Towards gender equity in artificial intelligence and machine learning applications in dermatology

There has been increased excitement around the use of machine learning (ML) and artificial intelligence (AI) in dermatology for the diagnosis of skin cancers and assessment of other dermatologic conditions. As these technologies continue to expand, it is essential to ensure they do not create or widen sex- and gender-based disparities in care. While desirable bias may result from the explicit inclusion of sex or gender in diagnostic criteria of diseases with gender-based differences, undesirable biases can result from usage of datasets with an underrepresentation of certain groups. We believe that sex and gender differences should be taken into consideration in ML/AI algorithms in dermatology because there are important differences in the epidemiology and clinical presentation of dermatologic conditions including skin cancers, sex-specific cancers, and autoimmune conditions. We present recommendations for ensuring sex and gender equity in the development of ML/AI tools in dermatology to increase desirable bias and avoid undesirable bias.     

CBL教学法在医学影像科实习中的应用

在医学影像科实习中,研究组采用CBL教学法完成医学影像学教学任务,对照组采用传统小讲课方式。由带教教师对其影像学病例分析进行评价、综合评分,探讨CBL在医学影像科实习中的实施及教学效果评价。     

基于深度学习和多种机器学习算法预测人体细胞色素P450抑制剂活性

人体细胞色素P450（CYP）受到抑制会导致药物-药物相互作用,从而产生严重的不良反应。因此,准确预测给定化合物对特定CYP亚型的抑制能力至关重要。本研究基于不同的分子表征,比较了11种机器学习方法和2种深度学习模型,实验结果表明,基于RDKit_2d + Morgan的CatBoost机器学习模型在准确率和马修斯系数方面优于其他模型,甚至优于先前发表的模型。此外,实验结果还显示,CatBoost模型不仅性能佳,而且计算资源消耗较低。最后,本文将表现较好的前3名模型结合为co_model,其在性能方面稍微优于单独使用CatBoost模型。     

Research and applications: Machine learning for predicting the response of breast cancer to neoadjuvant chemotherapy

Objective

To employ machine learning methods to predict the eventual therapeutic response of breast cancer patients after a single cycle of neoadjuvant chemotherapy (NAC).

Materials and methods

Quantitative dynamic contrast-enhanced MRI and diffusion-weighted MRI data were acquired on 28 patients before and after one cycle of NAC. A total of 118 semiquantitative and quantitative parameters were derived from these data and combined with 11 clinical variables. We used Bayesian logistic regression in combination with feature selection using a machine learning framework for predictive model building.

Results

The best predictive models using feature selection obtained an area under the curve of 0.86 and an accuracy of 0.86, with a sensitivity of 0.88 and a specificity of 0.82.

Discussion

With the numerous options for NAC available, development of a method to predict response early in the course of therapy is needed. Unfortunately, by the time most patients are found not to be responding, their disease may no longer be surgically resectable, and this situation could be avoided by the development of techniques to assess response earlier in the treatment regimen. The method outlined here is one possible solution to this important clinical problem.

Conclusions

Predictive modeling approaches based on machine learning using readily available clinical and quantitative MRI data show promise in distinguishing breast cancer responders from non-responders after the first cycle of NAC.     

Research and applications: Learning regular expressions for clinical text classification

Objectives

Natural language processing (NLP) applications typically use regular expressions that have been developed manually by human experts. Our goal is to automate both the creation and utilization of regular expressions in text classification.

Methods

We designed a novel regular expression discovery (RED) algorithm and implemented two text classifiers based on RED. The RED+ALIGN classifier combines RED with an alignment algorithm, and RED+SVM combines RED with a support vector machine (SVM) classifier. Two clinical datasets were used for testing and evaluation: the SMOKE dataset, containing 1091 text snippets describing smoking status; and the PAIN dataset, containing 702 snippets describing pain status. We performed 10-fold cross-validation to calculate accuracy, precision, recall, and F-measure metrics. In the evaluation, an SVM classifier was trained as the control.

Results

The two RED classifiers achieved 80.9–83.0% in overall accuracy on the two datasets, which is 1.3–3% higher than SVM''s accuracy (p<0.001). Similarly, small but consistent improvements have been observed in precision, recall, and F-measure when RED classifiers are compared with SVM alone. More significantly, RED+ALIGN correctly classified many instances that were misclassified by the SVM classifier (8.1–10.3% of the total instances and 43.8–53.0% of SVM''s misclassifications).

Conclusions

Machine-generated regular expressions can be effectively used in clinical text classification. The regular expression-based classifier can be combined with other classifiers, like SVM, to improve classification performance.     

可解释机器学习方法在疾病预测中的应用:脓毒血症患者死亡风险研究

目的 探索可解释机器学习方法在疾病预测中的应用。方法 本研究以脓毒血症死亡风险预测为例,从重症监护医学数据库(Medical Information Mart for Intensive Care, MIMIC)-Ⅳ中采集符合纳排标准的19 903例脓毒血症(sepsis-3)患者的临床数据,利用决策树、逻辑回归、随机森林、XGBoost、轻量梯度提升机(light gradient boosting machine,LightGBM)模型分别构建脓毒血症死亡预测模型。在此基础上,利用全局可解释方法(特征重要性、部分依赖图、个体条件期望、全局代理模型)和局部可解释方法(局部代理模型和Shapely值)对复杂机器学习模型进行解释,探索影响脓毒血症患者预后的危险因素。结果 解释性差的机器学习模型的预测性能[模型LightGBM、随机森林、XGBoost的曲线下面积(area under curve,AUC)值分别为0.913、0.892、0.872]高于具有内在解释性的模型(逻辑回归模型AUC=0.779,决策树模型AUC=0.791),并利用全局解释性方法、局部可解释性方法两种类型的解释方法对机器学习模型决策过程进行解释。结论 利用全局解释性方法可以解释在整个特征空间内机器学习模型的响应趋势,利用局部可解释性方法可以解释机器学习模型对特定病例的决策过程。     

Research and applications: Automatic lymphoma classification with sentence subgraph mining from pathology reports

Objective

Pathology reports are rich in narrative statements that encode a complex web of relations among medical concepts. These relations are routinely used by doctors to reason on diagnoses, but often require hand-crafted rules or supervised learning to extract into prespecified forms for computational disease modeling. We aim to automatically capture relations from narrative text without supervision.

Methods

We design a novel framework that translates sentences into graph representations, automatically mines sentence subgraphs, reduces redundancy in mined subgraphs, and automatically generates subgraph features for subsequent classification tasks. To ensure meaningful interpretations over the sentence graphs, we use the Unified Medical Language System Metathesaurus to map token subsequences to concepts, and in turn sentence graph nodes. We test our system with multiple lymphoma classification tasks that together mimic the differential diagnosis by a pathologist. To this end, we prevent our classifiers from looking at explicit mentions or synonyms of lymphomas in the text.

Results and Conclusions

We compare our system with three baseline classifiers using standard n-grams, full MetaMap concepts, and filtered MetaMap concepts. Our system achieves high F-measures on multiple binary classifications of lymphoma (Burkitt lymphoma, 0.8; diffuse large B-cell lymphoma, 0.909; follicular lymphoma, 0.84; Hodgkin lymphoma, 0.912). Significance tests show that our system outperforms all three baselines. Moreover, feature analysis identifies subgraph features that contribute to improved performance; these features agree with the state-of-the-art knowledge about lymphoma classification. We also highlight how these unsupervised relation features may provide meaningful insights into lymphoma classification.     

New onset delirium prediction using machine learning and long short-term memory (LSTM) in electronic health record

ObjectiveTo develop and test an accurate deep learning model for predicting new onset delirium in hospitalized adult patients.MethodsUsing electronic health record (EHR) data extracted from a large academic medical center, we developed a model combining long short-term memory (LSTM) and machine learning to predict new onset delirium and compared its performance with machine-learning-only models (logistic regression, random forest, support vector machine, neural network, and LightGBM). The labels of models were confusion assessment method (CAM) assessments. We evaluated models on a hold-out dataset. We calculated Shapley additive explanations (SHAP) measures to gauge the feature impact on the model.ResultsA total of 331 489 CAM assessments with 896 features from 34 035 patients were included. The LightGBM model achieved the best performance (AUC 0.927 [0.924, 0.929] and F1 0.626 [0.618, 0.634]) among the machine learning models. When combined with the LSTM model, the final model’s performance improved significantly (P = .001) with AUC 0.952 [0.950, 0.955] and F1 0.759 [0.755, 0.765]. The precision value of the combined model improved from 0.497 to 0.751 with a fixed recall of 0.8. Using the mean absolute SHAP values, we identified the top 20 features, including age, heart rate, Richmond Agitation-Sedation Scale score, Morse fall risk score, pulse, respiratory rate, and level of care.ConclusionLeveraging LSTM to capture temporal trends and combining it with the LightGBM model can significantly improve the prediction of new onset delirium, providing an algorithmic basis for the subsequent development of clinical decision support tools for proactive delirium interventions.     

基于机器学习的神经电生理在慢性疾病诊断中的应用

慢性疾病正逐步取代传染性疾病,成为威胁人类身体健康最为主要的因素,给家庭和社会带来沉重负担.因此,实现慢性疾病早期诊断意义重大.目前的临床诊断方法难以实现这一目标,因此需要一种新的慢性疾病诊断方法.心脑血管疾病、恶性肿瘤、糖尿病等常见的慢性疾病都会影响神经系统,造成神经电信号变化,因此神经电生理将会成为早期诊断慢性疾病...     

An explainable machine learning platform for pyrazinamide resistance prediction and genetic feature identification of Mycobacterium tuberculosis

ObjectiveTuberculosis is the leading cause of death from a single infectious agent. The emergence of antimicrobial resistant Mycobacterium tuberculosis strains makes the problem more severe. Pyrazinamide (PZA) is an important component for short-course treatment regimens and first- and second-line treatment regimens. This research aims for fast diagnosis of M. tuberculosis resistance to PZA and identification of genetic features causing resistance.Materials and MethodsWe use clinically collected genomic data of M. tuberculosis that are resistant or susceptible to PZA. A machine learning platform is built to diagnose PZA resistance using the whole genome sequence data, and to identify resistance genes and mutations. The platform consists of a deep convolutional neural network (DCNN) model for resistance diagnosis and a support vector machine (SVM) model as a surrogate to identify resistance genes and mutations.ResultsThe DCNN model achieves a PZA resistance diagnosis accuracy of 93%. Each prediction takes less than a second. The SVM has revealed 2 novel genes, embB and gyrA, besides the well-known pncA gene, and 9 mutations that harbor PZA resistance.DiscussionThe DCNN and SVM machine learning platform, if used together with the real-time genome sequencing machines, could allow for rapid PZA diagnosis, allowing for critical time to ensure good patient outcomes, and preventing outbreaks of deadly infections. Furthermore, identifying pertinent resistance genes and mutations will help researchers better understand the biological mechanisms behind resistance.ConclusionsMachine learning can be used to achieve high-accuracy resistance prediction, and identify genes and mutations causing the resistance.     

基于机器学习的老年创伤性颅脑损伤预后研究

目的建立老年创伤性颅脑损伤预后模型,并分析预后的影响因素。方法收集2009年1月—2019年1月颅脑外伤患者2272例资料,其中老年组患者680例（年龄≥65岁）,非老年组1592例（年龄＜65岁）。将伤后第3个月格拉斯哥结局评分、住院天数、并发症次数作为终点指标,利用多种机器学习的算法进行两组间预后因素差异的分析。结果老年组患者与非老年组患者相比,预后更差,住院时间更长,两组间差异有统计学意义（P=0.024,P＜0.001）。老年组经过筛选,老年组患者的3个终点指标使用多层感知器模型,非老年组中格拉斯哥结局评分使用多层感知器模型,住院天数、并发症次数的预测采用C5.0决策树模型。急诊GCS、具体年龄对老年组患者的预后有更大的影响。结论老年颅脑外伤患者与年轻人所适用的机器学习模型不尽相同,老年人预后更差,年龄和急诊GCS对预后的影响可能更大。     

基于机器学习算法的NT-proBNP灰值患者心力衰竭判别模型评价

目的 应用机器学习算法构建氨基末端脑钠尿肽(N-terminal pro-brain natriuretic peptide,NT-proBNP)灰值患者心力衰竭判别模型并评价.方法 收集2013年1月至2018年12月在上海市浦东新区公利医院进行NT-proBNP 实验室检测的患者临床资料和实验室检测信息,数据清洗后...     

Stroke risk prediction using machine learning: a prospective cohort study of 0.5 million Chinese adults

ObjectiveTo compare Cox models, machine learning (ML), and ensemble models combining both approaches, for prediction of stroke risk in a prospective study of Chinese adults.Materials and MethodsWe evaluated models for stroke risk at varying intervals of follow-up (<9 years, 0–3 years, 3–6 years, 6–9 years) in 503 842 adults without prior history of stroke recruited from 10 areas in China in 2004–2008. Inputs included sociodemographic factors, diet, medical history, physical activity, and physical measurements. We compared discrimination and calibration of Cox regression, logistic regression, support vector machines, random survival forests, gradient boosted trees (GBT), and multilayer perceptrons, benchmarking performance against the 2017 Framingham Stroke Risk Profile. We then developed an ensemble approach to identify individuals at high risk of stroke (>10% predicted 9-yr stroke risk) by selectively applying either a GBT or Cox model based on individual-level characteristics.ResultsFor 9-yr stroke risk prediction, GBT provided the best discrimination (AUROC: 0.833 in men, 0.836 in women) and calibration, with consistent results in each interval of follow-up. The ensemble approach yielded incrementally higher accuracy (men: 76%, women: 80%), specificity (men: 76%, women: 81%), and positive predictive value (men: 26%, women: 24%) compared to any of the single-model approaches.Discussion and ConclusionAmong several approaches, an ensemble model combining both GBT and Cox models achieved the best performance for identifying individuals at high risk of stroke in a contemporary study of Chinese adults. The results highlight the potential value of expanding the use of ML in clinical practice.     

Patient safety and quality improvement: Ethical principles for a regulatory approach to bias in healthcare machine learning

Accumulating evidence demonstrates the impact of bias that reflects social inequality on the performance of machine learning (ML) models in health care. Given their intended placement within healthcare decision making more broadly, ML tools require attention to adequately quantify the impact of bias and reduce its potential to exacerbate inequalities. We suggest that taking a patient safety and quality improvement approach to bias can support the quantification of bias-related effects on ML. Drawing from the ethical principles underpinning these approaches, we argue that patient safety and quality improvement lenses support the quantification of relevant performance metrics, in order to minimize harm while promoting accountability, justice, and transparency. We identify specific methods for operationalizing these principles with the goal of attending to bias to support better decision making in light of controllable and uncontrollable factors.     

机器学习在调强放疗计划剂量验证中的作用

目的：探讨机器学习对调强放疗计划剂量验证中的作用。方法：选取2019年3月至2020年8月在温州医科大学附属第一医院接受双弧容积调强放射治疗（VMAT）的141例患者,提取调强计划的13个复杂度参数并收集不同条件下的伽玛通过率（GPR）,将数据按照7:3随机划分为训练集与测试集。通过Pearson相关性分析和套索回归（LASSO）筛选参数,利用支持向量机的机器学习方法进行建模,对GPR分别进行数值和分类预测。均方根误差（RMSE）和平均绝对误差（MAE）用来评估模型数值预测的准确性,曲线下面积（AUC）用来评估模型分类的准确性。结果：在GPR数值预测中,在3%/3 mm、3%/2 mm、2%/2 mm条件下,测试集中RMSE分别为2.22、3.51、4.59;MAE分别为1.56、2.68、3.67。在GPR分类预测中,在3%/3 mm、3%/2 mm、2%/2 mm条件下测试集的AUC结果分别0.79、0.78、0.77。结论：基于机器学习对调强放疗计划进行剂量验证具有一定的临床应用价值,为质量保证提供了一种新思路。     

以问题为基础的学习方法在放射诊断学教学中的应用

在医学影像学专业的放射诊断学课程中引入以问题为基础的教学法（problem-basedlearning,PBL）,以问卷调查结果作为评估依据,与传统的以授课为基础的教学法进行对比分析。结果显示,PBL教学法对学生学科知识的掌握、扩展知识的深度和广度、学习积极性的提高有很大帮助,尤其是对培养自学能力和综合分析能力效果显著,教学质量明显提高。     

Using machine learning to improve the accuracy of patient deterioration predictions: Mayo Clinic Early Warning Score (MC-EWS)

ObjectiveWe aimed to develop a model for accurate prediction of general care inpatient deterioration.Materials and MethodsTraining and internal validation datasets were built using 2-year data from a quaternary hospital in the Midwest. Model training used gradient boosting and feature engineering (clinically relevant interactions, time-series information) to predict general care inpatient deterioration (resuscitation call, intensive care unit transfer, or rapid response team call) in 24 hours. Data from a tertiary care hospital in the Southwest were used for external validation. C-statistic, sensitivity, positive predictive value, and alert rate were calculated for different cutoffs and compared with the National Early Warning Score. Sensitivity analysis evaluated prediction of intensive care unit transfer or resuscitation call.ResultsTraining, internal validation, and external validation datasets included 24 500, 25 784 and 53 956 hospitalizations, respectively. The Mayo Clinic Early Warning Score (MC-EWS) demonstrated excellent discrimination in both the internal and external validation datasets (C-statistic = 0.913, 0.937, respectively), and results were consistent in the sensitivity analysis (C-statistic = 0.932 in external validation). At a sensitivity of 73%, MC-EWS would generate 0.7 alerts per day per 10 patients, 45% less than the National Early Warning Score.DiscussionLow alert rates are important for implementation of an alert system. Other early warning scores developed for the general care ward have achieved lower discrimination overall compared with MC-EWS, likely because MC-EWS includes both nursing assessments and extensive feature engineering.ConclusionsMC-EWS achieved superior prediction of general care inpatient deterioration using sophisticated feature engineering and a machine learning approach, reducing alert rate.     

基于机器学习的药物血浆蛋白结合率的预测

预测药物在血浆中的蛋白结合率,有助于了解药物的药代动力学特征,对药物发现的早期研究有重要的参考价值.本研究收集了 2452个临床药物的血浆蛋白结合率信息,用Molecular Operating Environment(MOE)和Mordred两种软件计算分子描述符,将算得的分子描述符作为模型的输入特征.使用极端梯度提...     

Attitudes toward artificial intelligence in radiology with learner needs assessment within radiology residency programmes: a national multi-programme survey

INTRODUCTIONWe aimed to assess the attitudes and learner needs of radiology residents and faculty radiologists regarding artificial intelligence (AI) and machine learning (ML) in radiology.METHODSA web-based questionnaire, designed using SurveyMonkey, was sent out to residents and faculty radiologists in all three radiology residency programmes in Singapore. The questionnaire comprised four sections and aimed to evaluate respondents’ current experience, attempts at self-learning, perceptions of career prospects and expectations of an AI/ML curriculum in their residency programme. Respondents’ anonymity was ensured.RESULTSA total of 125 respondents (86 male, 39 female; 70 residents, 55 faculty radiologists) completed the questionnaire. The majority agreed that AI/ML will drastically change radiology practice (88.8%) and makes radiology more exciting (76.0%), and most would still choose to specialise in radiology if given a choice (80.0%). 64.8% viewed themselves as novices in their understanding of AI/ML, 76.0% planned to further advance their AI/ML knowledge and 67.2% were keen to get involved in an AI/ML research project. An overwhelming majority (84.8%) believed that AI/ML knowledge should be taught during residency, and most opined that this was as important as imaging physics and clinical skills/knowledge curricula (80.0% and 72.8%, respectively). More than half thought that their residency programme had not adequately implemented AI/ML teaching (59.2%). In subgroup analyses, male and tech-savvy respondents were more involved in AI/ML activities, leading to better technical understanding.CONCLUSIONA growing optimism towards radiology undergoing technological transformation and AI/ML implementation has led to a strong demand for an AI/ML curriculum in residency education.