1. Introduction
The therapeutic approaches and prognosis for gliomas differ considerablyfordifferenttumorgrades,itishenceimportanttoaccuratelyassessgliomagradefortreatmentplanning.Inadditiontothe useofconventionalmeanvalue,severalstudieshavereportedthe use of histogram analysis ofADC in glioma grading [1,2]. How- ever only the differentiation between low grade and high grade was concerned, whereas differentiation between grade II and III remainschallenging.Alimitationin previousstudiescould be the use of multiple ROIs in a section of the tumor lesion, which may underestimate the heterogeneity of glioma cellularity. In this study, thedifferentialdiagnosticvalueofhistogramanalysisofADCsignalvaluebasedonentireregionofgradeIIandIIItumorisinvestigated,andfurtheraimedatrevealingthemicroscopicchangesof glioma in the evolution of low grade to high grade.
2. MethodFourteen patients with grade II glioma and 22 patients with grade III glioma were enrolled in this retrospective study, tumor grades were pathologically confirmed. All the participants underwent DWIonGESignaHDxt3.0TMRwholebodyscanner.ROIsthat contained the entire tumor and peripheral edema were drawn in eachsliceoftheADCmaps.Thenhistogramsvoxelwisemeasurementsoftheentiretumorvolumewereobtained.Histogramrelatedparametersincluding minintensity,max intensity,mean value,median intensity, the 10th,25th,50th,75th and 90th percentiles, range, voxel number, std deviation, variance, relative deviation, meandeviation,skewness,kurtosisanduniformitywererecorded. The obtained parameters were compared between groups through theSPSS18.0.Receiveroperatingcharacteristiccurve(ROC)was used to assess the ability of parameters between grade II and III glioma.All statistical results were P＜0.05 as statistically significant.
3. ResultsTheADC map and histogram of typical cases grade II and III glioma are shown in Figure 1.The histogram parameters of grade II andIIIandcomparisonresultsaresummarizedinTable1.Itcanbe seenthatminintensity,maxintensity,meanvalue,medianintensity,the10th,25th,50th,75thand90thpercentiles,skewness,kurtosisanduniformityaredecreasedingradeIIIthangradeII,andon thecontraryrange,voxelnumber,stddeviation,variance,relative deviation, mean deviation are increased.Among all, min intensity（p=0.045), 10th percentiles（p=0.043, voxel number（p=0.0041, std deviation（p=0.013）, skewness（p=0.021）showed significant differencebetweentwogroups.TheROCtestshowedthatMinIntensity, 10th percentiles, Voxel number, std Deviation, Skewness feature significant difference between grade II and III (Figure 2), TheAUC, cutoff value, sensitivity and specificity of the parameters are summarized in Table 2.
4. DiscussionandConclusionIn this study, the ROIs encompassed the tumor parenchyma and peripheral edema, without avoiding cystic, necrosis and hemorrhage area, which are characteristics of grade III glioma, as compared to grade II. In addition, the boundaries of the tumor lesions maynotbeclearsincethesurroundingareasmaybevascularized. Both facts support that the ROI based analysis should include the entiretumorvolumeratherthansubsections.Inthisstudy,minintensityand10thpercentilesshowedsignificantdifferencebetween grade II and III, suggesting that ADC value in low zone is more meaningful.Inotherwords,thelowerrangeofADCbetterreflects theprogressofhighercellularity.Thedifferenceofvoxelnumber between grade II and III reflects the faster growth rate and more invasion range of grade III. Standard deviation shows the level of datadispersion,higherstandarddeviationofADCindicateslarger regions of cystic, necrosis or haemorrhage. Skewness describes the symmetry of the curve distribution. Compared with grade II, theADCvalueofgradeIIIconcentrateonlowzone,thecenterof thehistogramcurvewasshiftedtoleft.Overall,itisseenthathis- togram analysis ofADC signal value based on entire tumor could providemoreinformationindifferentiationofgradeIIandIIIgli- oma. Several parameter showed superior diagnostic value.
References 1. Yusuhn Kang, Seung Hong Choi, Young-Jae Kim. Gliomas:Histram Analysis of Apparent Diffusion Coefficient Maps with Standard-orHigh-b-ValueDiffusionweightedMRImagingCorrelationwith Tumor Grade[J]. Radiology. 2011; 261(3): 883-884.
2. Lelande S, Hu, Shu luo Ning, JenniferVM. Muliti-Parametric MRIand TextureAnalysis to Visualize Spatial Histologic Heterogeneity and Tumor Extent in Glioblastoma[J]. PLOS ONE. 2015
Shesnia Salim. Whole Tumo rVolume Based Histogram Analysis of ADC for Differentiating between who Grade II and III Glioma. Annals of Clinical and Medical Case Reports 2022