中国医学影像学杂志
中國醫學影像學雜誌
중국의학영상학잡지
CHINESE JOURNAL OF MEDICAL IMAGING
2013年
4期
297-300
,共4页
刘卫芳%夏翃%周震%童隆正
劉衛芳%夏翃%週震%童隆正
류위방%하굉%주진%동륭정
多发性硬化%磁共振成像%脑白质
多髮性硬化%磁共振成像%腦白質
다발성경화%자공진성상%뇌백질
Multiple sclerosis%Magnetic resonance imaging%Brain white matter
目的利用灰度差分统计法和游程长矩阵分析MRI上多发性硬化患者病灶区脑白质(LWM)、表现正常脑白质(NAWM)及健康人脑白质(NWM)纹理特征的差异.资料与方法于26例多发性硬化患者和26例健康对照者的T2WI图像中提取感兴趣区,组成LWM、NAWM和NWM 3组样本.利用灰度差分统计法和游程长矩阵提取对比度等8个纹理特征,比较3组样本各纹理特征间的差异.结果 LWM、NAWM、NWM 3组样本能量呈递增趋势(F=99.178, P<0.001),对比度、平均值和熵呈递减趋势(F=45.947、95.637、77.837, P<0.001);长游程因子和灰度不均匀度因子呈递增趋势(F=11.456、102.747, P<0.001),短游程因子和游程长不均匀度因子呈递减趋势(F=54.195、51.087, P<0.001).3组样本间两两比较, NAWM和NWM组间对比度、长游程因子和灰度不均匀度因子差异无统计学意义(P>0.05),其余差异均有统计学意义(P<0.05、P<0.001).结论纹理分析可以揭示多发性硬化患者的隐匿性病变,有助于多发性硬化的早期发现和治疗.
目的利用灰度差分統計法和遊程長矩陣分析MRI上多髮性硬化患者病竈區腦白質(LWM)、錶現正常腦白質(NAWM)及健康人腦白質(NWM)紋理特徵的差異.資料與方法于26例多髮性硬化患者和26例健康對照者的T2WI圖像中提取感興趣區,組成LWM、NAWM和NWM 3組樣本.利用灰度差分統計法和遊程長矩陣提取對比度等8箇紋理特徵,比較3組樣本各紋理特徵間的差異.結果 LWM、NAWM、NWM 3組樣本能量呈遞增趨勢(F=99.178, P<0.001),對比度、平均值和熵呈遞減趨勢(F=45.947、95.637、77.837, P<0.001);長遊程因子和灰度不均勻度因子呈遞增趨勢(F=11.456、102.747, P<0.001),短遊程因子和遊程長不均勻度因子呈遞減趨勢(F=54.195、51.087, P<0.001).3組樣本間兩兩比較, NAWM和NWM組間對比度、長遊程因子和灰度不均勻度因子差異無統計學意義(P>0.05),其餘差異均有統計學意義(P<0.05、P<0.001).結論紋理分析可以揭示多髮性硬化患者的隱匿性病變,有助于多髮性硬化的早期髮現和治療.
목적이용회도차분통계법화유정장구진분석MRI상다발성경화환자병조구뇌백질(LWM)、표현정상뇌백질(NAWM)급건강인뇌백질(NWM)문리특정적차이.자료여방법우26례다발성경화환자화26례건강대조자적T2WI도상중제취감흥취구,조성LWM、NAWM화NWM 3조양본.이용회도차분통계법화유정장구진제취대비도등8개문리특정,비교3조양본각문리특정간적차이.결과 LWM、NAWM、NWM 3조양본능량정체증추세(F=99.178, P<0.001),대비도、평균치화적정체감추세(F=45.947、95.637、77.837, P<0.001);장유정인자화회도불균균도인자정체증추세(F=11.456、102.747, P<0.001),단유정인자화유정장불균균도인자정체감추세(F=54.195、51.087, P<0.001).3조양본간량량비교, NAWM화NWM조간대비도、장유정인자화회도불균균도인자차이무통계학의의(P>0.05),기여차이균유통계학의의(P<0.05、P<0.001).결론문리분석가이게시다발성경화환자적은닉성병변,유조우다발성경화적조기발현화치료.
Purpose To analyze the texture characteristics of lesion white matter (LWM) and normal appearing white matter (NAWM) in patients with multiple sclerosis (MS) and normal white matter (NWM) using gray-scale difference method and run-length matrix. Materials and Methods Regions of interest (ROIs) were extracted on T2WI and were classified into LWM, NAWM, and NWM groups for 26 MS patients and 26 normal controls. Eight texture characteristics were extracted using gray-scale difference method and run-length matrix, and the difference of texture characteristics were compared among the three groups. Results There was increased trends of energy, long run emphasis and grey level nonuniformity among LWM, NAWM and NWM group (F=99.178, 11.456, 102.747, P<0.001), and there were decreased trends of contrast, mean value, entropy, short run emphasis and run length nonuniformity (F=45.947, 95.637, 77.837, 54.195, 51.087, P<0.001). There was no significant difference for contrast, long run emphasis or gray-level nonuniformity compared between any two groups (P>0.05), and the others indexes showed statistical difference between any two group (P<0.05). Conclusion Texture analysis can reveal the occult lesions in MS, and is helpful for the early diagnosis and therapy.