临床医师头晕/眩晕规范化培训的现状调查

目的 了解头晕/眩晕诊疗的诊疗现状。 方法 对2015年9月-2016年1月全国眩晕巡讲培训的临床医师进行微信形式的问卷调查,收集数据 进行统计分析。 结果 共625例临床医生参与调查。住院医师和主治医师组对头晕/眩晕的相关知识掌握程度低于 副主任医师和主任医师组（P <0.001）。在规范化培训内容方面,仅有22.4%的医师认为自己熟练掌握 眩晕相关体格检查,60%的医生关注眩晕相关查体,57.92%的医生关注眩晕相关疾病鉴别诊断。 结论 应加强低年资住院医师和主治医师在头晕/眩晕相关知识方面的培训。     

~(11)C-METPET/CT判断脑胶质瘤复发的价值研究

目的评价11C-METPET/CT脑显像对判断脑胶质瘤复发的临床应用价值。方法28例脑胶质瘤手术切除后患者,临床怀疑复发行11C-METPET/CT显像,其中24例同期行FDG PET/CT检查,19例行头颅增强MRI检查,最后诊断根据病理检查和临床随访结果确定。结果28例患者中11C-METPET/CT显像阳性20例,其中1例假阳性;阴性8例,其中假阴性1例,11C-METPET/CT显像的敏感性、特异性和准确率分别为95%、87.5%、92.9%,FDG PET/CT分别为88.9%、50%、79.2%,增强MRI的敏感性、特异性和准确率分别为70%、42.9%、54.9%。结论评价脑胶质瘤复发,11C-METPET/CT较FDGPET/CT和增强MRI有更高的准确性和临床应用价值。     

MRI影像组学在脑胶质瘤分级中的价值研究

目的采用MRI影像组学方法对胶质瘤的高、低级别进行术前评估。方法纳入154例经病理证实的脑胶质瘤病人,其中WHO Ⅱ级(低级别胶质瘤)75例,WHO Ⅲ~Ⅳ级(高级别胶质瘤)79例,随机分为训练集和验证集各77例。应用受试者工作特征(ROC)曲线下面积(AUC)表示训练集和验证集性能。利用影像组学标签联合病理学检测构建评估高、低级别脑胶质瘤的预测模型,并采用影像组学诺模图反应测试模型。结果采用LASSO方法在388个影像组学特征中选择3个标签特征,联合病理结果进行二分类建模。诺模图显示联合影像组学标签及病理结果构建的模型图可以显著提高诊断效能。训练集中AUC达到0.850,特异性达81.8%,敏感性为77.3%;验证集中AUC达0.836,特异性达83.3%,敏感性为77.3%。LASSO构建的模型评估决策曲线高于其他模型。结论 MRI影像组学方法可在术前帮助区分脑胶质瘤的高、低级别。     

重症肌无力患者胸腺病变的CT和病理对照研究

目的　探讨胸部 CT对胸腺病变的诊断价值。方法　对 2 2例重症肌无力患者胸部 CT和胸腺病理检查结果进行对照研究。结果　胸部 CT对胸腺异常诊断的敏感性达 88.8% ,特异性为 81.8%。其中对正常胸腺、胸腺增生和胸腺瘤诊断的特异性分别为 6 6 .7%、5 0 .0 %和 91.7%。结论　胸部 CT是诊断胸腺病变的较好方法 ,其诊断的敏感性和特异性高低与胸腺病变的性质相关 ,对胸腺瘤的诊断价值最高 ,能为提示或除外胸腺瘤的诊断提供帮助 ,为临床治疗提供指导     

头颈联合64层CT血管成像诊断颅内外动脉狭窄的临床价值

目的 探讨头颈联合64层螺旋CT血管成像（64-slice spiral computed tomography angiography，64-SSCTA）诊断颅内外动脉狭窄的临床价值。方法 回顾性分析74例临床确诊为短暂性脑缺血发作（transient ischemic attack，TIA）和缺血性卒中患者行64-SSCTA及数字减影血管造影（digital subtraction angiography，DSA）影像资料。以DSA作为“金标准”，计算64-SSCTA显示血管狭窄的敏感性、特异性、准确率；并对比64-SSCTA和DSA显示血管狭窄的符合程度。结果 74例患者共有814支动脉血管接受评价，64-SSCTA显像中诊断为狭窄或闭塞的有117支，其中103支被DSA证实；有697支诊断为正常的血管，其中688支与DSA结果一致。以DSA作为“金标准”计算出64-SSCTA对患者头颈血管病变检出的敏感性、特异性及准确率分别为92.0%、98.0%和97.2%。64-SSCTA与DSA的吻合系数Kappa =0.81，P＜0.01。对比分析发现，64-SSCTA和DSA对血管狭窄检出率差异无统计学意义（χ2=0.076，P =0.439）。结论 64-SSCTA对颅内、外动脉血管狭窄诊断具有较高的敏感性、特异性和准确率，可能为颅内、外动脉血管狭窄提供一种常用的、无创的检测手段，并为颅内、外动脉狭窄的临床治疗提供较为详尽、客观、准确的影像依据。     

人工智能与脑网络组融合助力意识障碍预后预测研究

意识障碍(DOC)患者的预后预测对临床治疗及患者家属都有重要意义。但目前的预后预测模型的准确度不高,敏感性和特异性均较低。本研究在基于病因、年龄和病程作为重要预测指标的基础上,融合使用基于脑功能磁共振影像的患者脑功能网络特征,利用人工智能和机器学习算法,研发出了一个预测意识障碍患者能否苏醒的计算模型,准确率达到了88%。     

基于改进LeNet-5模型的WHOⅡ/Ⅲ级脑胶质瘤影像自动分级的临床研究

目的探讨基于改进LeNet-5模型的WHOⅡ/Ⅲ级脑胶质瘤影像自动分级的临床应用价值。方法收集经手术病理证实的98例WHOⅡ级和Ⅲ级胶质瘤患者的MRI资料;按照就诊时间顺序将前67例患者作为训练集,后31例患者作为测试集。首先,用深度学习技术及训练集的760张MRI T2WI图像,在卷积神经网络下(LeNet-5模型下)进行训练,模拟影像科医师的判断过程,从而建立了WHOⅡ/Ⅲ级脑胶质瘤影像自动分级系统。然后,对测试集的68张MRI T2WI图像进行临床验证。结果相较于原始LeNet-5模型,改进后的模型对WHOⅡ/Ⅲ级脑胶质瘤的测试准确率明显提高;当epoch=45时,正确率达到最高,epoch大于45时,趋于不变。同时改进后模型的测试错误个数均减少,总体错误率降低;表明改进后的模型对胶质瘤分级的准确率较改进前有提高。结论改进的LeNet-5网络模型能够较好地识别胶质瘤影像表现的深层特征,提高了胶质瘤影像分级的准确率,为临床诊断提供了帮助。     

脑海绵状血管瘤的影像表现

目的 探讨脑海绵状血管瘤的影像表现。方法 脑海绵状血管瘤58例全都作MRI检查,其25例曾作DSA和48例曾作CT检查。分析脑海绵状血管瘤的影像表现。结果 脑海绵状血管瘤DSA常呈阴性,CT的敏感性和特异性有欠缺,而MRI有较高的敏感性和特异性。结论 MRI对脑海绵状血管瘤有重要的临床诊断价值。     

多层螺旋CTA与DSA诊断颅内动脉瘤的应用价值及对比分析

目的探讨多层螺旋CT血管造影(CTA)及数字减影血管造影(DSA)在颅内动脉瘤诊断中的应用价值,对二者进行对比分析。方法回顾性分析经CTA、DSA双重检查并经手术证实的68例颅内动脉瘤患者临床及影像学资料。将CTA和DSA的诊断结果用SPSS13.0统计软件进行统计分析,差异显著性采用χ2检验,比较二者诊断颅内动脉瘤的准确性、敏感性和特异性。结果 CTA诊断的准确性、敏感性、特异性分别为93%、92.31%、98.72%;DSA诊断的准确性、敏感性、特异性分别为95%、94.87%、100%,二者无显著性差异。结论 CTA、DSA是截然不同的两种检查方法,都具有较高的临床应用价值,在诊断颅内动脉瘤方面各有其优势和不足,可以互补。     

基于Faster-RCNN的正常数字减影血管造影脑血管检测与时相分期研究

目的为实现对脑部数字减影血管造影(DSA)序列的自动化判读,探索在DSA影像中目标检测算法FasterRCNN对复杂血管结构鉴别的应用性。方法收集来自复旦大学附属华山医院DSA影像库2010年1月至2013年12月的正常颈内动脉正位造影图像共计388例,其中350例作为模型训练测试集(测试集)数据,38例作为独立验证集。①测试集DSA中筛选出曝光适度、显影清晰的影像共计680张,比例为8∶2。根据不同时期DSA影像的血管特征,标记不同的感兴趣区域,图片集总计标注了5类血管特征区域。搭建Faster-RCNN多目标检测网络,优化网络参数,保存最优模型。分析测试集各类血管结构的平均精度(AP)和多类别平均精度均值(mAP)。②独立验证集DSA数据依次输入模型进行血管结构检测,分析各图像血管结构的类别与出现的时间,以此为标准对每张图像的时相进行区分,从而确定每一例DSA的时相区间。将判定结果与专科医生标定的结果进行比较,计算各时期的区分准确率。结果测试集136张图片中,颈内动脉的AP为0.922、Willis环为0.991、大静脉为0.899、静脉血管为0.769、静脉窦为0.929。5类血管特征区域的多类别mAP为0.902。独立验证集中,动脉期、毛细血管期、静脉早期和静脉窦期分期准确率分别达到100%,92.1%,92.1%和78.9%。结论 Faster-RCNN算法可以分析DSA序列中的时间信息与结构信息从而对DSA影像进行自动判读,可在缩短读片时间前提下保证足够的判读准确度,为复杂脑血管的鉴别提供技术支持。     

颞叶癫痫发作间期PET-CT与VEEG检查对致癫灶的诊断价值比较

目的 比较分析颞叶癫痫发作间期正电子发射计算机断层显像（PET-CT）和视频脑电图（VEEG）检查对致癫灶的诊断价值。方法 回顾性分析2016年3月至2018年3月手术治疗的80例单侧颞叶癫痫的临床资料。另选取同期健康体检者30例作为对照。术前进行PET-CT和VEEG监测定位致痫灶，以术中监测结果为定位致痫灶的金标准。利用受试者工作特征（ROC）曲线评价PET-CT放射性分布不对称指数（AI）对颞叶癫痫致癫灶的诊断价值。结果 PET-CT、VEEG确定致癫灶的灵敏度分别为88.73%、47.89%，特异度分别为88.89%、66.67%。颞叶癫痫发作间期病灶侧AI显著高于非病灶侧（P<0.05），同时也显著高于健康体检者颞叶内侧AI（P<0.05）及颞叶外侧AI（P<0.05）。ROC曲线分析结果显示，AI=0.153诊断颞叶癫痫致癫灶的曲线下面积为0.730，95%置信区间在0.544~0.916。结论 颞叶癫痫发作间期PET-CT脑显像定位准确率优于VEEG，对手术治疗准确定位有很高的应用价值，且AI为0.153时诊断癫痫灶的效果最好。     

Factors associated with hydrocephalus after subarachnoid hemorrhage. A report of the Cooperative Aneurysm Study

Hydrocephalus is an important complication of subarachnoid hemorrhage (SAH). We analyzed several factors possibly related to hydrocephalus following SAH in 3521 patients from the International Study on the Timing of Aneurysm Surgery. Hydrocephalus was diagnosed on admission computed tomographic (CT) scans in 15% of patients and was thought to be clinically symptomatic in 13.2% of patients. There was a 5.9% overlap between these groups. Using contingency table analysis, we found the following were significantly related to clinical hydrocephalus: increasing age; preexisting hypertension; admission blood pressure measurements; postoperative hypertension; admission CT findings of intraventricular hemorrhage, a diffuse collection of subarachnoid blood, and a thick focal collection of subarachnoid blood; posterior circulation site of aneurysm; focal ischemic deficits; use of antifibrinolytic drugs preoperatively; hyponatremia; admission level of consciousness; and a low score on the Glasgow outcome scale. Using discriminate factor analysis to predict clinical hydrocephalus, the most important variables in order were the following: CT hydrocephalus, intraventricular hemorrhage, admission level of consciousness, presubarachnoid hypertension, increasing age, subarachnoid blood noted on CT scan, posterior circulation aneurysm site, and hypertension postoperatively (canonical correlation = .399). We conclude that the development of hydrocephalus after SAH is multifactorial. Factors that compromise cerebrospinal fluid circulation acutely (eg, intraventricular hemorrhage, hemorrhage from a posterior circulation site of aneurysm, and diffuse spread of subarachnoid blood) contribute to the development of acute hydrocephalus. These same factors, plus the use of antifibrinolytic drugs preoperatively, are also important in the pathogenesis of clinical hydrocephalus, perhaps by promoting subarachnoid fibrosis.(ABSTRACT TRUNCATED AT 250 WORDS)     

Prospective validation of a proposal for diagnosis and management of patients attending the emergency department for mild head injury

BACKGROUND: In mild head injury, predictors to select patients for computed tomography (CT) and/or to plan proper management are needed. The strength of evidence of published recommendations is insufficient for current use. We assessed the diagnostic accuracy and the clinical validity of the proposal of the Neurotraumatology Committee of the World Federation of Neurosurgical Societies on mild head injury from an emergency department perspective. METHODS: In a three year period, 5578 adolescent and adult subjects were prospectively recruited and managed according to the proposed protocol. Outcome measures were: (a) any post-traumatic lesion; (b) need for neurosurgical intervention; (c) unfavourable outcome (death, permanent vegetative state or severe disability) after six months. The predictive value of a model based on five variables (Glasgow coma score, clinical findings, risk factors, neurological deficits, and skull fracture) was tested by logistic regression analysis. FINDINGS: At first CT evaluation 327 patients (5.9%) had intracranial post-traumatic lesions. In 16 cases (0.3%) previously undiagnosed lesions were detected after re-evaluation within seven days. Neurosurgical intervention was needed in 71 patients (1.3%) and an unfavourable outcome occurred in 39 cases (0.7%). The area under the ROC curve of the variables in predicting post-traumatic lesions was 0.906 (0.009) (sensitivity 70.0%, specificity 94.1% at best cut off), neurosurgical intervention was 0.926 (0.016) (sensitivity 81.7%, specificity 94.1%), and unfavourable outcome was 0.953 (0.014) (sensitivity 88.1%, specificity 95.1%). INTERPRETATION: The variables prove highly accurate in the prediction of clinically meaningful outcomes, when applied to a consecutive set of patients with mild head injury in the clinical setting of a 1st level emergency department.     

Distribution of cisternal blood in patients with acute hydrocephalus after subarachnoid hemorrhage.

The distribution of cisternal blood in relation to the development of acute hydrocephalus was studied in 246 consecutive patients with aneurysmal subarachnoid hemorrhage who were admitted within 72 hours. Patients with evidence on the initial computed tomograph (CT) of subarachnoid hemorrhage caused by other than a ruptured aneurysm and patients with a negative angiography were excluded. Acute hydrocephalus (defined as a bicaudate index, measured on the initial CT or on a repeat CT within 1 week after subarachnoid hemorrhage, exceeding the 95th percentile for age) was found on the initial CT in 50 (20%) of the 246 patients and on a repeat CT in 9 other patients. Ventricular blood was found significantly more often in patients with acute hydrocephalus than in those in whom acute hydrocephalus did not develop (28 of 59 [47%] versus 58 of 187 [31%]; chi 2 = 4.634, p = 0.031). When the analysis was restricted to the 86 patients with ventricular blood, no significant differences were found in the total amount of cisternal blood and in the distribution of cisternal blood between patients with and without hydrocephalus. In contrast, among the 160 patients without ventricular blood, hydrocephalus was associated with a slightly higher total amount of cisternal blood (Wilcoxon's rank sum test, p = 0.023), and significantly more patients with acute hydrocephalus had a higher score in both ambient cisterns than patients without acute hydrocephalus (20 of 31 [65%] versus 41 of 129 [32%]; chi 2 = 10.007, p = 0.002).(ABSTRACT TRUNCATED AT 250 WORDS)     

Image level detection of large vessel occlusion on 4D-CTA perfusion data using deep learning in acute stroke

ObjectivesAutomated image-level detection of large vessel occlusions (LVO) could expedite patient triage for mechanical thrombectomy. A few studies have previously attempted LVO detection using artificial intelligence (AI) on CT angiography (CTA) images. To our knowledge this is the first study to detect LVO existence and location on raw 4D-CTA/ CT perfusion (CTP) images using neural network (NN) models.Materials and MethodsRetrospective study using data from a level-I stroke center was performed. A total of 306 (187 with LVO, and 119 without) patients were evaluated. Image pre-processing included co-registration, normalization and skull stripping. Five consecutive time-points for each patient were selected to provide variable contrast density in data. Additional data augmentation included rotation and horizonal image flipping. Our model architecture consisted of two neural networks, first for classification (based on hemispheric asymmetry), followed by second model for exact site of LVO detection. Only cases deemed positive by the classification model were routed to the detection model, thereby reducing false positives and improving specificity. The results were compared with expert annotated LVO detection.ResultsUsing a 80:20 split for training and validation, the combination of both classification and detection model achieved a sensitivity of 86.5%, a specificity of 89.5%, and an accuracy of 87.5%. A 5-fold cross-validation using the entire data achieved a mean sensitivity of 82.7%, a specificity of 89.8%, and an accuracy of 85.5% and a mean AUC of 0.89 (95% CI: 0.85-0.93).ConclusionOur findings suggest that accurate image-level LVO detection is feasible on CTP raw images.     

Psychiatric residency training in the management of emergencies

As a psychiatry resident on the Emergency Service or on call at night, the resident learns to function on several levels, as consultant, liaison, colleague, and occasionally as primary physician. At the outset, the resident needs to know that he or she is not expected to know all that yet. It is especially necessary to emphasize to a beginning resident that it is acceptable and important to ask for help. The resident is never the only physician caring for an emergency department patient. There should always be other psychiatry residents, psychiatry attending physicians, other house staff, and emergency medicine attending physicians who can help and often have an interest in each patients. The resident has an obligation to involve other physicians if there is uncertainty about the proper moves to make, and an obligation to discuss the management of a patient with the physician who initially consulted, as well as any other doctors responsible for the patient. Emergency cases can be discussed at morning rounds as well as at various seminars. The resident should also know that an interested or helpful ear can always be found among more experienced residents and attending physicians. The supervisor may present the option to the resident that he or she is available to see the patient with the resident, afterwards, or not at all, and can teach by watching and commenting, by modeling, and by supervision of the observations and reactions that the resident presents. There are a number of benefits to these models of emergency psychiatry training. In the first place, all of the drawbacks seen in present systems are addressed or avoided. The psychiatry resident retains his or her identity as a physician, working in consultation and cooperation with other physicians in a medical facility, rather than in a community clinic or psychiatric hospital. When a resident sees an emergency patient in revisit one or two days later, the resident has the opportunity to observe the effect of the psychotherapeutic intervention. One outcome of this observation is a decreased reliance on medication, and an increased use of outpatient modalities. The resident also gains an appreciation of the mental health system as he or she must work with many agencies and many levels of care.     

脑积水脑室穿刺术后继发导管相关性出血的危险因素

目的 探讨脑积水脑室穿刺术后继发导管相关性出血的危险因素。方法 回顾性分析2015年4月至2020年7月收治的187例脑积水的临床资料。结果 187例中,120例经枕角穿刺行脑室-腹腔分流术,67例经额角穿刺行脑室外引流术。术后发生导管相关性出血13例,发生率为 6.9%;其中枕角穿刺出血7例,额角穿刺出血6例;保守治疗2例,手术治疗11例;出院时GOS评分4~5分7例,2~3分4例,1分2例;出院后随访6个月,GOS评分无明显变化。多因素logistic回归分析显示,颅内压急剧下降（OR=6.39;95% CI 1.67~24.5;P=0.007）、堵管后重新置管（OR=5.45;95% CI 1.45~20.4;P=0.010）、脑室穿刺>3次（OR=10.4;95% CI 2.33~46.6;P=0.002）是导管相关性出血的独立危险因素。结论 导管相关性出血是脑积水脑室穿刺术后较为少见的、严重并发症,围手术期应综合评估,以降低术后出血概率;术后应密切观察病情并及时复查头颅CT,出血量大的病人,预后较差。     

The use of the medical team model on a consultation-liaison teaching service

This paper describes a consultation-liaison (C/L) teaching service that uses a medical team rounding model; four teams cover distinct geographic areas of the hospital, each team consisting of an attending psychiatrist, a resident, two medical students, and a psychology graduate student. Daily attending rounds on the medical/surgical wards provide prompt and direct patient care supervision for the team members, allowing for coordination of their activities and communication with the attending physician, who serves as a role model. Psychological testing can be readily integrated into the clinical setting; neurology and family practice residents can also get their psychiatric training in this setting. The medical team model of rounding is different from other models used on C/L services; its pros and cons are discussed.     

Detecting panic disorder in emergency department chest pain patients: A validated model to improve recognition

Objective

To develop and validate a detection model to improve the probability of recognizing panic disorder in patients consulting the emergency department for chest pain.

Methods

Through logistic regression analysis, demographic, self-report psychological, and pain variables were explored as factors predictive of the presence of panic disorder in 180 consecutive patients consulting an emergency department with a chief complaint of chest pain. The detection model was then prospectively validated on a sample of 212 patients recruited following the same proceduce.

Results

Panic-agoraphobia (Agoraphobia Cognitions Questionnaire, Mobility Inventory for Agoraphobia), chest pain quality (Short Form McGill Pain Questionnaire), pain loci, and gender variables were the best predictors of the presence of panic disorder. These variables correctly classified 84% of chest pain subjects in panic and non-panic disorder categories. Model properties: sensitivity 59%; specificity 93%; positive predictive power 75%; negative predictive power 87% at a panic disorder sample prevalence of 26%. The model correctly classified 73% of subjects in the validation phase.

Conclusion

The scales in this model take approximately ten minutes to complete and score. It may improve upon current physician recognition of panic disorder in patients consulting for chest pain.     

Transfer learning artificial intelligence for automated detection of atrial fibrillation in patients undergoing evaluation for suspected obstructive sleep apnoea: a feasibility study

BackgroundIndividuals with obstructive sleep apnoea (OSA) experience a higher burden of atrial fibrillation (AF) than the general population, and many cases of AF remain undetected. We tested the feasibility of an artificial intelligence (AI) approach to opportunistic detection of AF from single-lead electrocardiograms (ECGs) which are routinely recorded during in-laboratory polysomnographic sleep studies.MethodsUsing transfer learning, an existing ECG AI model was applied to 1839 single-lead ECG traces recorded during in-laboratory sleep studies without any training of the algorithm. Manual review of all traces was performed by two trained clinicians who were blinded to each other's review. Discrepancies between the two investigators were resolved by two cardiologists who were also unaware of each other's scoring. The diagnostic accuracy of the AI algorithm was calculated against the results of the manual ECG review which were considered gold standard.ResultsManual review identified AF in 144 of the 1839 single-lead ECGs (7.8%). The AI detected all cases of manually confirmed AF (sensitivity = 100%, 95% CI: 97.5–100.0). The AI model misclassified many ECGs with artefacts as AF, resulting in a specificity of 76.0 (95% CI: 73.9–78.0), and an overall diagnostic accuracy of 77.9% (95% CI: 75.9%–97.8%).ConclusionTransfer learning AI, without additional training, can be successfully applied to disparate ECG signals, with excellent negative predictive values, and can exclude AF among patients undergoing evaluation for suspected OSA. Further signal-specific training is likely to improve the AI's specificity and decrease the need for manual verification.