ICA-R算法在静息态功能磁共振图像的应用研究

目的：针对传统独立成分分析（ICA）方法在处理功能磁共振（fMRI）数据时存在计算量大、效率低等问题,提出基于ICA-R算法处理数据。方法：将脑默认网络的幅值信息作为源信号的部分先验知识,以参考信号的形式引入到传统ICA算法提取静息态fMRI（rsfMRI）数据的默认网络。结果：ICA—R算法能够有效提取符合大脑在静息状态时的自发活动特征的脑默认网络。结论：ICA-R算法克服了传统ICA算法以分离所有的源信号为目标造成的效率低等缺点,避免了传统ICA算法中需要后续处理的步骤,提高了算法效率,并且能够准确地提取脑脑默认网络。     

心脏骤停后院内死亡机器学习预测模型的建立

目的 探索影响心脏骤停患者预后的相关因素,并通过机器学习建立一个准确、快速的预后预测模型。方法 对美国重症监护医学信息数据库(MIMIC)中1 772例18岁以上心脏骤停患者的数据进行回顾性分析,通过三种机器学习算法建立预测模型,包括逻辑回归(logistic regression, LR)、极致梯度提升(extreme gradient boosting, XGBoost)和支持向量机(support vector machine, SVM)算法,用于预测患者心脏骤停后院内病死率。计算受试者工作特征曲线下面积(area under the curve, AUC)、准确度、精确度、召回率和F1分数,以评估所建立模型的预测性能。结果 XGBoost算法的表现优于另外两种算法。XGBoost算法建立的预测模型准确度、召回率、精确度和F1分数分别为0.762、0.812、0.765和0.788。XGBoost模型的AUC大于LR和SVM模型(0.847 vs. 0.834和0.820)。XGBoost模型中最重要的前10个特征是入院24h内乳酸、格拉斯哥昏迷评分(GCS)量表、尿素氮、血糖、...     

基于Visual C++和Matlab的磁共振影像增强后处理系统

目的:随着医学影像学的飞速发展,手术导航技术的应用及脑功能等图像分析研究的不断深入,基于医学数字成像和通信标准的医学影像分析与处理也随之成为医学图像处理领域中的热点。为便于科研人员研究相应的磁共振图像局部增强等后处理算法及进行图像分析,提出一种基于VisualC 和Matlab的磁共振影像增强后处理研发平台。方法:对北京协和医院放射科2005-01/10获取的部分磁共振医学数字成像和通信图像利用灰度扩展进行全局增强,利用基于数学形态学方法进行局部增强算法的研究。在程序实现上使用Matlab引擎实现VC 和Matlab混合编程处理医学数字成像和通信图像。结果:为磁共振图像进行增强局部对比度算法研究提供了一个研发平台,实现了位图图像与医学数字成像和通信图像的数据转换接口功能。结论:处理后的图像具有更好的应用价值,为图像局部对比度增强算法的研究提供一个有效的平台。在算法研制阶段采用VC和Matlab混合编程的方法可以提高算法研究效率。     

基于人工智能算法的步态分析在疾病临床诊疗中的应用进展

<正>在过去几十年中，人工智能（artificial intelligence,AI）已被广泛地应用到搜索引擎、语音识别、人脸识别、自动驾驶等多个领域。近年来，现代医疗领域已成为人工智能领域新的研究热点。基于人工智能算法的步态分析是医疗领域的一个重要研究分支，本文总结了人工智能算法在临床步态分析中的应用进展。     

自动化机器学习在预测重症监护病房感染患者死亡中的应用

目的 探索自动化机器学习(AutoML)在预测重症监护病房(ICU)感染患者死亡中的应用。方法 以2019～2020年四川省自贡市重症监护病房开源数据库中感染患者作为研究对象，基于H₂O平台建立AutoML死亡预测模型。算法包括梯度提升模型(GBM)、极端梯度增强算法(XGBoost)、广义线性模型(GLM)、深度学习(DL)、随机森林(RF)。数据集按照3∶1随机分为训练集和验证集，训练集用于模型的构建，验证集用于评价模型效果。模型表现指标为受试者工作特征(ROC)曲线下面积(AUC),此外通过变量重要性排序、Shapley加法解释图(SHAP)、部分依赖关系和独立模型局部解释(LIME)等方法来解释模型。结果 共计1 151和380例患者分别被纳入训练集和验证集来进行AutoML建模。在验证集中，基于XGBoost算法的AutoML模型表现最优，拥有最高的AUC(0.753)和最高的准确率(0.713),优于第2名GBM模型(AUC 0.748)、第3名GLM模型(AUC 0.745)。在XGBoost模型中，重要的变量包括诊断疾病、活化部分凝血活酶时间(AP...     

基于Visual C＋＋和Matlab的磁共振影像增强后处理系统

目的：随着医学影像学的飞速发展，手术导航技术的应用及脑功能等图像分析研究的不断深入，基于医学数字成像和通信标准的医学影像分析与处理也随之成为医学图像处理领域中的热点.为便于科研人员研究相应的磁共振图像局部增强等后处理算法及进行图像分析，提出一种基于Visual C＋＋和Matlab的磁共振影像增强后处理研发平台. 方法：对北京协和医院放射科2005-01／10获取的部分磁共振医学数字成像和通信图像利用灰度扩展进行全局增强，利用基于数学形态学方法进行局部增强算法的研究：在程序实现上使用Matlab引擎实现VC＋＋和Matlab混合编程处理医学数字成像和通信图像. 结果：为磁共振图像进行增强局部对比度算法研究提供了一个研发平台，实现了位图图像与医学数字成像和通信图像的数据转换接口功能。 结论：处理后的图像具有更好的应用价值，为图像局部对比度增强算法的研究提供一个有效的平台：在算法研制阶段采用VC和Matlab混合编程的方法可以提高算法研究效率.     

监督学习在预测危险因素诱导下急性呼吸窘迫综合征发生风险中的应用

急性呼吸窘迫综合征(acute respiratory distress syndrome, ARDS)缺乏特异性诊断标准，且诱因复杂，在临床实践中往往难以做到早期识别、及时干预，这就需要一种精确、高效的手段辅助识别其发生。基于大数据的机器学习作为一种可以处理海量数据、高效利用有效知识的学习方法，在众多领域发挥了不同作用，在医学领域的重要性日益凸显。截至目前，在医学领域已有大量的机器学习成功应用的案例，其中监督学习算法凭借其可以预测风险的优势，获得众多研究者青睐。本文旨在阐述机器学习算法中监督学习算法在预测危险因素诱导下ARDS发生风险的临床应用。     

腹腔镜疝修补术后患者复发疝气的预测模型构建与验证：一项8年的回顾性研究

目的 建立一个识别腹腔镜疝修补术前、术中及术后高危因素的机器学习模型并预测患者术后疝复发情况。方法 纳入2010—2018年腹腔镜疝修补术后患者,收集患者的29项特征变量,包括患者的人口统计学特征、慢性病史、实验室检查特征、手术信息以及患者的术后情况等。采用梯度提升机(XGBoost)、随机森林(RF)、支持向量机(SVM)以及K最临近(KNN)四种机器学习算法构建预测模型。同时应用SHAP分析对模型进行可视化解释并采用k-折交叉验证法、ROC曲线图、校正曲线、决策曲线分析等指标评价模型性能。结果 共纳入1 178例腹股沟疝患者,其中114例出现术后疝复发。4种预测模型中,XGBoost算法具有最佳效能,其在训练集中的AUC值为0.985,在验证集中的AUC值为0.917,预测准确度很高。k-折交叉验证法、校正曲线、DCA曲线结果均显示XGBoost模型稳定、临床实用性强。此外,独立验证集的AUC值为0.86,说明XGBoost预测模型具有较好的外推性。SHAP分析结果显示,补片尺寸、补片的固定情况、糖尿病史、低蛋白血症、肥胖、吸烟史、术中SpO₂以及术中体温低均...     

基于Visualization Toolkit的脑模型三维重建方法研究

目的利用可视化工具包VisualizationToolkit(VTK)结合VC++实现医学图像三维可视化。方法基于头部CT测量数据,采用MarchingCubes算法和Raycasting算法重建出头模型的表皮和颅骨。结果和结论VTK使用灵活,功能强大,利用它进行图像重建,具有重建步骤简单、效果好、速度快、交互能力强等优点,可以被广泛应用于医学图像的重建中。     

人工智能在病理诊断领域的进展

基于近几年机器视觉的发展,深度学习的人工智能方法应用于组织病理极大程度上促进了病理学家解决临床上的诊断问题,用该种方法解决病理学问题可被称为计算机病理学。人工智能可以做到帮助病理学家初筛大部分良性数据、辅助诊断、疗效预测、识别生物标志物等,甚至可以做到对药效治疗监测以及识别药物发现未知的信号。基于深度学习在病理领域的深入研究,让计算机自动处理病理数据成为可能。人工智能诊断决策建立在大数据之上,很多有可能做到对每个病人的个性化管理,对于大多普遍性的疾病诊断有着更加快速准确的优势。但数字病理学的发展仍受到一些问题的限制,以至于现阶段没有广泛应用于数字病理诊断平台。本文总结了近几年人工智能在病理诊断领域的最新进展,并讨论这种技术的可行性,补充说明在数字病理学中遇到的困难和挑战,并提出在该领域实用性上的展望。      

Supervised machine learning for the early prediction of acute respiratory distress syndrome (ARDS)

PurposeAcute respiratory distress syndrome (ARDS) is a serious respiratory condition with high mortality and associated morbidity. The objective of this study is to develop and evaluate a novel application of gradient boosted tree models trained on patient health record data for the early prediction of ARDS.Materials and methods9919 patient encounters were retrospectively analyzed from the Medical Information Mart for Intensive Care III (MIMIC-III) data base. XGBoost gradient boosted tree models for early ARDS prediction were created using routinely collected clinical variables and numerical representations of radiology reports as inputs. XGBoost models were iteratively trained and validated using 10-fold cross validation.ResultsOn a hold-out test set, algorithm classifiers attained area under the receiver operating characteristic curve (AUROC) values of 0.905 when tested for the detection of ARDS at onset and 0.827, 0.810, and 0.790 for the prediction of ARDS at 12-, 24-, and 48-h windows prior to onset, respectively.ConclusionSupervised machine learning predictions may help predict patients with ARDS up to 48 h prior to onset.     

Improving Target Acquisition for Computer Users With Athetosis

Prior work has highlighted the challenges faced by people with athetosis when trying to acquire on-screen targets using a mouse or trackball. The difficulty of positioning the mouse cursor within a confined area has been identified as a challenging task. We have developed a target acquisition assistance algorithm that features transition assistance via directional gain variation based on target prediction, settling assistance via gain reduction in the vicinity of a predicted target, and expansion of the predicted target as the cursor approaches it. We evaluated the algorithm on improving target acquisition efficiency among seven participants with athetoid cerebral palsy. Our results showed that the algorithm significantly reduced the overall movement time by about 20%. Considering the target acquisition occurs countless times in the course of regular computer use, the accumulative effect of such improvements can be significant for improving the efficiency of computer interaction among people with athetosis.     

基于危险因素和常规实验室指标的子痫前期风险预测模型研究

目的通过对妊娠妇女6～10周的一般资料、危险因素和常规实验室指标水平进行数据分析,构建妊娠早期子痫前期(PE)预测模型,比较Logistic回归模型和极端梯度提升(XGBoost)模型的预测能力。方法回顾性分析2015年1月至2020年8月北京大学第三医院925例PE患者和7 613例正常对照组的一般资料、PE发病危险因素和27项常规实验室指标(妊娠6～10周),包括血脂、肝肾功能、凝血、血细胞计数等指标,采用Mann-Whitney U检验、Logistic回归、XGBoost等统计学方法进行数据分析,分别建立预测模型,绘制ROC曲线抗磷脂综合征,计算曲线下面积(AUC~(ROC))、敏感性、特异性;并用XGBoost绘制特征重要性条形图。结果两组孕妇是否有糖尿病、SLE、抗磷脂综合征、肾病、子痫或PE史以及是否为初产妇的比例差异均有统计学意义(P均0.05)。27个常规实验室指标中,两组除Plt/Lym的水平差异无统计学意义(P均0.05)外,其他所有指标差异均有统计学意义(P均0.05)。仅纳入危险因素(7项)建立Logistic回归模型,AUC~(ROC)为0.621(95%CI:0.601～0.640),敏感性为34.8%,特异性为81.5%;纳入危险因素和实验室指标(6项危险因素+14项实验室指标)建立Logistic模型,AUC~(ROC)为0.752(95%CI:0.735～0.769),敏感性为64.2%,特异性为76.0%;建立XGBoost模型,AUC~(ROC)为0.867(95%CI:0.839～0.896),敏感性为73.0%,特异性为82.3%。采用XGBoost模型进行PE发病早期预测的能力最优。XGBoost筛选出重要性排在前三的指标依次为TG、Lp(a)、C1q。结论单独使用临床危险因素预测PE的效能不高,PE发病危险因素结合常规实验室指标进行妊娠早期预测PE发病风险的效果更优,而XGBoost模型早期预测PE发病的性能优于Logistic回归模型。TG、Lp(a)、C1q是早期预测PE发病的重要变量。     

Forecasting respiratory motion with accurate online support vector regression (SVRpred)

Object

To accurately deliver radiation in image-guided robotic radiosurgery, highly precise prediction algorithms are required. A new prediction method is presented and evaluated.

Materials and methods

SVRpred, a new prediction method based on support vector regression (SVR), has been developed and tested. Computer-generated data mimicking human respiratory motion with a prediction horizon of 150 ms was used for lab tests. The algorithm was subsequently evaluated on a respiratory motion signal recorded during actual radiosurgical treatment using the CyberKnife^®. The algorithm’s performance was compared to the MULIN prediction methods and Wavelet-based multi scale autoregression (wLMS).

Results

The SVRpred algorithm clearly outperformed both the MULIN and the wLMS algorithms on both real (by 15 and 16 percentage points, respectively) and noise-corrupted simulated data (by 13 and 48 percentage points, respectively). Only on noise-free artificial data, the SVRpred algorithm did perform as well as the MULIN algorithms but not as well as the wLMS algorithm.

Conclusion

This new algorithm is a feasible tool for the prediction of human respiratory motion signals significantly outperforming previous algorithms. The only drawback is the high computational complexity and the resulting slow prediction speed. High performance computers will be needed to use the algorithm in live prediction of signals sampled at a high resolution.     

基于截面投影Otsu法的颈动脉内中膜超声图像分割

颈动脉内中膜厚度与多种心脑血管疾病密切相关,是预测该类心脑血管疾病的重要指标。对此,本文将基于截面投影Otsu法引入颈动脉内中膜分割,提出了一种高质量的颈动脉内中膜分割方法,该方法首先将截面投影Otsu法扩展到多阈值情况,以便处理颈动脉图像的多目标性;然后,采用微粒群算法搜索最优阈值,以提高算法效率。临床数据测试结果表明:该方法在抗噪性和时效性方面具有明显优势,可为心血管疾病预防、诊断、治疗及计算机辅助诊断提供重要参考。     

A rate control algorithm for encoding multispectral images from remote sensing satellites

This letter presents a rate allocation algorithm that can be used when the Consultative Committee for Space Data Systems (CCSDS) lossy image compressor is applied to encode multispectral images. This algorithm is a numerical solution to the rate control problem when multiple information sources are compressed independently. Given the rate distortion curves and a total bit rate, the algorithm performs an efficient search to find the best rate distribution among the components that produces the lowest distortion. It is shown that the algorithm introduced here is asymptotically optimal. Furthermore, the relationship between performance and complexity can be easily controlled. The efficiency of the algorithm is tested when CCSDS compressor is used to encode the Karhunen and Loève Transform components of multispectral images. The test is performed for a group of images from different multispectral sensors. The convergence speed and the rate distortion curve are evaluated. Numerical results are compared to the performance of the same compressor when it uses the rate control algorithm based on the Gaussian rate distortion function. It is shown that the proposed algorithm can produce improvements over 2 dB in the Peak Signal-to-Noise Ratio even with a reduced computational complexity.     

BP神经网络在疾病分析影响因素中的作用

背景:疾病产生原因复杂多样,临床医生对大量样本病历数据挖掘的探讨往往缺乏有效的手段,信息技术的应用能力有待提高.目的:利用人工神经网络的BP算法,对临床大样本量的病历进行分析,以找出某种疾病的致病因素与疾病本身之间的内在关系.方法:以高血压病为例,以某中医院2010-07的高血压患者病历数据为实验数据,对疾病的影响因素进行建模,优选 Microsoft SQL Server 2005 Analysis Services智能工具,分析其挖掘结果,并利用单独查询进行预测与决策支持.结果与结论:应用基于BP算法的人工神经网络分析疾病的致病因素对疾病本身的影响有较好的预测效果,有利于提升医务人员借助信息技术方法在临床诊断的水平,提高疾病诊断效率.

Abstract：

BACKGROUND: Disease pathogenic factors are complicated. There is not an effective method to analyze large sample data mining, and application ability of information technology of clinical doctors needs to be improved. OBJECTIVE: Using BP algorithm of artificial neural network to analyze large sample clinical cases, in order to explore inner relations between disease pathogenic factors and diseases.METHODS: Take hypertension for example, medical data of patients with hypertension in a traditional Chinese medical hospital served as experimental data, and the influence factors of the disease were simulated with Microsoft SQL Server 2005 Analysis Services, the mining data was analyzed, and a single query was used as prediction and decision support.RESULTS AND CONCLUSION: Analysis of effect of disease pathogenic factors on disease itself based on artificial neural network with BP algorithm has good predictive effect in clinical diagnosis, which is of benefit to enhance the diagnostic efficiency of medical personnel using information technology.     

BP神经网络在疾病分析影响因素中的作用

背景：疾病产生原因复杂多样,临床医生对大量样本病历数据挖掘的探讨往往缺乏有效的手段,信息技术的应用能力有待提高。目的：利用人工神经网络的BP算法,对临床大样本量的病历进行分析,以找出某种疾病的致病因素与疾病本身之间的内在关系。方法：以高血压病为例,以某中医院2010-07的高血压患者病历数据为实验数据,对疾病的影响因素进行建模,优选Microsoft SQL Server 2005 Analysis Services智能工具,分析其挖掘结果,并利用单独查询进行预测与决策支持。结果与结论：应用基于BP算法的人工神经网络分析疾病的致病因素对疾病本身的影响有较好的预测效果,有利于提升医务人员借助信息技术方法在临床诊断的水平,提高疾病诊断效率。     

Application of machine learning algorithm for predicting gestational diabetes mellitus in early pregnancy

Objective: To study the application of a machine learning algorithm for predicting gestational diabetes mellitus (GDM) in early pregnancy.Methods: This study identified indicators related to GDM through a literature review and exper t discussion. Pregnant women who had attended medical institutions for an antenatal examination from November 2017 to August 2018 were selected for analysis, and the collected indicators were retrospectively analyzed. Based on Python, the indicators were classified and modeled using a random forest regression algorithm, and the performance of the prediction model was analyzed. Results: We obtained 4806 analyzable data from 1625 pregnant women. Among these, 3265 samples with all 67 indicators were used to establish data set F1; 4806 samples with 38 identical indicators were used to establish data set F2. Each of F1 and F2 was used for training the random forest algorithm. The overall predictive accuracy of the F1 model was 93.10％, area under the receiver operating characteristic curve (AUC) was 0.66, and the predictive accuracy of GDM-positive cases was 37.10％. The corresponding values for the F2 model were 88.70％, 0.87, and 79.44％. The results thus showed that the F2 prediction model performed better than the F1 model. To explore the impact of sacrificial indicators on GDM prediction, the F3 data set was established using 3265 samples (F1) with 38 indicators (F2). After training, the overall predictive accuracy of the F3 model was 91.60％, AUC was 0.58, and the predictive accuracy of positive cases was 15.85％. Conclusions: In this study, a model for predicting GDM with several input variables (e.g., physical examination, past history, personal history, family history, and laboratory indicators) was established using a random forest regression algorithm. The trained prediction model exhibited a good performance and is valuable as a reference for predicting GDM in women at an early stage of pregnancy. In addition, there are cer tain requirements for the propor tions of negative and positive cases in sample data sets when the random forest algorithm is applied to the early prediction of GDM.     

影像组学在卵巢癌中的研究进展

作为一种能将医学图像转换为高维、可挖掘数据的新技术,影像组学已广泛用于研究多种肿瘤,并建立了系统的工作流程。本研究对影像组学在卵巢癌诊断及分型、预后预测和分子生物学解释等方面的研究进展及其发展方向进行综述。