An interpretable framework to identify responsive subgroups from clinical trials regarding treatment effects: Application to treatment of intracerebral hemorrhage.

Randomized Clinical trials (RCT) suffer from a high failure rate which could be caused by heterogeneous responses to treatment. Despite many models being developed to estimate heterogeneous treatment effects (HTE), there remains a lack of interpretable methods to identify responsive subgroups. This work aims to develop a framework to identify subgroups based on treatment effects that prioritize model interpretability. The proposed framework leverages an ensemble uplift tree method to generate descriptive decision rules that separate samples given estimated responses to the treatment. Subsequently, we select a complementary set of these decision rules and rank them using a sparse linear model. To address the trial's limited sample size problem, we proposed a data augmentation strategy by borrowing control patients from external studies and generating synthetic data. We apply the proposed framework to a failed randomized clinical trial for investigating an intracerebral hemorrhage therapy plan. The Qini-scores show that the proposed data augmentation strategy plan can boost the model's performance and the framework achieves greater interpretability by selecting complementary descriptive rules without compromising estimation quality. Our model derives clinically meaningful subgroups. Specifically, we find those patients with Diastolic Blood Pressure≥70 mm hg and Systolic Blood Pressure<215 mm hg benefit more from intensive blood pressure reduction therapy. The proposed interpretable HTE analysis framework offers a promising potential for extracting meaningful insight from RCTs with neutral treatment effects. By identifying responsive subgroups, our framework can contribute to developing personalized treatment strategies for patients more efficiently.

Identifying predictors of the tooth loss phenotype in a large periodontitis patient cohort using a machine learning approach.

OBJECTIVES: This study aimed to identify predictors associated with the tooth loss phenotype in a large periodontitis patient cohort in the university setting. METHODS: Information on periodontitis patients and nineteen factors identified at the initial visit was extracted from electronic health records. The primary outcome is tooth loss phenotype (presence or absence of tooth loss). Prediction models were built on significant factors (single or combinatory) selected by the RuleFit algorithm, and these factors were further adopted by regression models. Model performance was evaluated by Area Under the Receiver Operating Characteristic Curve (AUROC) and Area Under the Precision-Recall Curve (AUPRC). Associations between predictors and the tooth loss phenotype were also evaluated by classical statistical approaches to validate the performance of machine learning models. RESULTS: In total, 7840 patients were included. The machine learning model predicting the tooth loss phenotype achieved AUROC of 0.71 and AUPRC of 0.66. Age, periodontal diagnosis, number of missing teeth at baseline, furcation involvement, and tooth mobility were associated with the tooth loss phenotype in both machine learning and classical statistical models. CONCLUSIONS: The rule-based machine learning approach improves model explainability compared to classical statistical methods. However, the model's generalizability needs to be further validated by external datasets. CLINICAL SIGNIFICANCE: Predictors identified by the current machine learning approach using the RuleFit algorithm had clinically relevant thresholds in predicting the tooth loss phenotype in a large and diverse periodontitis patient cohort. The results of this study will assist clinicians in performing risk assessment for periodontitis at the initial visit.

aBnormal motION capture In aCute Stroke (BIONICS): A Low-Cost Tele-Evaluation Tool for Automated Assessment of Upper Extremity Function in Stroke Patients.

BACKGROUND: The incidence of stroke and stroke-related hemiparesis has been steadily increasing and is projected to become a serious social, financial, and physical burden on the aging population. Limited access to outpatient rehabilitation for these stroke survivors further deepens the healthcare issue and estranges the stroke patient demographic in rural areas. However, new advances in motion detection deep learning enable the use of handheld smartphone cameras for body tracking, offering unparalleled levels of accessibility. METHODS: In this study we want to develop an automated method for evaluation of a shortened variant of the Fugl-Meyer assessment, the standard stroke rehabilitation scale describing upper extremity motor function. We pair this technology with a series of machine learning models, including different neural network structures and an eXtreme Gradient Boosting model, to score 16 of 33 (49%) Fugl-Meyer item activities. RESULTS: In this observational study, 45 acute stroke patients completed at least 1 recorded Fugl-Meyer assessment for the training of the auto-scorers, which yielded average accuracies ranging from 78.1% to 82.7% item-wise. CONCLUSION: In this study, an automated method was developed for the evaluation of a shortened variant of the Fugl-Meyer assessment, the standard stroke rehabilitation scale describing upper extremity motor function. This novel method is demonstrated with potential to conduct telehealth rehabilitation evaluations and assessments with accuracy and availability.

Non-invasive arterial blood pressure measurement and SpO2 estimation using PPG signal: a deep learning framework.

BACKGROUND: Monitoring blood pressure and peripheral capillary oxygen saturation plays a crucial role in healthcare management for patients with chronic diseases, especially hypertension and vascular disease. However, current blood pressure measurement methods have intrinsic limitations; for instance, arterial blood pressure is measured by inserting a catheter in the artery causing discomfort and infection. METHOD: Photoplethysmogram (PPG) signals can be collected via non-invasive devices, and therefore have stimulated researchers' interest in exploring blood pressure estimation using machine learning and PPG signals as a non-invasive alternative. In this paper, we propose a Transformer-based deep learning architecture that utilizes PPG signals to conduct a personalized estimation of arterial systolic blood pressure, arterial diastolic blood pressure, and oxygen saturation. RESULTS: The proposed method was evaluated with a subset of 1,732 subjects from the publicly available ICU dataset MIMIC III. The mean absolute error is 2.52 ± 2.43 mmHg for systolic blood pressure, 1.37 ± 1.89 mmHg for diastolic blood pressure, and 0.58 ± 0.79% for oxygen saturation, which satisfies the requirements of the Association of Advancement of Medical Instrumentation standard and achieve grades A for the British Hypertension Society standard. CONCLUSIONS: The results indicate that our model meets clinical standards and could potentially boost the accuracy of blood pressure and oxygen saturation measurement to deliver high-quality healthcare.

A comprehensive evaluation of full-spectrum cell-free RNAs highlights cell-free RNA fragments for early-stage hepatocellular carcinoma detection.

BACKGROUND: Various studies have reported cell-free RNAs (cfRNAs) as noninvasive biomarkers for detecting hepatocellular carcinoma (HCC). However, they have not been independently validated, and some results are contradictory. We provided a comprehensive evaluation of various types of cfRNA biomarkers and a full mining of the biomarker potential of new features of cfRNA. METHODS: We first systematically reviewed reported cfRNA biomarkers and calculated dysregulated post-transcriptional events and cfRNA fragments. In 3 independent multicentre cohorts, we further selected 6 cfRNAs using RT-qPCR, built a panel called HCCMDP with AFP using machine learning, and internally and externally validated HCCMDP's performance. FINDINGS: We identified 23 cfRNA biomarker candidates from a systematic review and analysis of 5 cfRNA-seq datasets. Notably, we defined the cfRNA domain to describe cfRNA fragments systematically. In the verification cohort (n = 183), cfRNA fragments were more likely to be verified, while circRNA and chimeric RNA candidates were neither abundant nor stable as qPCR-based biomarkers. In the algorithm development cohort (n = 287), we build and test the panel HCCMDP with 6 cfRNA markers and AFP. In the independent validation cohort (n = 171), HCCMDP can distinguish HCC patients from control groups (all: AUC = 0.925; CHB: AUC = 0.909; LC: AUC = 0.916), and performs well in distinguishing early-stage HCC patients (all: AUC = 0.936; CHB: AUC = 0.917; LC: AUC = 0.928). INTERPRETATION: This study comprehensively evaluated full-spectrum cfRNA biomarker types for HCC detection, highlighted the cfRNA fragment as a promising biomarker type in HCC detection, and provided a panel HCCMDP. FUNDING: National Natural Science Foundation of China, and The National Key Basic Research Program (973 program).

Identifying predictors of tooth loss using a rule-based machine learning approach: A retrospective study at university-setting clinics.

BACKGROUND: This study aimed to identify predictors associated with tooth loss in a large periodontitis patient cohort in the university setting using the machine learning approach. METHODS: Information on periodontitis patients and 18 factors identified at the initial visit was extracted from electronic health records. A two-step machine learning pipeline was proposed to develop the tooth loss prediction model. The primary outcome is tooth loss count. The prediction model was built on significant factors (single or combination) selected by the RuleFit algorithm, and these factors were further adopted by the count regression model. Model performance was evaluated by root-mean-squared error (RMSE). Associations between predictors and tooth loss were also assessed by a classical statistical approach to validate the performance of the machine learning model. RESULTS: In total, 7840 patients were included. The machine learning model predicting tooth loss count achieved RMSE of 2.71. Age, smoking, frequency of brushing, frequency of flossing, periodontal diagnosis, bleeding on probing percentage, number of missing teeth at baseline, and tooth mobility were associated with tooth loss in both machine learning and classical statistical models. CONCLUSION: The two-step machine learning pipeline is feasible to predict tooth loss in periodontitis patients. Compared to classical statistical methods, this rule-based machine learning approach improves model explainability. However, the model's generalizability needs to be further validated by external datasets.

Explainable machine learning model for predicting the occurrence of postoperative malnutrition in children with congenital heart disease.

BACKGROUND & AIMS: Malnutrition is persistent in 50%-75% of children with congenital heart disease (CHD) after surgery, and early prediction is crucial for nutritional intervention. The aim of this study was to develop and validate machine learning (ML) models to predict the malnutrition status of children with CHD. We used explainable ML methods to provide insight into the model's predictions and outcomes. METHODS: This prospective cohort study included consecutive children with CHD admitted to the hospital from December 2017 to May 2020. The cohort data were divided into the training and test data sets based on the follow-up time. The outcome of the study was CHD child malnutrition 1 year after surgery, the primary outcome was an underweight status, and the secondary outcomes were stunted and wasting status. We used five ML algorithms with multiple features to construct prediction models, and the performance of these ML models was measured by an area under the receiver operating characteristic curve (AUC) analysis. We also used the permutation importance and SHapley Additive exPlanations (SHAP) to determine the importance of the selected features and interpret the ML models. RESULTS: We enrolled 536 children with CHD who underwent complete repair. The proportions of children with an underweight, stunted, or wasting status 1 year after surgery were 18.1% (97/536), 12.1% (65/536), and 17.5% (94/536), respectively. All patients contributed to the generation of 115 useable features, which allowed us to build models to predict malnutrition. Five prediction algorithms were used, and the XGBoost model achieved the greatest AUC in all outcomes. The results obtained from the permutation importance and SHAP analyses showed that the 1-month postoperative WAZ-score, discharge WAZ score and preoperative WAZ score were the top 3 important features in predicting an underweight status in the XGBoost algorithm. Regarding the stunted status, the top 3 important features were the 1-month postoperative HAZ score, discharge HAZ score, and aortic clamping time. Regarding the wasting status, the top 3 important features were the hospital length of stay, formula intake, and discharge WHZ-score. We also used a narrative case report as an example to describe the clinical manifestations and predicted the primary outcomes of two children. CONCLUSIONS: We developed an ML model (XGBoost) that provides accurate early predictions of malnutrition 1-year postoperatively in children with CHD. Because the ML model is explainable, it may better enable clinicians to better understand the reasoning underlying the outcome. Our study could aid in determining individual treatment and nutritional follow-up strategies for children with CHD.

Demand-driven spreading patterns of African swine fever in China.

African swine fever (ASF) is a highly contagious hemorrhagic viral disease of domestic and wild pigs. ASF has led to major economic losses and adverse impacts on livelihoods of stakeholders involved in the pork food system in many European and Asian countries. While the epidemiology of ASF virus (ASFV) is fairly well understood, there is neither any effective treatment nor vaccine. In this paper, we propose a novel method to model the spread of ASFV in China by integrating the data of pork import/export, transportation networks, and pork distribution centers. We first empirically analyze the overall spatiotemporal patterns of ASFV spread and conduct extensive experiments to evaluate the efficacy of a number of geographic distance measures. These empirical analyses of ASFV spread within China indicate that the first occurrence of ASFV has not been purely dependent on the geographical distance from existing infected regions. Instead, the pork supply-demand patterns have played an important role. Predictions based on a new distance measure achieve better performance in predicting ASFV spread among Chinese provinces and thus have the potential to enable the design of more effective control interventions.

Direct saponification preparation and analysis of free and conjugated phytosterols in sugarcane (Saccharum officinarum L.) by reversed-phase high-performance liquid chromatography.

A simple method based on direct saponification followed by RP-HPLC analysis was developed for quantification of free and conjugated sterols in sugarcane. Acid hydrolysis prior to alkaline saponification was used to determined acylated steryl glycoside and steryl glycoside in sugarcane. The applicability and generality of this method were improved with intensive investigation. Compared to traditional solvent extraction method, this method was more time saving and appropriate for characterization of sterol fractions in sugarcane. This method was successfully applied for determination of free and conjugated sterols in different sugarcane samples. The results exhibited that stigmasterol (varied from 883.3 ± 23.5 to 1823.9 ± 24.5 µg/g dry weigh) and ß-sitosterol (varied from 117.6 ± 19.9 to 801.4 ± 33.5 µg/g dry weight) were major phytosterols in the sugarcane sample, and their glycosylated forms accounted for almost 87.0% of stigmasterol and 87.5% of ß-sitosterol in sugarcane, respectively. In addition, among other parts of sugarcane, tips contained the greatest amount of phytosterols.

Extraction optimization by response surface methodology: Purification and characterization of phytosterol from sugarcane (Saccharum officinarum L.) rind.

A green, simple, and effective method for the extraction of sugarcane lipids from sugarcane rind was investigated by response surface methodology. The optimum conditions of technological progress obtained through response surface methodology were as follows: liquid-to-solid ratio 7.94: 1 mL/g, extraction temperature 50°C and extraction time 5.98 h. The practical sugarcane lipids extraction yield was 6.55 ± 0.28%, which was in good consistence with the predicted extraction yield of 6.47%. The results showed that the sugarcane lipids extraction yield obtained in optimum conditions increased by 1.16∼7.28-fold compared to the yields obtained in single-factor experiments. After saponification and SPE steps, the nonsaponifiable fraction of sugarcane lipids was analyzed by gas chromatography with mass spectrometry and high-performance liquid chromatography. ß-Sitosterol, stigmasterol, and campesterol were the prevailing phytosterols in the sample, while fucosterol, gramisterol, stigmast-7-en-3-ol, (3ß,5α,24S)-, stigmasta-4,6,22-trien-3α-ol, and cholest-8(14)-en-3ß-ol acetate were also identified as minor steroids. Furthermore, the content of ß-sitosterol and a mixture of campesterol and stigmasterol (quantified by high-performance liquid chromatography) was 44.18 mg/100 g dry weight and 43.20 mg stigmasterol/100 g dry weight, respectively. Our results indicate that sugarcane rind is a good source of phytosterol.