机器学习在儿童创伤后应激障碍识别及转归预测中的应用

创伤后应激障碍(PTSD)会给儿童发展带来负面效应, 其影响甚至延续至成年期。然而传统诊断方式难以做到快速、客观、准确的识别和诊断儿童PTSD, 机器学习作为一种处理大量变量和数据的新兴方法, 逐渐被应用到儿童PTSD的早期预测、识别及辅助诊断等研究中。机器学习凭借其性能、原理等方面的优势, 可被应用在儿童PTSD的识别与转归领域。相比自我报告式的诊断, 通过机器学习辅助识别和诊断儿童PTSD的过程具有效率高、客观准确、节约资源等独特优势。然而, 机器学习也在硬件成本、算法选择和预测准确度等方面存在局限性。未来研究人员需要进一步提高机器学习诊断识别儿童PTSD的准确率, 并将机器学习算法同传统诊断方法结合进行更多的探索和应用。     

机器学习在自闭症儿童早期识别和诊断领域的应用

早发现、早诊断、早干预是开展自闭症儿童教育康复工作的共识, 但传统识别和诊断方法局限及专业人员缺乏常导致自闭症儿童错失最佳干预期。为改善现状, 近年来机器学习凭借其客观准确、简便灵活等方面的优势, 逐渐被应用到自闭症的早期预测、筛查、诊断和评估过程管理中, 积累了较为丰富的成果。但是机器学习也在研究对象选取、分类数据采集和理论模型应用等方面存在局限性。未来研究应推动构建孕产期和新生儿病理生理信息追踪数据库和标准化模型分类指标体系, 同时继续优化算法, 加快智能化自闭症识别和诊断理论成果向实践转化。     

基于网络分析的抑郁症产生与演变预测

抑郁症是现代社会亟需解决的公共健康问题,预防是应对该问题最有效的方式之一。有效预防的关键在于准确识别潜在抑郁症患者,捕捉抑郁状态发生变化的预警信号,及时采取预防措施。抑郁是由多种症状相互作用而成的网络系统,该网络的结构特征和动力特征能为抑郁症发生与演变的预测提供新的理论视角和可测量的指标。以如何预测抑郁症发生与演化的关键问题为切入点,从理论的角度论述症状网络与抑郁的关系,进一步考察抑郁症状网络的拓扑结构特征、临界现象相关指标在预测抑郁发作及突变中的表现力。为增加早期预警信号在抑郁状态预测方面的准确性,未来研究应当构建更系统、全面的网络,通过使用综合的或基于机器学习的预警指标,优化抑郁状态确定方法。     

计算精神病学：抑郁症研究和临床应用的新视角

抑郁症是一种复杂而异质的精神疾病, 给全球带来沉重的疾病负担。尽管基于症状学的诊断方法已被广泛应用于各领域, 但这种方法并不利于病理机制的探讨。另外, 该诊断方法预测效度较低, 导致其难以准确评估和比较各种治疗方案的疗效。计算精神病学方法则能通过理论驱动和数据驱动两种互补的方法解决上述问题, 从而提高对抑郁症的认识、预防和治疗。理论驱动方法基于经验知识或假设, 利用计算建模方法对数据进行多水平分析; 数据驱动方法则基于机器学习算法分析高维数据, 提高抑郁症诊断和预测的准确性, 进而提高治疗的精准度。理论驱动和数据驱动方法的发展与结合, 以及人才和资源的整合, 将会更有效地推进抑郁症的防治。     

抑郁症的脑结构与功能异常:来自静息态下多模态脑影像的研究证据

传统单一模态、单一分析方法在揭示抑郁症脑机制上存在较多局限;而新近多种模态、多种分析方法的结合可在一定程度上较好地促进对抑郁症脑功能和结构的全面探索、挖掘,可以更加有效地运用和实施于早期辅助诊断、干预治疗当中。因此,本文首先简要介绍了多种模态下的脑影像指标及其分析技术,而后分别从结构及功能神经影像数据融合等方面,概述了抑郁症脑结构和功能的研究现状,发现抑郁症患者存在诸多脑区及相关环路结构及功能的异常。同时,通过对抑郁症多模态研究现状的梳理和总结,结合我们已有的相关前期研究工作,对未来抑郁症等情感障碍的进一步研究工作提出了一些思考和展望。     

抑郁症患者的自传体记忆特征及其应用自传体记忆测验的元分析

研究目的在于探讨抑郁症患者自传体记忆的特征以及自传体记忆测验(the autobiographical memory test, AMT)在该领域的应用情况。通过文献检索, 搜集了18篇应用AMT作为测评工具的抑郁症自传体记忆研究, 抑郁症患者566人, 正常对照组457人。元分析研究发现, 和正常对照组比较, 抑郁症组的具体性记忆减少, 概括化记忆增多, 反应迟缓。目前对这种现象的解释主要有功能性回避模型、认知执行受损模型和沉思模型; 年龄、抑郁情绪以及AMT测试程序对测试结果有明显影响; 发表偏倚和敏感性分析显示存在发表偏倚, 但稳定性较好。AMT在抑郁症研究存在灵敏度不足等局限, 作者从AMT程序、研究设计等方面提出了改善建议。     

基于社交媒体数据的心理指标识别建模:机器学习的方法

心理指标识别建模是基于海量数据结合计算机机器学习算法识别心理特征的一种新兴方式。由于传统纸笔测量方式所存在的诸多限制,本文对基于社交媒体数据的心理建模方法及应用于心理测量的可行性进行综述,介绍了特征及提取方法、常用机器学习算法以及应用场景,并对心理指标识别建模的优势和不足进行了总结与展望。该测量方法基于社交媒体数据,相比自我报告法具有时效性高、可回溯测量、生态效度好等独特优势。然而,基于社交媒体的心理指标识别建模方法也在学习成本、硬件成本等方面存在局限性。未来研究人员需要进一步探索社会媒体信息与用户心理变量间的关联机制,并将心理指标识别模型同传统心理学研究方法结合进行更多的探索和应用。心理指标识别建模结合心理测量基本原理和计算机领域机器学习的技术,将为心理学研究打开一扇新的大门。     

抑郁症患者的语言使用模式

语言使用模式能反映心理状态和精神病理学特征。抑郁症患者与健康人群的语言使用模式存在差异, 识别抑郁症患者的语言使用模式有助于抑郁症的预测和诊断。传统的心理学研究和基于社交媒体的研究均表明, 抑郁症患者更多地使用第一人称单数代词和消极情绪词, 更少地使用第一人称复数代词和积极情绪词。基于社交媒体的研究进一步发现了一些抑郁个体日常生活中的其他语言标志。建议未来的研究进一步确认更具抑郁特异性的语言标志, 并进一步探索语言标志与抑郁症状间的理论联系。     

抑郁症的整合情绪记忆模型述评

多种早期情绪记忆模型尝试揭示抑郁症的认知致病原因和机制。但是,抑郁症的致病因素存在多样性,而各个模型只侧重于某个方面,难以全面、系统、准确对其进行的诠释。整合情绪记忆模型将记忆建构、自我图式、自传体记忆、内隐记忆等认知结构进行系统的整合,从编码和提取的角度对抑郁症的的发生、持续和治疗提供新的解释,为抑郁症的预防、治疗提供了新思路。     

抑郁症认知治疗理论及实践进展

认知治疗分化为认知行为与认知分析治疗两个流派。抑郁的贝克认知模型、归因模型、自我价值关联模型以及抑郁的注意过程等认知理论为认知治疗奠定了基础。认知行为治疗、认知行为分析系统心理治疗以及基于冥想的认知治疗等已经在实践中被较有力的临床证据检验。认知治疗在抑郁症的治疗中已显示出广泛的应用前景。     

抑郁症的脑功能异常:来自静息态功能磁共振成像的证据

静息态功能磁共振成像是指在静息状态下测量的BOLD信号,即受试者安静地躺在扫描仪中,不给受试者任何特定的任务,受试者也不用做任何反应,此时受试者的大脑活动处于自发状态。通过使用该技术可以为抑郁症发作的临床现象提供神经影像学依据,以期为将来抑郁症的治疗提供生物标记。因此,本文综述了大量抑郁症患者在静息态脑功能方面的差异研究,发现了单、双相抑郁症,首、复发抑郁症,早发性、晚发性抑郁症,难治、非难治性抑郁症等不同类型的抑郁症在包括局部一致性(Re Ho)和低频振幅(ALFF)在内的功能分化以及包括功能连接密度(FCD)、功能同伦(VMHC)、复杂网络(Complex Network)和ROI功能连接在内的功能整合两大指标的改变。     

Screening for and Preventing Perinatal Depression

New York City (NYC) public hospitals recently mandated that all pregnant women be screened for depression, but no funds were allocated for screening or care coordination/treatment, and research suggests that unfunded mandates are not likely to be successful. To address this, we implemented an on-site depression prevention intervention (NYC ROSE) for positive depression screens among pregnant, mostly Black and Hispanic, lower-income women in one public hospital. In this paper, we used Aarons’ implementation model to describe the successes and challenges of screening and intervention. Patient tracking sheets and electronic medical records were abstracted. Key informant interviews and an informal focus group were conducted, and staff observations were reviewed; common implementation themes were identified and fit into Aarons’ model. We found that a lack of funding and staff training, which led to minimal psychoeducation for patients, were outer context factors that may have made depression screening difficult, screening results unreliable, and NYC ROSE enrollment challenging. Although leadership agreed to implement NYC ROSE, early involvement of all levels of staff and patients would have better informed important inner context factors, like workflow and logistical/practical challenges. There was also a mismatch between the treatment model and the population being served; patients often lived too far away to receive additional services on site, and economic issues were often a higher priority than mental health services. Screening and interventions for perinatal depression are essential for optimal family health, and a detailed, thoughtful and funded approach can help ensure effectiveness of such efforts.     

A novel machine learning approach for early detection of hepatocellular carcinoma patients

Liver cancer is quite common type of cancer among individuals worldwide. Hepatocellular carcinoma (HCC) is the malignancy of liver cancer. It has high impact on individual’s life and investigating it early can decline the number of annual deaths. This study proposes a new machine learning approach to detect HCC using 165 patients. Ten well-known machine learning algorithms are employed. In the preprocessing step, the normalization approach is used. The genetic algorithm coupled with stratified 5-fold cross-validation method is applied twice, first for parameter optimization and then for feature selection. In this work, support vector machine (SVM) (type C-SVC) with new 2level genetic optimizer (genetic training) and feature selection yielded the highest accuracy and F₁-Score of 0.8849 and 0.8762 respectively. Our proposed model can be used to test the performance with huge database and aid the clinicians.     

Depression Screening in Medically Ill Homecare Elderly

PURPOSE: This article provides data on a depression screening model (HOME) in acute home health care designed to detect clinical depression among medically ill homebound older patients. The model was developed to address the lack of mental health services in home health care settings and to specifically improve geriatric depression screening as part of routine care. Authors report on the concordance of homecare and research interview ratings of depression in older homecare patients. DESIGN AND METHODS: Using a prospective cohort design, data were collected from 289 elderly patients, aged 65 and older, from a large home health care agency to examine depression, cognitive functioning, medical comorbidity, functional status, and social isolation. Research interviews used the depression module of the structured clinical interview for DSM (SCID). RESULTS: The overall prevalence of major depression was 5.7 percent according to both homecare and research raters. The prevalence of subthreshold depressive disorder was 16.4 percent as reported by research raters. Observed agreement was 73 percent and kappa agreement was 0.42, indicating a fair to moderate agreement. We identified patient characteristics that may influence the accuracy of homecare worker estimates of depressive symptoms. IMPLICATIONS: Findings suggest that depression continues to be underdetected in medically ill homebound elderly patients. Ongoing training in depression screening methods, patient follow-up interviews, and appropriate referral would improve care of depressed elderly homecare patients.     

人工智能辅助的自闭症早期患者的筛查与诊断

自闭症谱系障碍(Autistic Spectrum Disorders, ASD)的症状早在婴幼儿期就会显现, 越早发现, 越早干预, 治疗效果越好。传统自闭症早期筛查与诊断在评估方法、流程上存在局限, 无法满足大规模筛查和诊断需求。随着人工智能技术的快速发展, 使用智能化方法进行自闭症早期大规模无感筛查与诊断逐渐成为可能。近10年间, 国内外对自闭症智能化识别方法的探索在经典任务行为、面部表情和情绪、眼动、脑影像、运动控制和运动模式、多模态6个领域积累了丰富的研究成果。未来研究应围绕构建国内自闭症早期智能医学筛查与诊断体系, 开发针对婴幼儿患者的筛查工具, 构建融合多模态数据的自闭症婴幼儿智能化识别模型, 建立结合脑影像技术的自闭症精细化诊断方法等方面来开展。     

What is morally at stake when using algorithms to make medical diagnoses? Expanding the discussion beyond risks and harms

In this paper, we examine the qualitative moral impact of machine learning-based clinical decision support systems in the process of medical diagnosis. To date, discussions about machine learning in this context have focused on problems that can be measured and assessed quantitatively, such as by estimating the extent of potential harm or calculating incurred risks. We maintain that such discussions neglect the qualitative moral impact of these technologies. Drawing on the philosophical approaches of technomoral change and technological mediation theory, which explore the interplay between technologies and morality, we present an analysis of concerns related to the adoption of machine learning-aided medical diagnosis. We analyze anticipated moral issues that machine learning systems pose for different stakeholders, such as bias and opacity in the way that models are trained to produce diagnoses, changes to how health care providers, patients, and developers understand their roles and professions, and challenges to existing forms of medical legislation. Albeit preliminary in nature, the insights offered by the technomoral change and the technological mediation approaches expand and enrich the current discussion about machine learning in diagnostic practices, bringing distinct and currently underexplored areas of concern to the forefront. These insights can contribute to a more encompassing and better informed decision-making process when adapting machine learning techniques to medical diagnosis, while acknowledging the interests of multiple stakeholders and the active role that technologies play in generating, perpetuating, and modifying ethical concerns in health care.

     

Insufficient Referral for Genetic Counseling in the Management of Hereditary Haemochromatosis in Portugal: A Study of Perceptions of Health Professionals Requesting HFE Genotyping

There is a general consensus that HFE- related Hereditary Haemochromatosis (HFE-HH) should be diagnosed at early stages in pre-symptomatic individuals, in order to prevent the most severe consequences of iron overload. In Portugal, despite an increasing number of requests for genetic diagnosis of this rare disease, there is not a corresponding increase in requests for genetic counselling. The objective of the present study was to evaluate physicians’ main motivations for requesting HFE genotyping or genetic counselling for HFE-HH. We assessed current medical practices regarding family testing and diagnosis and discuss whether these can be improved in order to increase the effectiveness of disease prevention. Our results show there is a general lack of knowledge about the selection of patient cases that should be sent for genetic counseling or for molecular testing of HFE-HH by physicians (especially by general practitioners). The lack of family-based screening may indirectly compromise the efficiency of disease prevention in terms of early diagnosis and treatment. We concluded it is necessary to circulate more information about Hereditary Haemochromatosis among health professionals in order to improve strategies for its early diagnosis.     

关于气质类型与相对学业不良的相关研究

本研究考察了226名大、中学生的气质类型与学业成就的关系,结果发现：无论哪个学习年限段的学生的气质类型中,与胆汁质、抑郁质有关的气质类型与学业不良有关。并认为其主要以任务坚持性和社会灵活性缺乏为特征影响学业不良学生的学习成就。     

Adherence predictors in an Internet-based Intervention program for depression

Internet-delivered psychotherapy has been demonstrated to be effective in the treatment of depression. Nevertheless, the study of the adherence in this type of the treatment reported divergent results. The main objective of this study is to analyze predictors of adherence in a primary care Internet-based intervention for depression in Spain. A multi-center, three arm, parallel, randomized controlled trial was conducted with 194 depressive patients, who were allocated in self-guided or supported-guided intervention. Sociodemographic and clinical characteristics were gathered using a case report form. The Mini international neuropsychiatric interview diagnoses major depression. Beck Depression Inventory was used to assess depression severity. The visual analogic scale assesses the respondent’s self-rated health and Short Form Health Survey was used to measure the health-related quality of life. Age results a predictor variable for both intervention groups (with and without therapist support). Perceived health is a negative predictor of adherence for the self-guided intervention when change in depression severity was included in the model. Change in depression severity results a predictor of adherence in the support-guided intervention. Our findings demonstrate that in our sample, there are differences in sociodemographic and clinical variables between active and dropout participants and we provide adherence predictors in each intervention condition of this Internet-based program for depression (self-guided and support-guided). It is important to point that further research in this area is essential to improve tailored interventions and to know specific patients groups can benefit from these interventions.