When using machine learning techniques in decision-making processes, the interpretability of models is important. In present paper, we adopted Shapley additive explanation (SHAP), which based on fair profit allocation among many stakeholders depending their contribution, for interpreting a gradient-boosting decision tree model hospital data. For better interpretability, propose two novel as follows: (1) new metric feature importance SHAP and (2) technique termed packing, packs multiple...

Background and Purpose— Several stroke prognostic scores have been developed to predict clinical outcomes after stroke. This study aimed develop validate novel data-driven predictive models for by referring previous in patients with acute ischemic a real-world setting. Methods— We used retrospective data of 4237 who were hospitalized single center Japan between January 2012 August 2017. first validated point-based (preadmission comorbidities, level consciousness, age, neurological deficit...

There are great expectations for artificial intelligence (AI) in medicine. We aimed to develop an AI prognostic model surgically resected non-small cell lung cancer (NSCLC). This study enrolled 1049 patients with pathological stage I-IIIA NSCLC at Kyushu University. set 17 clinicopathological factors and 30 preoperative 22 postoperative blood test results as explanatory variables. Disease-free survival (DFS), overall (OS), cancer-specific (CSS) were objective The eXtreme Gradient Boosting...

<sec> <title>BACKGROUND</title> Prolonged hospital stays can lead to inefficiencies in healthcare delivery and unnecessary consumption of medical resources. </sec> <title>OBJECTIVE</title> This study aimed identify key clinical variances associated with prolonged length stay (PLOS) pathways using a machine learning model trained on real-world data from the ePath system. <title>METHODS</title> We analyzed 480 patients lung cancer (mean age 68.3 ± 11.2 years, 51% men) who underwent...

This study aimed to determine whether body weight is associated with functional outcome after acute ischemic stroke. We measured the mass index (BMI) and assessed clinical outcomes in patients The BMI was categorized into underweight (< 18.5 kg/m2), normal (18.5-22.9 overweight (23.0-24.9 obesity (≥ 25.0 kg/m2). association between a poor (modified Rankin Scale [mRS] score: 3-6) evaluated. included 11,749 stroke (70.3 ± 12.2 years, 36.1% women). risk of 3-month higher for underweight, lower...

Ridge regression is one of the most popular shrinkage estimation methods for linear models. effectively estimates coefficients in presence high-dimensional regressors. Recently, a generalized ridge estimator was suggested that involved generalizing uniform to non-uniform shrinkage; this shown perform well sparse and In paper, we introduce our newly developed R package “g.ridge” (first version published on 7 December 2023) implements both estimator. The equipped with cross-validation...

Background Although machine learning models demonstrate significant potential in predicting postoperative delirium, the advantages of their implementation real-world settings remain unclear and require a comparison with conventional practical applications. Objective The objective this study was to validate temporal generalizability decision tree ensemble sparse linear regression for delirium after surgery compared that traditional logistic model. Methods health record data patients...

Abstract Introduction The reliability of data‐driven predictions in real‐world scenarios remains uncertain. This study aimed to develop and validate a machine‐learning‐based model for predicting clinical outcomes using data from an electronic pathway (ePath) system. Methods All available were collected patients with lung cancer who underwent video‐assisted thoracoscopic surgery at two independent hospitals utilizing the ePath primary outcome interest was prolonged air leak (PAL), defined as...

Machine learning (ML) techniques are widely employed across various domains to achieve accurate predictions. This study assessed the effectiveness of ML in predicting early mortality risk among patients with acute intracerebral hemorrhage (ICH) real-world settings.

Abstract Introduction Patients with stroke often experience pneumonia during the acute stage after onset. Oral care may be effective in reducing risk of stroke‐associated (SAP). We aimed to determine changes oral care, as well incidence SAP, patients intracerebral hemorrhage, following implementation a learning health system our hospital. Methods retrospectively analyzed data 1716 hemorrhage who were hospitalized at single center Japan between January 2012 and December 2018. Data stratified...

Diabetic retinopathy (DR) is the leading cause of visual impairment and blindness. Consequently, numerous deep learning models have been developed for early detection DR. Safety-critical applications employed in medical diagnosis must be robust to distribution shifts. Previous studies focused on model performance under shifts using natural image datasets such as ImageNet, CIFAR-10, SVHN. However, there a lack research specifically investigating datasets. To address this gap, we investigated...

Ridge regression is one of the most popular shrinkage estimation methods for linear models. effectively estimates coefficients in presence high-dimensional regressors. Recently, a generalized ridge estimator was suggested by generalizing uniform to non-uniform shrinkage, which shown perform well under sparse and In this paper, we introduce our newly developed R package “g.ridge” (the first version published on 2023-12-07 at https://cran.r-project.org/package=g.ridge) that implements both...

Delirium is common in the emergency department, and once it develops, there a risk of self-extubation drains tubes, so critical to predict delirium before occurs. Machine learning was used create two prediction models this study: one for predicting occurrence after delirium. Each model showed high discriminative performance, indicating possibility selecting high-risk cases. Visualization predictors using Shapley additive explanation (SHAP), machine interpretability method, that were...

The purpose of this study is investigation transmitting phased array antennas for a future wireless power transfer (WPT) demonstration satellite experiment. A sequential was used to construct an antenna improve the overall axial ratio antenna. We evaluate characteristics based on simulations three types arrays. conduct experiments using model determined in validate reliability simulations.

Delirium in hospitalized patients is a worldwide problem, causing burden on healthcare professionals and impacting patient prognosis. A machine learning interpretation method (ML method) presents the results of predictions promotes guided decisions. This study focuses visualizing predictors delirium using ML implementing analysis clinical practice. Retrospective data 55,389 single acute care center Japan between December 2017 February 2022 were collected. Patients categorized into three...

Background Although machine learning is a promising tool for making prognoses, the performance of in predicting outcomes after stroke remains to be examined. Objective This study aims examine how much data-driven models with improve predictive poststroke compared conventional prognostic scores and elucidate explanatory variables learning–based differ from items scores. Methods We used data 10,513 patients who were registered multicenter prospective registry Japan between 2007 2017. The poor...

The turnover of kindergarten teachers has drastically increased in the past 10 years. Reducing rates among preschool workers become an important issue worldwide. Parents have avoided enrolling children preschools due to insufficient care, which affects their ability work. Therefore, this study developed a diagnostic model understand workers' unwillingness continue working. A total 1002 full-time were divided into 2 groups. Predictors drawn from general questionnaires, including those for...

<sec> <title>BACKGROUND</title> Although machine learning is a promising tool for prognostication, the performance of in predicting outcomes after stroke remains to be examined. </sec> <title>OBJECTIVE</title> We aimed examine how much data-driven models with improve predictive post-stroke compared conventional prognostic scores and elucidate explanatory variables learning-based differ from items scores. <title>METHODS</title> used data 10513 patients registered multicenter prospective...

<sec> <title>BACKGROUND</title> Although machine learning models demonstrate significant potential in predicting postoperative delirium, the advantages of their implementation real-world settings remain unclear and require a comparison with conventional practical applications. </sec> <title>OBJECTIVE</title> The objective this study was to validate temporal generalizability decision tree ensemble sparse linear regression for delirium after surgery compared that traditional logistic model....

Identifying important predicative indicators for prognosis is useful since these factors help understanding diseases and determining treatments patients. We extracted of cerebral infarction from EHR. analyzed EHR data 1,697 patients with 1,602 variables using gradient boosting decision tree. Extracted include not only well-known such as NIHSS but also new albumin-globulin ratio.

<sec> <title>BACKGROUND</title> The occurrence of delirium in hospitalized patients is a worldwide problem, not only because the burden it places on healthcare professionals, but also its impact patient prognosis. Therefore, although there has been much research use machine learning to predict advance, are few cases where results have applied clinical practice. Explainable artificial intelligence (XAI) techniques being increasingly adopted recent years these models present AI predictions...