| ObjectiveTo study the impairment of facial emotion recognition after stroke is existent or not, and the dominant hemisphere of facial emotion recognition; determining the different vascular innervation area damage appear different types of facial emotion recognition impair. The severity of stroke is relevant to facial emotion recognition, and facial emotion recognition impairment influence on post stroke depression and cognitive dysfunction.Methods1、Using prospective case-control study method, we collected patients who meet the diagnosic standard of stroke were unilateral initial stroke, and admission to Kunming General Hospital of Chengdu Military during 2013 in May to 2014 in October. In addition, we collected healthy people as the normal control group (NC), and record the patient’s general information and clinical data. Recording the lesion site and the British Oxford County Community Stroke Project typing (OCSP) by medical imaging technology, and assessing the admission HIHSS score score, mRS score, HAMD score, MoCA score, In addition, the reaction time and correct rate were recorded using the E-prime system when you identify six kinds of emotional picture.2、We follow up the patients in the stroke group, the main follow-up content is that secondary prevention forms, HAMD score, MoCAscore, death and relapse.3、Using SPSS 17.0 statistical software for data management and data analysis. Measurement data using the mean±standard deviation (mean±SD) said, count data by percentage (%) said. Comparing two groups measurement data of normal distribution by T test, if the two groups measurement data is skewed distribution, we use the Mann-Whitney U test; non normal distribution of multiple groups of measurement data were compared with Kruskal Wallis H test; the two between three groups were compared with Wilcoxon test, we must adjust the test level which is equal to 0.017, if a < 0.017 was statistically significant;Comparing the two groups of count data by chi square test, if P< 0.05, there was statistically significant. Using multivariate linear regression analysis the severity of stroke to facial emotion recognition, and analysis the impact of facial emotion recognition to post stroke depression and cognitive dysfunction.Results1-The study collected 60 cases stroke patients (30 cases of right hemisphere stroke,30 cases of left hemisphere stroke), and 30 cases of normal controls. Tw o group baseline data comparison show that most patients with hypertension in str oke group (70% vs.26.67%, P<0.05), antihypertensive drugs are more in stroke g roup (65% vs.20%, P<0.05). Comparing the rest of baseline data, the difference i s not statistically significant (P>0.05). the difference of the baseline data betwe en right and left stroke group is not statistically significant (P>0.05).2、Spearman rank correlation analysis emotion recognition accuracy betwee n stroke group and control group, six kinds of facial emotion recognition acc uracy decreased in the stroke group. Sadness (65.95±14.67 vs.75.24±15.44%, P= 0.045), surprise (76.67±15.10 vs.82.86±12.67%, P= 0.054), fear (52.86±13.4 9 vs.66.19v18.56%, P<0.001), anger (63.57±17.92 vs.72.86±20.67%, P= 0. 011), happy (93.57±8.89 vs.99.05±3.62%, P= 0.001), disgust (49.52±20.11 vs. 39.05±20.52%, P= 0.012). Except surprise, the difference of rest emotions is n o statistically significant. There are emotional recognition impairasion after strok e. The comparison show that reaction time of stroke group is prolong, sadness (2109.79±1378.95 vs.2028.88±1456.17ms, P= 0.49), surprise (2588.92±1574.77 vs.2343.83±1600.30ms, P= 0.06), fear (3078.87±1983.13 vs.2684.14±1697.51 ms, P= 0.82), anger (2421.26±1668.38 vs.2403.53±1687.98ms, P= 0.90), disgust (3041.84±1919.03 vs.2711.54±1662.12ms, P= 0.84), the difference is not statist ically significant (P>0.05). The happy emotional reaction time was prolong (8 32.68± 537.53 vs.736.01±603.22ms), the difference was statistically significant (P= 0.04).3、Comparison between three groups of six kinds of facial expression recognition, the accuracy of positive emotion is higher than negative emotion. In group NC, positive emotions:happy (99.05%)>surprise (82.86%); negative emotions:sadness (75.24%)> anger (72.86%)> fear(66.19%)> disgust (49.52%). In group RHS, positive emotions:happy(90.47%)> surprise (71.42%); negative emotions:sadness (61.43%)> anger(58.10%)> fear (47.14%)> disgust (30.48%). In group LHS, positive emotions:happy(96.67%)> surprise (81.9%); negative emotions:anger (69.05%)> sadness (61.43%)> fear(58.57%)> disgust (47.62%). There is the difference of facial emotion recognition, emotional valence effect.4、NC group compared with RHS group, RHS group for emotional recogni tion accuracy declined, sadness (61.43±13.6 vs.75.24±15.44%, P<0.001), surp rise (71.42±15.92 vs.82.86±12.67%, P<0.001), fear (47.14±13.6 vs.66.19±18. 56%, P<0.001), anger (58.10±14.97 vs.72.86±20.67%, P<0.001), happy (9 0.47±10.16 vs.99.05±3.62%, P<0.001), disgust (30.48±18.28 vs.49.52±20.11 %, P<0.001), the difference is statistically significant (P<0.017). Comparing emotional recognition accuracy between NC group and LHS group, the emotion al recognition accuracy which include sadness (70.48±14.50 vs.75.24±15.44%, P = 0.001), fear (58.57±10.84 vs.66.19±18.56%, P=0.002), and anger (69.05± 19.16 vs.72.86±20.67%, P=0.001) is drop, the difference is statistically signific ant; the difference of surprise (82.86±12.67 vs.81.90±12.40%, P=0.07), happy (99.05±3.62 vs.96.67±6.15%, P=0.26), and disgust (49.52±20.11 vs.47.62±19. 25%, P=0.019) of emotional recognition accuracy is not statistically significant. Comparison RHS group and LHS group show a statistically significant differenc e (P< 0.017), it inculdes sadness (61.43±13.60 vs.70.48±14.50%, P=0.016), s urprise (71.42±15.92 vs.81.90±12.40%, P=0.002), fear (47.14±13.60 vs.56.67 ±12.15%, P= 0.001), anger (58.10±14.97 vs.69.05±19.16%, P=0.001), happ y (90.47±10.16 vs.96.67±6.15%, P=0.012), disgust (30.48±18.28 vs.47.62±19. 25%, P< 0.001)5、According to the OCSP classification, analysis shows that no matter wh at part of the blood supply area of stroke, positive emotions (happy, surprise) is th e highest recognition accuracy, negative emotions (sadness, fear, anger, disgust) reco gnition accuracy is different. TACI group recognize accuracy to facial emotion s:happy(93.65%)>surprise (76.19%)> sadness (74.60%)>anger (57.14%)>fear (53.96%)>disgust (33.33%); PACI group recognize accuracy to facial emotions: happy (92.59%)>surprise (76.19%)>sadness (65.61%)>anger (65.08%)>f ear (51.32%)> disgust (38.1%); LACI group recognize accuracy to facial emo tions:happy (94.96%)> surprise (79.83%)> anger (72.27%)> sadness (66.39 %)>fear (57.14%)> disgust (45.38%); facial recognition accuracy of POCI gr oup:happy (90.48%)> surprise (71.43%)=sadness (71.43%)>anger (61.90%) >fear (57.14%)=disgust (57.14%)6、The linear regression analysis showed that the severity of stroke is relevant to six kinds of emotional recognition accuracy. It is negatively correlative to sadness (B=-0.06, P= 0.026) and surprise (B=-0.04, P= 0.022)7、Multiple linear regression analysis show that the recognition accuracy of happy expression(B= -0.43, P<0.001) is negatively correlative to HAMD score. The recognition accuracy of happy expression (B= 0.11, P= 0.006) is positively correlative to the MoCA score.Conclusion1、normal person and stroke recognize positive emotion accuracy is higher than negative emotion, facial emotion recognition is effected by emotional valence.2、Patients after stroke is impaired with facial emotion recognition, the right brain is dominant in emotion recognition.3、No matter what part stroke of the blood supply area, The accuracy of ha ppy emotions recognition is the highest, the accuracy of disgust emotion recogni tion is the lowest. Negative emotion recognition is low in anterior circulation st roke, positive emotion recognition is low in posterior circulation stroke.4、The emotional recognition is affected by early stroke severity; the higher accuracy of happy emotional recognition, the lower HAMD score; the more accuracy of happy emotion recognition, the less impairment of cognitive function. |
PDF Full Text Request