Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

Surgery Surgery

Empowering Caseworkers to Better Serve the Most Vulnerable with a Cloud-Based Care Management Solution.

In Applied clinical informatics ; h5-index 22.0

BACKGROUND :  Care-management tools are typically utilized for chronic disease management. Sonoma County government agencies employed advanced health information technologies, artificial intelligence (AI), and interagency process improvements to help transform health and health care for socially disadvantaged groups and other displaced individuals.

OBJECTIVES :  The objective of this case report is to describe how an integrated data hub and care-management solution streamlined care coordination of government services during a time of community-wide crisis.

METHODS :  This innovative application of care-management tools created a bridge between social and clinical determinants of health and used a three-step approach-access, collaboration, and innovation. The program Accessing Coordinated Care to Empower Self Sufficiency Sonoma was established to identify and match the most vulnerable residents with services to improve their well-being. Sonoma County created an Interdepartmental Multidisciplinary Team to deploy coordinated cross-departmental services (e.g., health and human services, housing services, probation) to support individuals experiencing housing insecurity. Implementation of a data integration hub (DIH) and care management and coordination system (CMCS) enabled integration of siloed data and services into a unified view of citizen status, identification of clinical and social determinants of health from structured and unstructured sources, and algorithms to match clients across systems.

RESULTS :  The integrated toolset helped 77 at-risk individuals in crisis through coordinated care plans and access to services in a time of need. Two case examples illustrate the specific care and services provided individuals with complex needs after the 2017 Sonoma County wildfires.

CONCLUSION :  Unique application of a care-management solution transformed health and health care for individuals fleeing from their homes and socially disadvantaged groups displaced by the Sonoma County wildfires. Future directions include expanding the DIH and CMCS to neighboring counties to coordinate care regionally. Such solutions might enable innovative care-management solutions across a variety of public, private, and nonprofit services.

Snowdon Jane L, Robinson Barbie, Staats Carolyn, Wolsey Kenneth, Sands-Lincoln Megan, Strasheim Thomas, Brotman David, Keating Katie, Schnitter Elizabeth, Jackson Gretchen, Kassler William


General General

Artificial Intelligence-Assisted Colonoscopy for Detection of Colon Polyps: a Prospective, Randomized Cohort Study.

In Journal of gastrointestinal surgery : official journal of the Society for Surgery of the Alimentary Tract

BACKGROUND AND AIMS : Improving the rate of polyp detection is an important measure to prevent colorectal cancer (CRC). Real-time automatic polyp detection systems, through deep learning methods, can learn and perform specific endoscopic tasks previously performed by endoscopists. The purpose of this study was to explore whether a high-performance, real-time automatic polyp detection system could improve the polyp detection rate (PDR) in the actual clinical environment.

METHODS : The selected patients underwent same-day, back-to-back colonoscopies in a random order, with either traditional colonoscopy or artificial intelligence (AI)-assisted colonoscopy performed first by different experienced endoscopists (> 3000 colonoscopies). The primary outcome was the PDR. It was registered with . (NCT047126265).

RESULTS : In this study, we randomized 150 patients. The AI system significantly increased the PDR (34.0% vs 38.7%, p < 0.001). In addition, AI-assisted colonoscopy increased the detection of polyps smaller than 6 mm (69 vs 91, p < 0.001), but no difference was found with regard to larger lesions.

CONCLUSIONS : A real-time automatic polyp detection system can increase the PDR, primarily for diminutive polyps. However, a larger sample size is still needed in the follow-up study to further verify this conclusion.

TRIAL REGISTRATION : Identifier: NCT047126265.

Luo Yuchen, Zhang Yi, Liu Ming, Lai Yihong, Liu Panpan, Wang Zhen, Xing Tongyin, Huang Ying, Li Yue, Li Aiming, Wang Yadong, Luo Xiaobei, Liu Side, Han Zelong


Artificial intelligence, Colonoscopy, Computer-aided diagnose

General General

An immune-related gene signature for determining Ewing sarcoma prognosis based on machine learning.

In Journal of cancer research and clinical oncology

PURPOSE : Ewing sarcoma (ES) is one of the most common malignant bone tumors in children and adolescents. The immune microenvironment plays an important role in the development of ES. Here, we developed an optimal signature for determining ES patient prognosis based on immune-related genes (IRGs).

METHODS : We analyzed the ES gene expression profile dataset, GSE17679, from the GEO database and extracted differential expressed IRGs (DEIRGs). Then, we conducted functional correlation and protein-protein interaction (PPI) analyses of the DEIRGs and used the machine learning algorithm-iterative Lasso Cox regression analysis to build an optimal DEIRG signature. In addition, we applied ES samples from the ICGC database to test the optimal gene signature. We performed univariate and multivariate Cox regressions on clinicopathological characteristics and optimal gene signature to evaluate whether signature is an important prognostic factor. Finally, we calculated the infiltration of 24 immune cells in ES using the ssGSEA algorithm, and analyzed the correlation between the DEIRGs in the optimal gene signature and immune cells.

RESULTS : A total of 249 DEIRGs were screened and an 11-gene signature with the strongest correlation with patient prognoses was analyzed using a machine learning algorithm. The 11-gene signature also had a high prognostic value in the ES external verification set. Univariate and multivariate Cox regression analyses showed that 11-gene signature is an independent prognostic factor. We found that macrophages and cytotoxic, CD8 T, NK, mast, B, NK CD56bright, TEM, TCM, and Th2 cells were significantly related to patient prognoses; the infiltration of cytotoxic and CD8 T cells in ES was significantly different. By correlating prognostic biomarkers with immune cell infiltration, we found that FABP4 and macrophages, and NDRG1 and Th2 cells had the strongest correlation.

CONCLUSION : Overall, the IRG-related 11-gene signature can be used as a reliable ES prognostic biomarker and can provide guidance for personalized ES therapy.

Ren En-Hui, Deng Ya-Jun, Yuan Wen-Hua, Wu Zuo-Long, Zhang Guang-Zhi, Xie Qi-Qi


Ewing sarcoma, Immune cell infiltration, Iterative Lasso regression, Machine learning, Prognosis analysis

Internal Medicine Internal Medicine

Risk prediction of delirium in hospitalized patients using machine learning: An implementation and prospective evaluation study.

In Journal of the American Medical Informatics Association : JAMIA

OBJECTIVE : Machine learning models trained on electronic health records have achieved high prognostic accuracy in test datasets, but little is known about their embedding into clinical workflows. We implemented a random forest-based algorithm to identify hospitalized patients at high risk for delirium, and evaluated its performance in a clinical setting.

MATERIALS AND METHODS : Delirium was predicted at admission and recalculated on the evening of admission. The defined prediction outcome was a delirium coded for the recent hospital stay. During 7 months of prospective evaluation, 5530 predictions were analyzed. In addition, 119 predictions for internal medicine patients were compared with ratings of clinical experts in a blinded and nonblinded setting.

RESULTS : During clinical application, the algorithm achieved a sensitivity of 74.1% and a specificity of 82.2%. Discrimination on prospective data (area under the receiver-operating characteristic curve = 0.86) was as good as in the test dataset, but calibration was poor. The predictions correlated strongly with delirium risk perceived by experts in the blinded (r = 0.81) and nonblinded (r = 0.62) settings. A major advantage of our setting was the timely prediction without additional data entry.

DISCUSSION : The implemented machine learning algorithm achieved a stable performance predicting delirium in high agreement with expert ratings, but improvement of calibration is needed. Future research should evaluate the acceptance of implemented machine learning algorithms by health professionals.

CONCLUSIONS : Our study provides new insights into the implementation process of a machine learning algorithm into a clinical workflow and demonstrates its predictive power for delirium.

Jauk Stefanie, Kramer Diether, Gro├čauer Birgit, Rienm├╝ller Susanne, Avian Alexander, Berghold Andrea, Leodolter Werner, Schulz Stefan


Machine learning, clinical decision support, delirium, electronic health records, prospective studies

General General

The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task on clinical concept normalization for clinical records.

In Journal of the American Medical Informatics Association : JAMIA

OBJECTIVE : The 2019 National Natural language processing (NLP) Clinical Challenges (n2c2)/Open Health NLP (OHNLP) shared task track 3, focused on medical concept normalization (MCN) in clinical records. This track aimed to assess the state of the art in identifying and matching salient medical concepts to a controlled vocabulary. In this paper, we describe the task, describe the data set used, compare the participating systems, present results, identify the strengths and limitations of the current state of the art, and identify directions for future research.

MATERIALS AND METHODS : Participating teams were provided with narrative discharge summaries in which text spans corresponding to medical concepts were identified. This paper refers to these text spans as mentions. Teams were tasked with normalizing these mentions to concepts, represented by concept unique identifiers, within the Unified Medical Language System. Submitted systems represented 4 broad categories of approaches: cascading dictionary matching, cosine distance, deep learning, and retrieve-and-rank systems. Disambiguation modules were common across all approaches.

RESULTS : A total of 33 teams participated in the MCN task. The best-performing team achieved an accuracy of 0.8526. The median and mean performances among all teams were 0.7733 and 0.7426, respectively.

CONCLUSIONS : Overall performance among the top 10 teams was high. However, several mention types were challenging for all teams. These included mentions requiring disambiguation of misspelled words, acronyms, abbreviations, and mentions with more than 1 possible semantic type. Also challenging were complex mentions of long, multi-word terms that may require new ways of extracting and representing mention meaning, the use of domain knowledge, parse trees, or hand-crafted rules.

Henry Sam, Wang Yanshan, Shen Feichen, Uzuner Ozlem


clinical narratives, concept normalization, machine learning, natural language processing

General General

Graph-based regularization for regression problems with alignment and highly-correlated designs.

In SIAM journal on mathematics of data science

Sparse models for high-dimensional linear regression and machine learning have received substantial attention over the past two decades. Model selection, or determining which features or covariates are the best explanatory variables, is critical to the interpretability of a learned model. Much of the current literature assumes that covariates are only mildly correlated. However, in many modern applications covariates are highly correlated and do not exhibit key properties (such as the restricted eigenvalue condition, restricted isometry property, or other related assumptions). This work considers a high-dimensional regression setting in which a graph governs both correlations among the covariates and the similarity among regression coefficients - meaning there is alignment between the covariates and regression coefficients. Using side information about the strength of correlations among features, we form a graph with edge weights corresponding to pairwise covariances. This graph is used to define a graph total variation regularizer that promotes similar weights for correlated features. This work shows how the proposed graph-based regularization yields mean-squared error guarantees for a broad range of covariance graph structures. These guarantees are optimal for many specific covariance graphs, including block and lattice graphs. Our proposed approach outperforms other methods for highly-correlated design in a variety of experiments on synthetic data and real biochemistry data.

Li Yuan, Mark Benjamin, Raskutti Garvesh, Willett Rebecca, Song Hyebin, Neiman David