A Patient Similarity Network (CHDmap) to Predict Outcomes After Congenital Heart Surgery: Development and Validation Study.

Background: Although evidence-based medicine proposes personalized care that considers the best evidence, it still fails to address personal treatment in many real clinical scenarios where the complexity of the situation makes none of the available evidence applicable. "Medicine-based evidence" (MBE), in which big data and machine learning techniques are embraced to derive treatment responses from appropriately matched patients in real-world clinical practice, was proposed. However, many challenges remain in translating this conceptual framework into practice. Objective: This study aimed to technically translate the MBE conceptual framework into practice and evaluate its performance in providing general decision support services for outcomes after congenital heart disease (CHD) surgery. Methods: Data from 4774 CHD surgeries were collected. A total of 66 indicators and all diagnoses were extracted from each echocardiographic report using natural language processing technology. Combined with some basic clinical and surgical information, the distances between each patient were measured by a series of calculation formulas. Inspired by structure-mapping theory, the fusion of distances between different dimensions can be modulated by clinical experts. In addition to supporting direct analogical reasoning, a machine learning model can be constructed based on similar patients to provide personalized prediction. A user-operable patient similarity network (PSN) of CHD called CHDmap was proposed and developed to provide general decision support services based on the MBE approach. Results: Using 256 CHD cases, CHDmap was evaluated on 2 different types of postoperative prognostic prediction tasks: a binary classification task to predict postoperative complications and a multiple classification task to predict mechanical ventilation duration. A simple poll of the k-most similar patients provided by the PSN can achieve better prediction results than the average performance of 3 clinicians. Constructing logistic regression models for prediction using similar patients obtained from the PSN can further improve the performance of the 2 tasks (best area under the receiver operating characteristic curve=0.810 and 0.926, respectively). With the support of CHDmap, clinicians substantially improved their predictive capabilities. Conclusions: Without individual optimization, CHDmap demonstrates competitive performance compared to clinical experts. In addition, CHDmap has the advantage of enabling clinicians to use their superior cognitive abilities in conjunction with it to make decisions that are sometimes even superior to those made using artificial intelligence models. The MBE approach can be embraced in clinical practice, and its full potential can be realized.

Improved Transcranial Plane-Wave Imaging With Learned Speed-of-Sound Maps.

Although transcranial ultrasound plane-wave imaging (PWI) has promising clinical application prospects, studies have shown that variable speed-of-sound (SoS) would seriously damage the quality of ultrasound images. The mismatch between the conventional constant velocity assumption and the actual SoS distribution leads to the general blurring of ultrasound images. The optimization scheme for reconstructing transcranial ultrasound image is often solved using iterative methods like full-waveform inversion. These iterative methods are computationally expensive and based on prior magnetic resonance imaging (MRI) or computed tomography (CT) information. In contrast, the multi-stencils fast marching (MSFM) method can produce accurate time travel maps for the skull with heterogeneous acoustic speed. In this study, we first propose a convolutional neural network (CNN) to predict SoS maps of the skull from PWI channel data. Then, use these maps to correct the travel time to reduce transcranial aberration. To validate the performance of the proposed method, numerical, phantom and intact human skull studies were conducted using a linear array transducer (L11-5v, 128 elements, pitch = 0.3 mm). Numerical simulations demonstrate that for point targets, the lateral resolution of MSFM-restored images increased by 65%, and the center position shift decreased by 89%. For the cyst targets, the eccentricity of the fitting ellipse decreased by 75%, and the center position shift decreased by 58%. In the phantom study, the lateral resolution of MSFM-restored images was increased by 49%, and the position shift was reduced by 1.72 mm. This pipeline, termed AutoSoS, thus shows the potential to correct distortions in real-time transcranial ultrasound imaging, as demonstrated by experiments on the intact human skull.

Intraoperative Transfusion of Autologous Blood Protects from Acute Kidney Injury after Pediatric Congenital Heart Surgery.

Background: Acute kidney injury (AKI) is a common complication after pediatric cardiac surgery. And autologous blood transfusion (ABT) is an important predictor of postoperative AKI. Unlike previous studies, which mainly focused on the correlation between ABT and AKI, the current study focuses heavily on the causal relationship between them, thus providing guidance for the treatment of patients during hospitalization to reduce the occurrence of AKI. Methods: A retrospective cohort of 3386 patients extracted from the Pediatric Intensive Care database was used for statistical analysis, multifactorial analysis, and causal inference. Characteristics that were correlated with ABT and AKI were categorized as confounders, instrumental variables, and effect modifiers, and were entered into the DoWhy causal inference model to determine causality. The calculated average treatment effect (ATE) was compared with the results of the multifactorial analysis. Results: The adjusted odds ratio (OR) for ABT volume was obtained by multifactorial analysis as 0.964. The DoWhy model refute test was able to indicate a causal relationship between ABT and AKI. Any ABT reduces AKI about 15.3%-18.8% by different estimation methods. The ATE regarding the amount of ABT was -0.0088, suggesting that every 1 mL/kg of ABT reduced the risk of AKI by 0.88%. Conclusions: Intraoperative transfusion of autologous blood can have a protective effect against postoperative AKI.

A time-aware attention model for prediction of acute kidney injury after pediatric cardiac surgery.

OBJECTIVE: Acute kidney injury (AKI) is a common complication after pediatric cardiac surgery, and the early detection of AKI may allow for timely preventive or therapeutic measures. However, current AKI prediction researches pay less attention to time information among time-series clinical data and model building strategies that meet complex clinical application scenario. This study aims to develop and validate a model for predicting postoperative AKI that operates sequentially over individual time-series clinical data. MATERIALS AND METHODS: A retrospective cohort of 3386 pediatric patients extracted from PIC database was used for training, calibrating, and testing purposes. A time-aware deep learning model was developed and evaluated from 3 clinical perspectives that use different data collection windows and prediction windows to answer different AKI prediction questions encountered in clinical practice. We compared our model with existing state-of-the-art models from 3 clinical perspectives using the area under the receiver operating characteristic curve (ROC AUC) and the area under the precision-recall curve (PR AUC). RESULTS: Our proposed model significantly outperformed the existing state-of-the-art models with an improved average performance for any AKI prediction from the 3 evaluation perspectives. This model predicted 91% of all AKI episodes using data collected at 24 h after surgery, resulting in a ROC AUC of 0.908 and a PR AUC of 0.898. On average, our model predicted 83% of all AKI episodes that occurred within the different time windows in the 3 evaluation perspectives. The calibration performance of the proposed model was substantially higher than the existing state-of-the-art models. CONCLUSIONS: This study showed that a deep learning model can accurately predict postoperative AKI using perioperative time-series data. It has the potential to be integrated into real-time clinical decision support systems to support postoperative care planning.

An ensemble learning with active sampling to predict the prognosis of postoperative non-small cell lung cancer patients.

BACKGROUND: Lung cancer is the leading cause of cancer death worldwide. Prognostic prediction plays a vital role in the decision-making process for postoperative non-small cell lung cancer (NSCLC) patients. However, the high imbalance ratio of prognostic data limits the development of effective prognostic prediction models. METHODS: In this study, we present a novel approach, namely ensemble learning with active sampling (ELAS), to tackle the imbalanced data problem in NSCLC prognostic prediction. ELAS first applies an active sampling mechanism to query the most informative samples to update the base classifier to give it a new perspective. This training process is repeated until no enough samples are queried. Next, an internal validation set is employed to evaluate the base classifiers, and the ones with the best performances are integrated as the ensemble model. Besides, we set up multiple initial training data seeds and internal validation sets to ensure the stability and generalization of the model. RESULTS: We verified the effectiveness of the ELAS on a real clinical dataset containing 1848 postoperative NSCLC patients. Experimental results showed that the ELAS achieved the best averaged 0.736 AUROC value and 0.453 AUPRC value for 6 prognostic tasks and obtained significant improvements in comparison with the SVM, AdaBoost, Bagging, SMOTE and TomekLinks. CONCLUSIONS: We conclude that the ELAS can effectively alleviate the imbalanced data problem in NSCLC prognostic prediction and demonstrates good potential for future postoperative NSCLC prognostic prediction.

An end-to-end tracking method for polyp detectors in colonoscopy videos.

Deep learning based computer-aided diagnosis technology demonstrates an encouraging performance in aspect of polyp lesion detection on reducing the miss rate of polyps during colonoscopies. However, to date, few studies have been conducted for tracking polyps that have been detected in colonoscopy videos, which is an essential and intuitive issue in clinical intelligent video analysis task (e.g. lesion counting, lesion retrieval, report generation). In the paradigm of conventional tracking-by-detection system, detection task for lesion localization is separated from the tracking task for cropped lesions re-identification. In the multi object tracking problem, each target is supposed to be tracked by invoking a tracker after the detector, which introduces multiple inferences and leads to external resource and time consumption. To tackle these problems, we proposed a plug-in module named instance tracking head (ITH) for synchronous polyp detection and tracking, which can be simply inserted into object detection frameworks. It embeds a feature-based polyp tracking procedure into the detector frameworks to achieve multi-task model training. ITH and detection head share the model backbone for low level feature extraction, and then low level feature flows into the separate branches for task-driven model training. For feature maps from the same receptive field, the region of interest head assigns these features to the detection head and the ITH, respectively, and outputs the object category, bounding box coordinates, and instance feature embedding simultaneously for each specific polyp target. We also proposed a method based on similarity metric learning. The method makes full use of the prior boxes in the object detector to provide richer and denser instance training pairs, to improve the performance of the model evaluation on the tracking task. Compared with advanced tracking-by-detection paradigm methods, detectors with proposed ITH can obtain comparative tracking performance but approximate 30% faster speed. Optimized model based on Scaled-YOLOv4 detector with ITH illustrates good trade-off between detection (mAP 91.70%) and tracking (MOTA 92.50% and Rank-1 Acc 88.31%) task at the frame rate of 66 FPS. The proposed structure demonstrates the potential to aid clinicians in real-time detection with online tracking or offline retargeting of polyp instances during colonoscopies.

Making Specific Plan Improves Physical Activity and Healthy Eating for Community-Dwelling Patients With Chronic Conditions: A Systematic Review and Meta-Analysis.

Background: Implementation intention formed by making a specific action plan has been proved effective in improving physical activity (PA) and dietary behavior (DB) for the general, healthy population, but there has been no meta-analysis of their effectiveness for patients with chronic conditions. This research aims to analyze several explanatory factors and overall effect of implementation intention on behavioral and health-related outcomes among community-dwelling patients. Methods: We searched CIHNAL (EBSCO), PUBMED, Web of Science, Science Direct, SAGE Online, Springer Link, Taylor & Francis, Scopus, Wiley Online Library, CNKI, and five other databases for eligible studies. Random-effects meta-analysis was conducted to estimate effect sizes of implementation intention on outcomes, including PA, DB, weight, and body mass index. And the eligible studies were assessed by the Cochrane Collaboration's tool for risk of bias assessment. Sensitivity analysis adopted sequential algorithm and the p-curve analysis method. Results: A total of 54 studies were identified. Significant small effect sizes of the intervention were found for PA [standard mean difference (SMD) 0.24, 95% confidence interval (CI) (0.10, 0.39)] and for the DB outcome [SMD -0.25, 95% CI (-0.34, -0.15)]. In moderation analysis, the intervention was more effective in improving PA for men (p < 0.001), older adults (p = 0.006), and obese/overweight patients with complications (p = 0.048) and when the intervention was delivered by a healthcare provider (p = 0.01). Conclusion: Implementation intentions are effective in improving PA and DB for community dwelling patients with chronic conditions. The review provides evidence to support the future application of implementation intention intervention. Besides, the findings from this review offer different directions to enhance the effectiveness of this brief and potential intervention in improving patients' PA and DB. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=160491.

Multi-label recognition of cancer-related lesions with clinical priors on white-light endoscopy.

Deep learning-based computer-aided diagnosis techniques have demonstrated encouraging performance in endoscopic lesion identification and detection, and have reduced the rate of missed and false detections of disease during endoscopy. However, the interpretability of the model-based results has not been adequately addressed by existing methods. This phenomenon is directly manifested by a significant bias in the representation of feature localization. Good recognition models experience severe feature localization errors, particularly for lesions with subtle morphological features, and such unsatisfactory performance hinders the clinical deployment of models. To effectively alleviate this problem, we proposed a solution to optimize the localization bias in feature representations of cancer-related recognition models that is difficult to accurately label and identify in clinical practice. Optimization was performed in the training phase of the model through the proposed data augmentation method and auxiliary loss function based on clinical priors. The data augmentation method, called partial jigsaw, can "break" the spatial structure of lesion-independent image blocks and enrich the data feature space to decouple the interference of background features on the space and focus on fine-grained lesion features. The annotation-based auxiliary loss function used class activation maps for sample distribution correction and led the model to present localization representation converging on the gold standard annotation of visualization maps. The results show that with the improvement of our method, the precision of model recognition reached an average of 92.79%, an F1-score of 92.61%, and accuracy of 95.56% based on a dataset constructed from 23 hospitals. In addition, we quantified the evaluation representation of visualization feature maps. The improved model yielded significant offset correction results for visualized feature maps compared with the baseline model. The average visualization-weighted positive coverage improved from 51.85% to 83.76%. The proposed approach did not change the deployment capability and inference speed of the original model and can be incorporated into any state-of-the-art neural network. It also shows the potential to provide more accurate localization inference results and assist in clinical examinations during endoscopies.

Explainable machine-learning predictions for complications after pediatric congenital heart surgery.

The quality of treatment and prognosis after pediatric congenital heart surgery remains unsatisfactory. A reliable prediction model for postoperative complications of congenital heart surgery patients is essential to enable prompt initiation of therapy and improve the quality of prognosis. Here, we develop an interpretable machine-learning-based model that integrates patient demographics, surgery-specific features and intraoperative blood pressure data for accurately predicting complications after pediatric congenital heart surgery. We used blood pressure variability and the k-means algorithm combined with a smoothed formulation of dynamic time wrapping to extract features from time-series data. In addition, SHAP framework was used to provide explanations of the prediction. Our model achieved the best performance both in binary and multi-label classification compared with other consensus-based risk models. In addition, this explainable model explains why a prediction was made to help improve the clinical understanding of complication risk and generate actionable knowledge in practice. The combination of model performance and interpretability is easy for clinicians to trust and provide insight into how they should respond before the condition worsens after pediatric congenital heart surgery.

Simultaneous Recognition of Atrophic Gastritis and Intestinal Metaplasia on White Light Endoscopic Images Based on Convolutional Neural Networks: A Multicenter Study.

INTRODUCTION: Patients with atrophic gastritis (AG) or gastric intestinal metaplasia (GIM) have elevated risk of gastric adenocarcinoma. Endoscopic screening and surveillance have been implemented in high incidence countries. The study aimed to evaluate the accuracy of a deep convolutional neural network (CNN) for simultaneous recognition of AG and GIM. METHODS: Archived endoscopic white light images with corresponding gastric biopsies were collected from 14 hospitals located in different regions of China. Corresponding images by anatomic sites containing AG, GIM, and chronic non-AG were categorized using pathology reports. The participants were randomly assigned (8:1:1) to the training cohort for developing the CNN model (TResNet), the validation cohort for fine-tuning, and the test cohort for evaluating the diagnostic accuracy. The area under the curve (AUC), sensitivity, specificity, and accuracy with 95% confidence interval (CI) were calculated. RESULTS: A total of 7,037 endoscopic images from 2,741 participants were used to develop the CNN for recognition of AG and/or GIM. The AUC for recognizing AG was 0.98 (95% CI 0.97-0.99) with sensitivity, specificity, and accuracy of 96.2% (95% CI 94.2%-97.6%), 96.4% (95% CI 94.8%-97.9%), and 96.4% (95% CI 94.4%-97.8%), respectively. The AUC for recognizing GIM was 0.99 (95% CI 0.98-1.00) with sensitivity, specificity, and accuracy of 97.9% (95% CI 96.2%-98.9%), 97.5% (95% CI 95.8%-98.6%), and 97.6% (95% CI 95.8%-98.6%), respectively. DISCUSSION: CNN using endoscopic white light images achieved high diagnostic accuracy in recognizing AG and GIM.

A novel deep learning approach to extract Chinese clinical entities for lung cancer screening and staging.

BACKGROUND: Computed tomography (CT) reports record a large volume of valuable information about patients' conditions and the interpretations of radiology images from radiologists, which can be used for clinical decision-making and further academic study. However, the free-text nature of clinical reports is a critical barrier to use this data more effectively. In this study, we investigate a novel deep learning method to extract entities from Chinese CT reports for lung cancer screening and TNM staging. METHODS: The proposed approach presents a new named entity recognition algorithm, namely the BERT-based-BiLSTM-Transformer network (BERT-BTN) with pre-training, to extract clinical entities for lung cancer screening and staging. Specifically, instead of traditional word embedding methods, BERT is applied to learn the deep semantic representations of characters. Following the long short-term memory layer, a Transformer layer is added to capture the global dependencies between characters. Besides, pre-training technique is employed to alleviate the problem of insufficient labeled data. RESULTS: We verify the effectiveness of the proposed approach on a clinical dataset containing 359 CT reports collected from the Department of Thoracic Surgery II of Peking University Cancer Hospital. The experimental results show that the proposed approach achieves an 85.96% macro-F1 score under exact match scheme, which improves the performance by 1.38%, 1.84%, 3.81%,4.29%,5.12%,5.29% and 8.84% compared to BERT-BTN, BERT-LSTM, BERT-fine-tune, BERT-Transformer, FastText-BTN, FastText-BiLSTM and FastText-Transformer, respectively. CONCLUSIONS: In this study, we developed a novel deep learning method, i.e., BERT-BTN with pre-training, to extract the clinical entities from Chinese CT reports. The experimental results indicate that the proposed approach can efficiently recognize various clinical entities about lung cancer screening and staging, which shows the potential for further clinical decision-making and academic research.

Combining gene expression signature with clinical features for survival stratification of gastric cancer.

The AJCC staging system is considered as the golden standard in clinical practice. However, it remains some pitfalls in assessing the prognosis of gastric cancer (GC) patients with similar clinicopathological characteristics. We aim to develop a new clinic and genetic risk score (CGRS) to improve the prognosis prediction of GC patients. We established genetic risk score (GRS) based on nine-gene signature including APOD, CCDC92, CYS1, GSDME, ST8SIA5, STARD3NL, TIMEM245, TSPYL5, and VAT1 based on the gene expression profiles of the training set from the Asian Cancer Research Group (ACRG) cohort by LASSO-Cox regression algorithms. CGRS was established by integrating GRS with clinical risk score (CRS) derived from Surveillance, Epidemiology, and End Results (SEER) database. GRS and CGRS dichotomized GC patients into high and low risk groups with significantly different prognosis in four independent cohorts with different data types, such as microarray, RNA sequencing and qRT-PCR (all HR > 1, all P < 0.001). Both GRS and CGRS were prognostic signatures independent of the AJCC staging system. Receiver operating characteristic (ROC) analysis showed that area under ROC curve of CGRS was larger than that of the AJCC staging system in most cohorts we studied. Nomogram and web tool (http://39.100.117.92/CGRS/) based on CGRS were developed for clinicians to conveniently assess GC prognosis in clinical practice. CGRS integrating genetic signature with clinical features shows strong robustness in predicting GC prognosis, and can be easily applied in clinical practice through the web application.

Using Z-score to optimize population-specific DDH screening: a retrospective study in Hangzhou, China.

BACKGROUND: DDH (Developmental Dysplasia of the Hip) screening can potentially avert many morbidities and reduce costs. The debate about universal vs. selective DDH ultrasonography screening in different countries revolves to a large extent around effectiveness, cost, and the possibility of overdiagnosis and overtreatment. In this study, we proposed and evaluated a Z-score enhanced Graf method to optimize population-specific DDH screening. METHODS: A total of 39,710 history ultrasonography hip examinations were collected to establish a sex, side specific and age-based Z-scores model using the local regression method. The correlation between Z-scores and classic Graf types was analyzed. Four thousand two hundred twenty-nine cases with follow-up ultrasonographic examinations and 5284 cases with follow-up X-ray examinations were used to evaluate the false positive rate of the first examination based on the subsequent examinations. The results using classic Graf types and the Z-score enhanced types were compared. RESULTS: The Z-score enhanced Graf types were highly correlated with the classic Graf's classification (R = 0.67, p < 0.001). Using the Z-scores ≥2 as a threshold could reduce by 86.56 and 80.44% the false positives in the left and right hips based on the follow-up ultrasonographic examinations, and reduce by 78.99% false-positive cases based on the follow-up X-ray examinations, respectively. CONCLUSIONS: Using an age, sex and side specific Z-scores enhanced Graf's method can better control the false positive rate in DDH screening among different populations.

Understanding risk factors for postoperative mortality in neonates based on explainable machine learning technology.

PURPOSE: We aimed to introduce an explainable machine learning technology to help clinicians understand the risk factors for neonatal postoperative mortality at different levels. METHODS: A total of 1481 neonatal surgeries performed between May 2016 and December 2019 at a children's hospital were included in this study. Perioperative variables, including vital signs during surgery, were collected and used to predict postoperative mortality. Several widely used machine learning methods were trained and evaluated on split datasets. The model with the best performance was explained by SHAP (SHapley Additive exPlanations) at different levels. RESULTS: The random forest model achieved the best performance with an area under the receiver operating characteristic curve of 0.72 in the validation set. TreeExplainer of SHAP was used to identify the risk factors for neonatal postoperative mortality. The explainable machine learning model not only explains the risk factors identified by traditional statistical analysis but also identifies additional risk factors. The visualization of feature contributions at different levels by SHAP makes the "black-box" machine learning model easily understood by clinicians and families. Based on this explanation, vital signs during surgery play an important role in eventual survival. CONCLUSIONS: The explainable machine learning model not only exhibited good performance in predicting neonatal surgical mortality but also helped clinicians understand each risk factor and each individual case.

Prediction of complications after paediatric cardiac surgery.

OBJECTIVES: Our objectives were to identify the risk factors for postoperative complications after paediatric cardiac surgery, develop a tool for predicting postoperative complications and compare it with other risk adjustment tools of congenital heart disease. METHODS: A total of 2308 paediatric patients who had undergone cardiac surgeries with cardiopulmonary bypass support in a single centre were included in this study. A univariate analysis was performed to determine the association between perioperative variables and postoperative complications. Statistically significant variables were integrated into a synthetic minority oversampling technique-based XGBoost model which is an implementation of gradient boosted decision trees designed for speed and performance. The 7 traditional risk assessment tools used to generate the logistic regression model as the benchmark in the evaluation included the Aristotle Basic score and category, Risk Adjustment for Congenital Heart Surgery (RACHS-1), Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery (STS-EACTS) mortality score and category and STS morbidity score and category. RESULTS: Our XGBoost prediction model showed the best prediction performance (area under the receiver operating characteristic curve = 0.82) when compared with these risk adjustment models. However, all of these models exhibited a relatively lower sensitivity due to imbalanced classes. The sensitivity of our optimization approach (synthetic minority oversampling technique-based XGBoost) was 0.74, which was significantly higher than the average sensitivity of the traditional models of 0.26. Furthermore, the postoperative length of hospital stay, length of cardiac intensive care unit stay and length of mechanical ventilation duration were significantly increased for patients who experienced postoperative complications. CONCLUSIONS: Postoperative complications of paediatric cardiac surgery can be predicted based on perioperative data using our synthetic minority oversampling technique-based XGBoost model before deleterious outcomes ensue.

Renal Biopsy Recommendation Based on Text Understanding.

Due to various etiologies and pathogenesis of kidney diseases, an invasive procedure called renal biopsy may be needed to determine the specific type of kidney disease, its severity, and the best treatment for it. This study aims to detmine if a text understanding technology based on admission records can recommend such an invasive procedure objectively. To understand clinical documents from nephrology, a semi-automatic learning-based lexicon construction method based on CRF and Word2vec was used. We constructed a dictionary of symptom terms for the nephrology department from clinical document, and then extracted patients' symptoms and detected their negation from admission notes. Combined with the preliminary diagnosis given by the doctor, an eigenvector was produced and fed to a machine learning classifier. When compared to the gold standard marked by physicians, the final recommendation achieved 83.5% accuracy, 80.6% precision, 76.6% recall, and 78.6% f1-measure respectively.

Real-time gastric polyp detection using convolutional neural networks.

Computer-aided polyp detection in gastric gastroscopy has been the subject of research over the past few decades. However, despite significant advances, automatic polyp detection in real time is still an unsolved problem. In this paper, we report on a convolutional neural network (CNN) for polyp detection that is constructed based on Single Shot MultiBox Detector (SSD) architecture and which we call SSD for Gastric Polyps (SSD-GPNet). To take full advantages of feature maps' information from the feature pyramid and to acquire higher accuracy, we re-use information that is abandoned by Max-Pooling layers. In other words, we reuse the lost data from the pooling layers and concatenate that data as extra feature maps to contribute to classification and detection. Meanwhile, in the feature pyramid, we concatenate feature maps of the lower layers and feature maps that are deconvolved from upper layers to make explicit relationships between layers and to effectively increase the number of channels. The results show that our enhanced SSD for gastric polyp detection can realize real-time polyp detection with 50 frames per second (FPS) and can improve the mean average precision (mAP) from 88.5% to 90.4%, with only a little loss in time-performance. And the further experiment shows that SSD-GPNet has excellent performance in improving polyp detection recalls over 10% (p = 0.00053), especially in small polyp detection. This can help endoscopic physicians more easily find missed polyps and decrease the gastric polyp miss rate. It may be applicable in daily clinical practice to reduce the burden on physicians.

A meta-model for computer executable dynamic clinical safety checklists.

BACKGROUND: Safety checklist is a type of cognitive tool enforcing short term memory of medical workers with the purpose of reducing medical errors caused by overlook and ignorance. To facilitate the daily use of safety checklists, computerized systems embedded in the clinical workflow and adapted to patient-context are increasingly developed. However, the current hard-coded approach of implementing checklists in these systems increase the cognitive efforts of clinical experts and coding efforts for informaticists. This is due to the lack of a formal representation format that is both understandable by clinical experts and executable by computer programs. METHODS: We developed a dynamic checklist meta-model with a three-step approach. Dynamic checklist modeling requirements were extracted by performing a domain analysis. Then, existing modeling approaches and tools were investigated with the purpose of reusing these languages. Finally, the meta-model was developed by eliciting domain concepts and their hierarchies. The feasibility of using the meta-model was validated by two case studies. The meta-model was mapped to specific modeling languages according to the requirements of hospitals. RESULTS: Using the proposed meta-model, a comprehensive coronary artery bypass graft peri-operative checklist set and a percutaneous coronary intervention peri-operative checklist set have been developed in a Dutch hospital and a Chinese hospital, respectively. The result shows that it is feasible to use the meta-model to facilitate the modeling and execution of dynamic checklists. CONCLUSIONS: We proposed a novel meta-model for the dynamic checklist with the purpose of facilitating creating dynamic checklists. The meta-model is a framework of reusing existing modeling languages and tools to model dynamic checklists. The feasibility of using the meta-model is validated by implementing a use case in the system.

Gastric precancerous diseases classification using CNN with a concise model.

Gastric precancerous diseases (GPD) may deteriorate into early gastric cancer if misdiagnosed, so it is important to help doctors recognize GPD accurately and quickly. In this paper, we realize the classification of 3-class GPD, namely, polyp, erosion, and ulcer using convolutional neural networks (CNN) with a concise model called the Gastric Precancerous Disease Network (GPDNet). GPDNet introduces fire modules from SqueezeNet to reduce the model size and parameters about 10 times while improving speed for quick classification. To maintain classification accuracy with fewer parameters, we propose an innovative method called iterative reinforced learning (IRL). After training GPDNet from scratch, we apply IRL to fine-tune the parameters whose values are close to 0, and then we take the modified model as a pretrained model for the next training. The result shows that IRL can improve the accuracy about 9% after 6 iterations. The final classification accuracy of our GPDNet was 88.90%, which is promising for clinical GPD recognition.