningioma grade I (P < 0.001) and metastasis was significantly higher than grade III and IV astrocytic tumors (P = 0.004). The analysis of Ki-67 proliferation index and mean ADC values in gliomas showed a significant inverse correlation between the parameters (r = –0.0429, P < 0.001) and direct correlation between Ki-67 and mean standard deviation of the ADC (r = 0.551, P < 0.001). As an index for the ADC to differentiate high-grade and low-grade tumors, the cut-off values of 1.40*10⁻³ mm²/s for mean ADC and 45*10⁻³ mm²/s for mean standard deviation have the highest combination of sensitivity, specificity, and area under the curve.

Conclusion

The mean value and standard deviation of the ADC could be considered for differentiating between low-grade and high-grade brain tumors, as two available non-invasive methods.

Keywords

Brain tumor apparent diffusion coefficient diffusion-weighted magnetic resonance imaging MRI magnetic resonance imaging Ki-67 antigen

Introduction

Gliomas are the most primary brain tumor in adults and make up almost 80% of malignant brain tumors (1). They have been classified into four grades according to the histological characteristics of the tumor according to the World Health Organization (WHO) classification of the central nervous system tumors (2). Grades I and II are known as low-grade tumors and grades III and IV as high-grade tumors. Prognosis and biological behavior of tumors differ between grades (3). Overall prognosis differs with the current treatment in the two groups. While new imaging modalities have been developed, the differentiation of these grades still remains a challenge (4).

Although magnetic resonance imaging (MRI) has been widely used for differentiating healthy cells from tumors and exhibiting the location of brain tumors, they are not always trustworthy in providing the differentiation of high-grade and low-grade tumors, which occasionally leads to an incorrect grading (5).

Conventional MRI has been used to obtain anatomic information; physiologic and metabolic information has also been provided by advanced MRI techniques (6). Diffusion-weighted imaging (DWI) evaluates the random diffusion of water molecules in biological tissues. It is used to evaluate malignancy (7), metastasis (8), and response to therapy (9).

The apparent diffusion coefficient (ADC) has been obtained at quantitative analysis of DWI. High-grade lesions elevate cellular proliferation and reduce extracellular space. However, the diffusion of water molecules are limited in the extracellular space. As a result of the inverse correlation of ADC with tumor cellularity, the ADC map has been increasingly used to differentiate high-grade from low-grade tumors (10).

There is high cellular pleomorphism in high-grade malignant brain tumors. Hypothetically, restricted diffusion is heterogeneous in high-grade brain tumors. Thus, ADC value measurements in viable and active regions of tumors have determined the tumor grade. Previous reports showed the usefulness of the ADC role on the differentiation of high-grade and low-grade gliomas in the supratentorial region (11), benign and malignant meningiomas (12), as well as distinguishing between meningiomas and gliomas, meningiomas, and schwannoma (13,14). However, the role of the standard deviation (SD) of ADC is unclear.

Ki-67 is a proliferation marker expressed during the proliferative cell cycle. The Ki-67 proliferation index increases in malignant cells (15). Previous studies have shown the correlation between Ki-67 and brain tumor malignancy class is useful in the evaluation of clinical course and patient outcome (16,17).

Therefore, the aim of this study was to evaluate the mean ADC in brain tumors, examine the mean SD of ADC as a predictive value of diffusion heterogeneity for tumor grading, and evaluate the role of cut-off ADC values in the differentiation of high-grade and low-grade tumors. We also evaluated the correlation between Ki67 proliferation index and ADC values of glioma as a most frequent malignant brain tumor.

Material and Methods

Patients

After approval from the local medical ethics committee (registration code IR.TBZMED.REC.1396.923) and written informed consent and from patients or their family members before inclusion in the cross-sectional study, consecutive individuals were recruited between January 2016 and June 2017. The sample size was calculated at 124 according to previous studies and considering 80% power, 5% type I error rate, and 20% type II error rate. The inclusion criteria were individuals with features of brain masses on imaging modalities and clinical findings that had pathologically proven brain tumors, except patients with metastasis. Exclusion criteria were a contraindication for MRI including patients with an implanted pacemaker, cochlear implant, aneurysm clips, and other metallic foreign bodies. Patients with claustrophobia were also excluded. All patients entered the imaging department in a tertiary referral center of Tabriz for MRI examination, as part of pre-treatment evaluation.

A total of 124 participants were enrolled (58 men/boys [46.8%], 66 women/girls [53.2%]; age range = 4–88 years; mean age ± SD = 45.82 ± 18.68).

MRI processing

MRI was accomplished on a 1.5-T MR system (MAGNETOM Avanto, Siemens, Erlangen, Germany) and a circularly polarized head coil was as part of the radiological assessment. Cotton pads have been used to reduce the motion artifacts during MRI. All patients who underwent MRI in the prone position included these sequences: transverse T1-weighted (T2W) images (unenhanced and contrast-enhanced); transverse T2-weighted (T2W) images (unenhanced); transverse FLAIR; and transverse DWI (unenhanced).

A standard dose (0.1 mmol/Kg) of gadodiamide (GE Healthcare Ireland, Cork, Ireland) was administered for contrast-enhanced images. The following transverse T1W spin-echo MR sequence parameters were used: repetition time/echo time (TR/TE) = 380/8.7 ms; slice thickness = 6 mm; field of view (FOV) = 230 × 183 mm; and image matrix = 204 × 256. Transverse T2W spin-echo MR sequence parameters were obtained with the following parameters: TR/TE = 3000/117 ms; slice thickness = 6 mm; FOV = 230 × 189 mm; image matrix = 230 × 384.

The transverse FLAIR images were performed with the following parameters: TR/TE = 9370/85 ms; TI = 2220 ms; slice thickness = 6 mm; FOV = 230 × 172 mm; image matrix = 163 × 256.

Transverse DWI was achieved using a single-shot T2W echo-planar spin-echo sequence in b-values of 0 and 1000 s/mm². The following DW-MRI parameters were used: TR/TE = 3000/98 ms; slice thickness = 7.5 mm; FOV = 310 × 310 mm; image matrix = 192 × 192.

Syngo.MR General Engine (Siemens, Erlangen, Germany), an imaging acquisition and post-processing software for MAGNETOM device, was used to determine the ADC values. ADC was calculated using manual localization of a region of interest (ROI) in the tumor on an ADC map by two expert neuroradiologists. A value of Cohen’s Kappa was used to determine the level of agreement between two neuroradiologists.

In patients with contrast-enhanced tumors, ROIs were placed at the site of enhanced lesions at T1W images; in those with weakly enhanced or completely unenhanced tumors, ROI were selected after detecting the tumor as a hyperintense area on FLAIR images. Hemorrhagic lesions have been classified as hyperintense areas in unenhanced T1W images. Cystic lesions have been classified as hyperintense areas in T2W images and hypointense areas in FLAIR images. ROI were placed in the solid parts of tumor avoiding cystic, necrotic, and hemorrhagic components (Figs. 1 and 2). The neuroradiologists were blind to the patients’ pathologic grading.

Fig. 1.

A 68-year-old man with glioblastoma. (a) FLAIR image shows a large mass with peritumoral edema at right temporal lobe. (b) Contrast-enhanced T1W MR image shows ring-enhancing of mass. (c) DW images show diffusional properties. (d) ADC map shows a ROI.

Fig. 2.

A 47-year-old man with lung cancer metastasis. (a) FLAIR image shows multiple hemorrhagic masses with extensive vasogenic edema. (b) Contrast-enhanced T1W MR image shows intense enhancement of masses. (c) DW image shows diffusional properties. (d) ADC map shows ROIs.

Immunohistochemistry

For the species with glioma diagnosis, the original hematoxylin and eosin (H&E) and immunohistochemistry slides were used for the determination of the Ki-67 proliferation index. Glioma proliferation indexes were reported as the percentage of tumor cell nuclei labeled with the Ki-67 monoclonal antibody in formalin-fixed paraffin tissue sections. The fields with the highest number of positively labeled cells were selected for counting.

The Ki-67 proliferation index values compared between high-grade and low-grade glioma and the correlation between ADC and Ki67 proliferation in glioma have been examined.

Statistical analysis

All individuals were classified into several groups according to their pathological grade. Age, gender, mean SD of ADC, and mean ADC were measured for each group. ANOVA and Chi-square tests were applied in order to compare age and gender, respectively, between the groups. MR-related variables including ROI area in ADC map, ROI area in FLAIR, and mean and SD of ADC in ADC map were analyzed using the Kruskal–Wallis test.

A receiver operating characteristic (ROC) analysis was conducted to determine the minimum cut-off values of ADC which had the highest sensitivity and specificity to differentiate high-grade tumors from low-grade tumors.

Mann–Whitney U test was conducted to compare the mean and SD of ADC between tumor groups. Multivariate regression analysis was performed for ADC and SD values, as dependent variables, and WHO grade groups to determine any significant relationship. Spearman’s correlation coefficient was performed to evaluate correlation between ADC parameters and immunohistochemistry findings.

All statistical analyses were performed using IBM SPSS software (version 24; SPSS, Chicago, IL, USA). P values ≤ 0.05 were considered significant.

Results

Based on WHO grading system, 37 (29.8%) patients were grade I, 10 (8.1%) grade II, 9 (7.3%) grade III, 28 (22.6%) grade IV, 29 (23.4%) metastasis, and 11 (8.9%) other types of brain tumors. Patient characteristics and imaging variables of each tumor grade are shown in Table 1.

Table 1.

Patient characteristics and imaging-related variables.

Variables	Pathologic grade						P value
Variables	I	II	III	IV	Metastasis	Other tumors	P value
Age (years)	47.2 ± 20.3	37.3 ± 14.4	53.3 ± 20.8	47.1 ± 19.8	53.8 ± 11.9	51.2 ± 17.3	>0.05
Gender
Female (n (%))	23 (62.2)	5 (50)	5 (55.6)	12 (42.9)	12 (41.4)	4 (36.4)
Male (n (%))	14 (37.8)	5 (50)	4 (44.4)	16 (57.1)	17 (58.6)	7 (63.6)	0.05
ROI of ADC map (AUC) (mean ± SD)	3.66 ± 2.49	3.78 ± 2.81	4.46 ± 3.09	2.94 ± 2.8	3.33 ± 3.09	2.86 ± 3.46	>0.05
ROI of FLAIR (mean ± SD)	1.27 ± 1.18	1.03 ± 0.78	3.98 ± 5.33	0.81 ± 0.83	0.79 ± 0.78	1.03 ± 1.22	>0.05
Mean ADC (mean ± SD)	1.28 ± 0.29	1.27 ± 0.40	1.15 ± 0.29	0.96 ± 0.77	0.98 ± 0.27	1.02 ± 0.98	<0.05
Signal intensity (mean ± SD)	0.48 ± 0.12	0.48 ± 0.08	0.48 ± 0.16	0.54 ± 0.07	0.46 ± 0.12	0.43 ± 0.06	>0.05
SD ADC (mean ± SD)	25.8 ± 20.9	54.8 ± 54.8	66.7 ± 34.9	74.8 ± 29.7	79.2 ± 29.2	45.8 ± 29.4	<0.05

AUC, area under curve; ADC, apparent diffusion coefficient; SD, standard deviation; ROI, region of interest.

There were no significant differences between the groups in the (P = 0.183), gender (P = 0.477), ROI area in ADC map (P = 0.695), and ROI area in FLAIR (P = 0.123). Randomization of individuals was successful based on these data and the mentioned variables did not interfere with the effect of all other variables.

The value of Cohen’s Kappa was 0.86, showing an excellent agreement level between the two neuroradiologists. Mean ADC values in the ADC map were significantly higher in grade I than grade IV and metastasis (P < 0.001). Mean SD of ADC in ADC map was significantly lower in grade I than grades II, III, IV, and metastasis (P < 0.001) as well as in grade II compared to grade IV and metastasis (P < 0.001). Mean ADC for tumor groups and the comparison between groups are shown in Table 2.

Table 2.

Comparison of mean and SD of ADC between tumor groups

Valuable	Tumor groups		P value
ADC (mean ± SD)	Metastasis, GBM, medulloblastoma0.97 ± 0.22	Grade I1.22 ± 0.36	0.011
	Grade III, IV1.01 ± 0.22	Metastasis1.13 ± 0.32	0.389
	GBM1.01 ± 0.16	Meningioma1.13 ± 0.29	0.077
	GBM1.01 ± 0.16	Medulloblastoma0.76 ± 0.36	<0.001
	GBM1.01 ± 0.16	Primary lymphoma0.44 ± 0.01	<0.001
	Metastasis1.13 ± 0.32	Primary lymphoma0.44 ± 0.01	<0.001
SD (mean ± SD)	Metastasis, GBM, medulloblastoma 77.05 ± 29.34	Grade I25.03 ± 21.98	0.001
	Grade III, IV72.85 ± 30.8	Metastasis89.45 ± 31.59	0.004
	GBM84.33 ± 25.17	Meningioma27.66 ± 23.46	<0.001
	GBM84.33 ± 25.17	Medulloblastoma39.89 ± 16.15	0.001
	GBM84.33 ± 25.17	Primary lymphoma75.61 ± 0.08	0.832
	Metastasis89.45 ± 31.59	Primary lymphoma75.61 ± 0.08	0.741

ADC, apparent diffusion coefficient; SD, standard deviation; GBM, glioblastoma multiform.

According to ROC analysis, cut-off values of 1.38, 1.4, and 1.43*10⁻³ mm²/s were statistically significant for mean ADC index. The cut-off value of 1.40*10⁻³ mm²/s had the highest combination of sensitivity, specificity, and area under the curve (AUC). It can be considered an index for ADC to distinguish high-grade tumors from low-grade tumors.

The mean Ki-67 proliferation index in grades I, II, III, and IV astrocytomas were 3.41 ± 3.71, 7.23 ± 3.94, 27.38 ± 6.04, and 37.58 ± 6.95, respectively.

There was a significant statistical difference in Ki-67 values between grades II and III (P < 0.001) and also grades III and IV (P < 0.001), whereas the difference of Ki-67 values between grades I and II was not significant (P = 0.395). The analysis of the Ki-67 proliferation index and mean ADC values in gliomas showed a significant inverse correlation between the parameters (r = –0.429, P < 0.001). In addition, there was a significant direct correlation between the mean SD of ADC values and Ki-67 proliferation index (r = 0.551, P < 0.001).

The cut-off values of 35, 40, and 45*10⁻³ mm²/s were statistically significant for predicting tumor pathology grades. Because the cut-off value of 40*10⁻³ mm²/s had the highest combination of sensitivity, specificity, and AUC, it is the preferred value. It means that each mean SD value > 40*10⁻³ mm²/s will change the tumor grade.

Multivariate logistic regression analysis (Table 3) has shown that the mean ADC and mean SD of ADC variables were statistically significant in grades I and II with grade IV differentiation (P < 0.001) and the mean SD of ADC was a better predictor compared to mean ADC in the differentiation of grades. However, this test showed that the mean ADC and mean SD of ADC variables in grade III and IV differentiation were not statistically significant (P = 0.691, P = 0.132, respectively).

Table 3.

Multivariate logistic regression analysis to determine the predictive factors of the outcome of the grade of brain tumors.

Tumor grade*	Variable	P value	OR	95% CI
I	ADC	<0.001	1.002	1.001–1.003
I	SD	<0.001	0.951	0.932–0.969
II	ADC	<0.001	1.001	1.000–1.002
II	SD	<0.001	0.970	0.954–0.985
III	ADC	0.691	1	0.999–1.001
III	SD	0.132	0.987	0.971–1.004

The reference level was grade IV.

OR, odds ratio; CI, confidence interval.

Discussion

Pleomorphism can be an indirect criterion of malignancy in pathology (18). Indeed, malignant tumors have a lower internal regulation. Previous studies related to the mean of diffusion have emphasized the higher restricted diffusion, more compact onto cells, and in some cases it is an indirect sign of malignancy (19,20). Little has been known about the role of the SD on tissue diffusion (21). In this context, the administration discussed whether tumors with higher SD are more non-homogeneous, more polymorphic, and/or more malignant.

We have concluded that the mean SD of ADC can differentiate high-grade and low-grade brain tumors; it was an interesting finding of the study. To our knowledge, this is the first study that evaluates the mean SD of ADC in brain tumors. Specifically, the mean SD of ADC values might differentiate between high-grade and low-grade brain tumors, and were more sensitive than the mean of ADC values in the three groups: (i) metastasis, GBM, medulloblastoma, and grade I tumors; (ii) grade III, IV and metastasis; (iii) GBM and meningioma grade I. Analysis of the mean SD of ADC in different brain tumors indicated that the 1.43*10⁻³ value could be a divergent index between the tumors. In addition, mean ADC was more sensitive than the mean SD of ADC to differentiate between GBM and metastasis from primary brain lymphoma (PBL).

GBM showed a significant increase in the mean of SD; it can be concluded that the higher mean SD of GBM was originated from more heterogeneity in the tumor and the mean ADC cannot be used to differentiate GBM from meningioma. Comparing the mean ADC of GBM with two other malignant tumors showed that it was significantly higher than that of medulloblastoma and PBL. The mean SD was higher in GBM than that of medulloblastoma, which can be a result of a more heterogeneous tissue.

Contrary to our study, some previous authors have reported that ADC map was not a helpful parameter in the preoperative grading of gliomas (22), differentiating gliomas from metastasis and lymphoma (23,24), and evaluating the tumor extension (25). Most recent studies, however, have shown the increasing importance of ADC in differentiation and grading of brain tumors (26).

A similar retrospective study with a good sample size on ADC of brain tumors showed a significant negative relationship between ADC and astrocytoma tumors grade (27). These findings were fairly consistent with our results and have been confirmed by Catalaa et al. (28).

Based on the present study, the mean ADC value of PBL, as a high cellular tumor, was lower than metastasis. Previous studies have mentioned that high-grade tumors have lower ADC values compared to low-grade tumors (29,30) and ADC measurement can help conventional MRI to diagnose and differentiate brain tumors (31).

A recent well-designed study on the mean and minimum ADC of brain tumors revealed that there are no significant changes between high-grade gliomas and metastasis regarding intratumoral mean ADC values. The authors also showed that this relationship was significant in mean ADC minimum between two groups (32). Consistent with these reports (33,34), we have observed no statistical difference between grade III and IV brain tumors with metastasis in mean ADC. Although we did not measure the minimum of ADC, the difference between the two groups regarding the mean SD of ADC was significant. The latter also reported no significant difference in the mean ADC of grade II and III supratentorial gliomas. In standard and high b-values, the mean ADC was higher in grade II than that of grade IV. However, it was higher in grade III than grade IV using only high b-values (34).

Of note, it has been observed that in primary brain tumors and PBL, the minimum ADC of the DW sequence in GBM was lower than low-grade glioma, although this relationship has not been observed between GBM and PBL (35). On the contrary, in agreement with this, it has recently been shown that the mean ADC of PBL was significantly lower than that of GBM (26,36).

The cut-off value of 1.302*10⁻³ for differentiating glioblastoma and metastasis, with 83% sensitivity and 79% specificity, has been reported by Lee et al. (30). We found that the cut-off value of 1.40*10⁻³ with 90.5% sensitivity and 49.2% specificity had the highest combination for brain tumor differentiation.

The mean of ADC was significantly lower than other tumors in high-grade pediatric brain tumors. They reported that ADC has sufficient accuracy in differentiating different tumor grades (36). Consistent with our findings, ADC of pediatric brain tumor shows the differentiation between high-grade and low-grade brain tumors (37). A recent study displayed a multivariable logistic regression and reported intracellular and extracellular diffusion coefficient as the best predictive variable in the grading of pediatric tumors (38).

Nevertheless, several factors affect the differences of ADC values; selection methods such as using the minimum, normalized, or mean of ADC (39,40), the difference in measurement accuracy and b values (21), magnetic field strength, and difference in the ROI selection methods (41).

Entropy is a novel radiological method for identifying malignancy imaging phenotype and it is useful in determining the abnormality that differentiates the texture of abnormal lesions (42). It also provides high diagnostic accuracy for distinguishing between high-grade and low-grade gliomas and also grade III and IV (43). In comparison, the SD of ADC is easier to estimate than entropy and other parameters. Thus, the SD of ADC may be recommended as a surrogate parameter for the inhomogeneity of brain tumors.

Many reports showed similar statistically significant differences in the Ki-67 proliferation index between high-grade and low-grade astrocytomas (44,45). The association between Ki-67 proliferation and ADC is important because it predicts the behavior of brain malignancies. Wakimoto et al. (46) and Rathi et al. (47) showed a significant difference in the Ki-67 proliferation index between anaplastic astrocytoma and glioblastoma while others did not show a significant statistical difference between anaplastic astrocytoma and glioblastoma (48). Similar to the current study, Higano et al. (49) and Surov et al. (3) revealed a significant negative correlation between Ki-67 and ADC values for malignant astrocytic tumors and meningioma.

The present study had some limitations. For instance, the large 7.5-mm slice thickness was a limitation enhancing partial volume effect in the current study. Furthermore, the correlation between SD of ADC in brain tumors, link between cause and effect, divest the tumors of non-surgical treatment regarding lack of pathology data, and evidence of absence of experimental tissue analysis for brain metastasis. Herein, further research is needed to demonstrate the role of SD of ADC in brain tumor prognosis clearly.

In conclusion, the findings of this study suggest that the mean SD and mean of ADC values are effective and non-invasive indicators for the differentiation and classification of high-grade and low-grade brain gliomas.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Berens

Rutka

Rosenblum

ML.

Brain tumor epidemiology, growth, and invasion.

Neurosurg Clin N Am 1990; 1:1–18.

Louis

Ohgaki

Wiestler

, et al. The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 2007; 114:97–109.

Surov

Hamerla

Meyer

, et al. Whole lesion histogram analysis of meningiomas derived from ADC values. Correlation with several cellularity parameters, proliferation index KI 67, nucleic content, and membrane permeability. Magn Reson Imaging 2018; 51:158–162.

Murakami

Hirai

Kitajima

, et al. Magnetic resonance imaging of pilocytic astrocytomas: usefulness of the minimum apparent diffusion coefficient (ADC) value for differentiation from high-grade gliomas. Acta Radiol 2008; 49:462–467.

Suh

Kim

Jung

, et al. Diffusion-weighted imaging and diffusion tensor imaging for differentiating high-grade glioma from solitary brain metastasis: a systematic review and meta-analysis. AJNR Am J Neuroradiol 2018; 39:1208–1214.

Law

Cha

Knopp

, et al. High-grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 2002; 222:715–721.

Brown

Nagala

McLean

, et al. Multi-institutional validation of a novel textural analysis tool for preoperative stratification of suspected thyroid tumors on diffusion-weighted MRI. Magn Reson Med 2016; 75:1708–1716.

Mata-Mbemba

Zapotocky

Laughlin

, et al. MRI characteristics of primary tumors and metastatic lesions in molecular subgroups of pediatric medulloblastoma: a single-center study. AJNR Am J Neuroradiol 2018; 39:949–955.

Galban

Hoff

Chenevert

, et al. Diffusion MRI in early cancer therapeutic response assessment. NMR Biomed 2017; 30:3458.

10.

Gihr

Horvath-Rizea

Garnov

, et al. Diffusion profiling via a histogram approach distinguishes low-grade from high-grade meningiomas, can reflect the respective proliferative potential and progesterone receptor status. Mol Imaging Biol 2018; 20:632–640.

11.

Fan

Deng

, et al. Usefulness of diffusion/perfusion-weighted MRI in patients with non-enhancing supratentorial brain gliomas: a valuable tool to predict tumour grading? Br J Radiol 2006; 79:652–658.

12.

Surov

Ginat

Sanverdi

, et al. Use of diffusion weighted imaging in differentiating between maligant and benign meningiomas. A multicenter analysis. World Neurosurg 2016; 88:598–602.

13.

Hong

, et al. Radiological indeterminate vestibular schwannoma and meningioma in cerebellopontine angle area: differentiating using whole-tumor histogram analysis of apparent diffusion coefficient. Int J Neurosci 2017; 127:183–190.

14.

Svolos

Tsolaki

Kapsalaki

, et al. Investigating brain tumor differentiation with diffusion and perfusion metrics at 3T MRI using pattern recognition techniques. Magn Reson Imaging 2013; 31:1567–1577.

15.

Jiang

Chen

, et al. Ki67 is a promising molecular target in the diagnosis of cancer (review). Mol Med Rep 2015; 11:1566–1572.

16.

Stoyanov

Dzhenkov

Kitanova

, et al. Correlation between Ki-67 Index, World Health Organization grade and patient survival in glial tumors with astrocytic differentiation. Cureus 2017; 9:e1396.

17.

Surov

Ginat

Lim

, et al. Histogram analysis parameters apparent diffusion coefficient for distinguishing high and low-grade meningiomas: a multicenter study. Transl Oncol 2018; 11:1074–1079.

18.

Homma

Fukushima

Vaccarella

, et al. Correlation among pathology, genotype, and patient outcomes in glioblastoma. J Neuropathol Exp Neurol 2006; 65:846–854.

19.

Karavaeva

Harris

Leu

, et al. Relationship between [18F]FDOPA PET uptake, apparent diffusion coefficient (ADC), and proliferation rate in recurrent malignant gliomas. Mol Imaging Biol 2015; 17:434–442.

20.

Huang

Chen

Zhang

, et al. Intravoxel incoherent motion diffusion-weighted imaging of resectable oesophageal squamous cell carcinoma: association with tumour stage. Br J Radiol 2018; 91:20170421.

21.

Nilsen

Fangberget

Geier

, et al. Quantitative analysis of diffusion-weighted magnetic resonance imaging in malignant breast lesions using different b value combinations. Eur Radiol 2013; 23:1027–1033.

22.

Lam

Poon

Metreweli

Diffusion MR imaging in glioma: does it have any role in the pre-operation determination of grading of glioma?

Clin Radiol 2002; 57:219–225.

23.

Stadnik

Chaskis

Michotte

, et al. Diffusion-weighted MR imaging of intracerebral masses: comparison with conventional MR imaging and histologic findings. AJNR Am J Neuroradiol 2001; 22:969–976.

24.

Rollin

Guyotat

Streichenberger

, et al. Clinical relevance of diffusion and perfusion magnetic resonance imaging in assessing intra-axial brain tumors. Neuroradiology 2006; 48:150–159.

25.

Neska-Matuszewska

Bladowska

Sasiadek

, et al. Differentiation of glioblastoma multiforme, metastases and primary central nervous system lymphomas using multiparametric perfusion and diffusion MR imaging of a tumor core and a peritumoral zone-Searching for a practical approach. PLoS One 2018; 13:e0191341.

26.

Yamasaki

Kurisu

Satoh

, et al. Apparent diffusion coefficient of human brain tumors at MR imaging. Radiology 2005; 235:985–991.

27.

Bulakbasi

Kocaoglu

Ors

, et al. Combination of single-voxel proton MR spectroscopy and apparent diffusion coefficient calculation in the evaluation of common brain tumors. AJNR Am J Neuroradiol 2003; 24:225–233.

28.

Catalaa

Henry

Dillon

, et al. Perfusion, diffusion and spectroscopy values in newly diagnosed cerebral gliomas. NMR Biomed 2006; 19:463–475.

29.

Calli

Kitis

Yunten

, et al. Perfusion and diffusion MR imaging in enhancing malignant cerebral tumors. Eur J Radiol 2006; 58:394–403.

30.

Lee

terBrugge

Mikulis

, et al. Diagnostic value of peritumoral minimum apparent diffusion coefficient for differentiation of glioblastoma multiforme from solitary metastatic lesions. AJR Am J Roentgenol 2011; 196:71–76.

31.

Caravan

Ciortea

Contis

, et al. Diagnostic value of apparent diffusion coefficient in differentiating between high-grade gliomas and brain metastases. Acta Radiol 2018; 59:599–605.

32.

Darwiesh

AMN

Abd-El Maboud

Khalil

AMR

, et al. Role of magnetic resonance spectroscopy & diffusion weighted imaging in differentiation of supratentorial brain tumors. Egyptian Journal of Radiology and Nuclear Medicine 2016; 47:1037–1042.

33.

Cihangiroglu

Ozturk-Isik

Firat

, et al. Preoperative grading of supratentorial gliomas using high or standard b-value diffusion-weighted MR imaging at 3T. Diagn Interv Imaging 2017; 98:261–268.

34.

Hino-Shishikura

Tateishi

Shibata

, et al. Tumor hypoxia and microscopic diffusion capacity in brain tumors: a comparison of (62)Cu-Diacetyl-Bis (N4-Methylthiosemicarbazone) PET/CT and diffusion-weighted MR imaging. Eur J Nucl Med Mol Imaging 2014; 41:1419–1427.

35.

Choi

Kim

Ahn

, et al. Glioma grading capability: comparisons among parameters from dynamic contrast-enhanced MRI and ADC value on DWI. Korean J Radiol 2013; 14:487–492.

36.

Vajapeyam

Brown

Johnston

, et al. Multiparametric analysis of permeability and ADC histogram metrics for classification of pediatric brain tumors by tumor grade. AJNR Am J Neuroradiol 2018; 39:552–557.

37.

Sui

Wang

Liu

, et al. Differentiation of low- and high-grade pediatric brain tumors with high b-value diffusion-weighted MR imaging and a fractional order calculus model. Radiology 2015; 277:489–496.

38.

Burrowes

Fangusaro

Nelson

, et al. Extended diffusion weighted magnetic resonance imaging with two-compartment and anomalous diffusion models for differentiation of low-grade and high-grade brain tumors in pediatric patients. Neuroradiology 2017; 59:803–811.

39.

Chen

Lou

, et al. Diagnostic value of minimum apparent diffusion coefficient values in prediction of neuroepithelial tumor grading. J Magn Reson Imaging 2010; 31:1331–1338.

40.

Pierce

Provenzale

JM.

Evaluation of apparent diffusion coefficient thresholds for diagnosis of medulloblastoma using diffusion-weighted imaging.

Neuroradiol J 2014; 27:63–74.

41.

Haradome

Usui

, et al. Orbital lymphoproliferative disorders (OLPDs): value of MR imaging for differentiating orbital lymphoma from benign OPLDs. AJNR Am J Neuroradiol 2014; 35:1976–1982.

42.

Dercle

Ammari

Bateson

, et al. Limits of radiomic-based entropy as a surrogate of tumor heterogeneity: ROI-area, acquisition protocol and tissue site exert substantial influence. Sci Rep 2017; 7:7952.

43.

Ryu

Choi

Park

, et al. Glioma: application of whole-tumor texture analysis of diffusion-weighted imaging for the evaluation of tumor heterogeneity. PLoS One 2014; 9:e108335.

44.

Abd El Atti

Abou Gabal

Osman

, et al. Insights into the prognostic value of DJ-1 and MIB-1 in astrocytic tumors. Diagn Pathol 2013; 8:126.

45.

Surov

Meyer

Wienke

Associations between apparent diffusion coefficient (ADC) and KI 67 in different tumors: a meta-analysis. Part 1: ADC(mean). Oncotarget 2017; 8:75434–75444.

46.

Wakimoto

Aoyagi

Nakayama

, et al. Prognostic significance of Ki-67 labeling indices obtained using MIB-1 monoclonal antibody in patients with supratentorial astrocytomas. Cancer 1996; 77:373–380.

47.

Rathi

Radotra

Khosla

VK.

Proliferative index in astrocytic tumours. Indian J Pathol Microbiol 2007; 50:754–758.

48.

Rodriguez-Pereira

Suarez-Penaranda

Vazquez-Salvado

, et al. Value of MIB-1 labelling index (LI) in gliomas and its correlation with other prognostic factors. A clinicopathologic study. J Neurosurg Sci 2000; 44:203–209; discussion 209–210.

49.

Higano

Yun

Kumabe

, et al. Malignant astrocytic tumors: clinical importance of apparent diffusion coefficient in prediction of grade and prognosis. Radiology 2006; 241:839–846.