Comparing Radiomics features of tumour and healthy liver tissue in a limited CT dataset: A machine learning study

IntroductionLiver cancer lesions on Computed Tomography (CT) withholds a great amount of data, which is not visible to the radiologists and radiographer. Radiomics features can be extracted from the lesions and used to train Machine Learning (ML) algorithms to predict between tumour and liver tissue. The purpose of this study was to investigate and classify Radiomics features extracted from liver tumours and normal liver tissue in a limited CT dataset.MethodsThe Liver Tumour Segmentation Benchmark (LiTS) dataset consisting of 131 CT scans of the liver with segmentations of tumour tissue and healthy liver was used to extract Radiomic features. Extracted Radiomic features included size, shape, and location extracted with morphological and statistical techniques according to the International Symposium on Biomedical Imaging manual. Relevant features was selected with chi2 correlation and principal component analysis (PCA) with tumour and healthy liver tissue as outcome according to a consensus between three experienced radiologists. Logistic regression, random forest and support vector machine was used to train and validate the dataset with a 10-fold cross-validation method and the Grid Search as hyper-parameter tuning. Performance was evaluated with sensitivity, specificity and accuracy.ResultsThe performance of the ML algorithms achieved sensitivities, specificities and accuracy ranging from 96.30% (95% CI: 81.03%–99.91%) to 100.00% (95% CI: 86.77%–100.00%), 91.30% (95% CI: 71.96%–98.93%) to 100.00% (95% CI: 83.89%–100.00%)and 94.00% (95% CI: 83.45%–98.75%) to 100.00% (95% CI: 92.45%–100.00%), respectively.ConclusionML algorithms classifies Radiomics features extracted from healthy liver and tumour tissue with perfect accuracy. The Radiomics signature allows for a prognostic biomarker for hepatic tumour screening on liver CT.Implications for practiceDifferentiation between tumour and liver tissue with Radiomics ML algorithms have the potential to increase the diagnostic accuracy, assist in the decision-making of supplementary multiphasic enhanced medical imaging, as well as for developing novel prognostic biomarkers for liver cancer patients.     

UK reporting radiographers’ perceptions of AI in radiographic image interpretation – Current perspectives and future developments

IntroductionRadiographer reporting is accepted practice in the UK. With a national shortage of radiographers and radiologists, artificial intelligence (AI) support in reporting may help minimise the backlog of unreported images. Modern AI is not well understood by human end-users. This may have ethical implications and impact human trust in these systems, due to over- and under-reliance. This study investigates the perceptions of reporting radiographers about AI, gathers information to explain how they may interact with AI in future and identifies features perceived as necessary for appropriate trust in these systems.MethodsA Qualtrics® survey was designed and piloted by a team of UK AI expert radiographers. This paper reports the third part of the survey, open to reporting radiographers only.Results86 responses were received. Respondents were confident in how an AI reached its decision (n = 53, 62%). Less than a third of respondents would be confident communicating the AI decision to stakeholders. Affirmation from AI would improve confidence (n = 49, 57%) and disagreement would make respondents seek a second opinion (n = 60, 70%). There is a moderate trust level in AI for image interpretation. System performance data and AI visual explanations would increase trust.ConclusionsResponses indicate that AI will have a strong impact on reporting radiographers’ decision making in the future. Respondents are confident in how an AI makes decisions but less confident explaining this to others. Trust levels could be improved with explainable AI solutions.Implications for practiceThis survey clarifies UK reporting radiographers’ perceptions of AI, used for image interpretation, highlighting key issues with AI integration.     

Elaboración de modelos predictivos de la gravedad y la mortalidad en pacientes con COVID-19 que acuden al servicio de urgencias,incluida la radiografía torácica

ObjectivesTo develop prognosis prediction models for COVID-19 patients attending an emergency department (ED) based on initial chest X-ray (CXR), demographics, clinical and laboratory parameters.MethodsAll symptomatic confirmed COVID-19 patients admitted to our hospital ED between February 24th and April 24th 2020 were recruited. CXR features, clinical and laboratory variables and CXR abnormality indices extracted by a convolutional neural network (CNN) diagnostic tool were considered potential predictors on this first visit. The most serious individual outcome defined the three severity level: 0) home discharge or hospitalization ≤ 3 days, 1) hospital stay >3 days and 2) intensive care requirement or death. Severity and in-hospital mortality multivariable prediction models were developed and internally validated. The Youden index was used for the optimal threshold selection of the classification model.ResultsA total of 440 patients were enrolled (median 64 years; 55.9% male); 13.6% patients were discharged, 64% hospitalized, 6.6% required intensive care and 15.7% died. The severity prediction model included oxygen saturation/inspired oxygen fraction (SatO2/FiO2), age, C-reactive protein (CRP), lymphocyte count, extent score of lung involvement on CXR (ExtScoreCXR), lactate dehydrogenase (LDH), D-dimer level and platelets count, with AUC-ROC = 0.94 and AUC-PRC = 0.88. The mortality prediction model included age, SatO2/FiO2, CRP, LDH, CXR extent score, lymphocyte count and D-dimer level, with AUC-ROC = 0.97 and AUC-PRC = 0.78. The addition of CXR CNN-based indices did not improve significantly the predictive metrics.ConclusionThe developed and internally validated severity and mortality prediction models could be useful as triage tools in ED for patients with COVID-19 or other virus infections with similar behaviour.     

Using big data from probabilistic genotyping to solve crime

Forensic Science South Australia (FSSA) has been using STRmix™ software to deconvolute all reported DNA mixtures since 2012. Almost a decade of deconvolutions had led to a substantial repository of analysed profile data that can be interrogated to observe trends in case type, location or occurrence. In addition, deconvolutions can be compared in order to identify common DNA donors and reveal new intelligence information in cases where DNA profiling has previously provided no investigative information. As a proof of concept all samples deconvoluted as part of criminal casework (suspect or no-suspect) were interrogated and compared to each other using the mixture-to-mixture comparison feature in STRmix™. Within the Adelaide region there were 32 groups of cases that had evidence samples linked by a common DNA donor with LR > 1 million which was in addition to direct links and mixture searching links identified previously. These groups of cases can then be interrogated to reveal additional information to inform Police intelligence gathering. Our paper reports on the findings of this proof-of-concept study.     

Machine learning models for Parkinson’s disease detection and stage classification based on spatial-temporal gait parameters

BackgroundParkinson’s disease (PD) is a chronic and progressive neurodegenerative disease with no cure, presenting a challenging diagnosis and management. However, despite a significant number of criteria and guidelines have been proposed to improve the diagnosis of PD and to determine the PD stage, the gold standard for diagnosis and symptoms monitoring of PD is still mainly based on clinical evaluation, which includes several subjective factors. The use of machine learning (ML) algorithms in spatial-temporal gait parameters is an interesting advance with easy interpretation and objective factors that may assist in PD diagnostic and follow up.Research questionThis article studies ML algorithms for: i) distinguish people with PD vs. matched-healthy individuals; and ii) to discriminate PD stages, based on selected spatial-temporal parameters, including variability and asymmetry.MethodsGait data acquired from 63 people with PD with different levels of PD motor symptoms severity, and 63 matched-control group individuals, during self-selected walking speed, was study in the experiments.ResultsIn the PD diagnosis, a classification accuracy of 84.6 %, with a precision of 0.923 and a recall of 0.800, was achieved by the Naïve Bayes algorithm. We found four significant gait features in PD diagnosis: step length, velocity and width, and step width variability. As to the PD stage identification, the Random Forest outperformed the other studied ML algorithms, by reaching an Area Under the ROC curve of 0.786. We found two relevant gait features in identifying the PD stage: stride width variability and step double support time variability.SignificanceThe results showed that the studied ML algorithms have potential both to PD diagnosis and stage identification by analysing gait parameters.     

Cloud-based fusion imaging improves operative metrics during fenestrated endovascular aneurysm repair

ObjectiveEndovascular treatment of complex aortic pathology has been associated with increases in procedural-related metrics, including the operative time and radiation exposure. Three-dimensional fusion imaging technology has decreased the radiation dose and iodinated contrast use during endovascular aneurysm repair. The aim of the present study was to report our institutional experience with the use of a cloud-based fusion imaging platform during fenestrated endovascular aneurysm repair (FEVAR).MethodsA retrospective review of a prospectively maintained aortic database was performed to identify all patients who had undergone FEVAR with commercially available devices (Zenith Fenestrated; Cook Medical Inc, Bloomington, IN) between 2013 and 2020 and all endovascular aneurysm repairs performed using Cydar EV Intelligent Maps (Cydar Medical, Cambridge, UK). The Cydar EV cohort was reviewed further to select all FEVARs performed with overlay map guidance. The patient demographic, clinical, and procedure metrics were analyzed, with a comparative analysis of FEVAR performed without and with the Cydar EV imaging platform. Patients were excluded from comparative analysis if the data were incomplete in the dataset or they had a documented history of prior open or endovascular abdominal aortic aneurysm repair.ResultsDuring the 7-year study period, 191 FEVARs had been performed. The Cydar EV imaging platform was implemented in 2018 and used in 124 complex endovascular aneurysm repairs, including 69 consecutive FEVARs. A complete dataset was available for 137 FEVARs. With exclusion to select for de novo FEVAR, a comparative analysis was performed of 53 FEVAR without and 63 with Cydar EV imaging guidance. The cohorts were similar in patient demographics, medical comorbidities, and aortic aneurysm characteristics. No significant difference was noted between the two groups for major adverse postoperative events, length of stay, or length of intensive care unit stay. The use of Cydar EV resulted in nonsignificant decreases in the mean fluoroscopy time (69.3 ± 28 minutes vs 66.2 ± 33 minutes; P = .598) and operative time (204.4 ± 64 minutes vs 186 ± 105 minutes; P = .278). A statistically significant decrease was found in the iodinated contrast volume (105 ± 44 mL vs 83 ± 32 mL; P = .005), patient radiation exposure using the dose area product (1,049,841 mGy/cm² vs 630,990 mGy/cm²; P < .001) and cumulative air kerma levels (4518 mGy vs 3084 mGy; P = .02) for patients undergoing FEVAR with Cydar EV guidance.ConclusionsAt our aortic center, we have observed a trend toward shorter operative times and significant reductions in both iodinated contrast use and radiation exposure during FEVAR using the Cydar EV intelligent maps. Intelligent map guidance improved the efficiency of complex endovascular aneurysm repair, providing a safer intervention for both patient and practitioner.     

人工智能在骨质疏松症中应用的研究进展

骨质疏松症(osteoporosis, OP)是一种与增龄相关的骨骼疾患,其起病隐匿,呈渐进性发展,患者初期无明显的临床表现,但随着病情的进展,骨量不断流失及骨组织微结构破坏,进而出现骨痛、脊柱变形,甚至出现骨质疏松性骨折(osteoporotic fracture, OPF)等严重并发症。双能X线吸收法是目前临床诊断OP的金标准,但由于其诊断的准确度受到体重、腰椎退行性改变及主动脉壁钙化等因素的影响,存在假阴性诊断的可能。近年来,人工智能(artificial intelligence, AI)在医学领域快速发展,目前AI已广泛应用于OP的研究中,其在OP筛查、诊断及预测领域的研究已成为一个新的热点。该文从AI应用于OP的早期筛查、医学影像学表现、临床诊疗资料及OPF风险预测等4个方面,阐述AI在OP诊疗过程中的应用现状及优势,为OP的精准诊疗提供新方向。     

青光眼影像人工智能深度学习研究现状与展望

青光眼是一组异质性神经退行性疾病,其特征是视网膜神经节细胞及其轴突逐渐消失,现已成为全球不可逆性失明的主要原因。人工智能(AI)是由机器展示的智能,而深度学习(DL)是其中一个基于深度神经网络的分支,在医学成像领域取得了重大突破。在青光眼影像方面,已有越来越多的研究将DL应用于眼底图像以及光学相干断层扫描(OCT),以检测青光眼性视神经病变。有很好的结果显示,将DL技术整合到影像中进行青光眼评估是高效、准确的,这可能会解决当前实践和临床工作流程中的一些难题。但是,未来进一步的研究对于解决现存挑战至关重要,例如为不同研究之间的图像标记建立标准,将“黑匣子”的学习过程进行可视化,提高模型在未知数据集上的泛化能力,开发基于DL的实际应用程序,以及建立合理的临床工作流程,进行前瞻性验证和成本效益分析等。本文总结了AI应用于青光眼影像的最新研究现状,并讨论了对临床的潜在影响和未来的研究方向。     

高乳酸血症-卒中样发作综合征型线粒体脑肌病一家系两代四例报告

线粒体脑肌病属于罕见性母系遗传病,本文回顾性分析了1家4例高乳酸血症-卒中样发作综合征（MELAS）型线粒体脑肌病患者,其主要表现为卒中样发作、头痛、癫痫、高乳酸血症、肌肉不耐受疲劳、高级智能下降、听力下降和身材矮小等,结合特征性影像学变化、基因检测及肌肉活检明确诊断,并结合文献对只有女儿能将其线粒体DNA（mt-DNA）传递给下一代的母系遗传MELAS型线粒体脑肌病临床特点进行了总结分析,旨在帮助临床认识此病,进一步提高MELAS型线粒体脑肌病的临床诊断率。     

超声在新型冠状病毒肺炎诊疗中的应用

新型冠状病毒肺炎（COVID-19）传播速度快,感染率高,疫情已构成全球大流行。超声设备具有便携、方便消毒、检查模式多样等优点,可应用于COVID-19的辅助诊断、病情监测及治疗评估,为诊治决策的调整提供指导。超声专家还可通过智能化超声装备在4G/5G网络支持下进行超声远程会诊,通过异地操作超声机器人机械臂完成实时扫查,有效缓解隔离区超声医师不足的问题,降低感染风险。目前,超声是唯一能进入隔离区对COVID-19患者进行床旁检查的可视化影像设备。本文分析了超声技术在COVID-19所致肺部损伤及全身多脏器病变检查诊断、病情监测、治疗评估中的应用价值,并就远程超声、超声人工智能等领域的应用前景进行探讨。