The placenta is one of the most vital organs when a woman is pregnant. If it's not working correctly, the consequences can be dire: Children may experience stunted growth and neurological disorders, and their mothers are at increased risk of blood conditions like preeclampsia, which can impair kidney and liver function. Unfortunately, assessing placental health is difficult because of the limited information that can be gleaned from imaging. Traditional ultrasounds are cheap, portable, and easy to perform, but they can't always capture enough detail. This has spurred researchers to explore the potential of magnetic resonance imaging (MRI).
Artificial intelligence based on medical claims data outperforms traditional models in stratifying patient risk. ABSTRACT Objectives: Current models for patient risk prediction rely on practitioner expertise and domain knowledge. This study presents a deep learning model--a type of machine learning that does not require human inputs--to analyze complex clinical and financial data for population risk stratification. Methods: "Skip-Gram," an unsupervised deep learning approach that uses neural networks for prediction modeling, used data from 2014 and 2015 to predict the risk of hospitalization in 2016. The area under the curve (AUC) of the deep learning model was compared with that of both the Clinical Classifications Software and the commercial DxCG Intelligence predictive risk models, each with and without demographic and utilization features.
Artificial intelligence could be used to help catch paedophiles operating on the dark web. The technology would target the most dangerous and sophisticated offenders in efforts to tackle child sexual abuse, the Home Office said. Earlier this month Chancellor Sajid Javid announced £30 million would be set aside to tackle online child sexual exploitation. The Government has pledged to spend more money on the Child Abuse Image Database (CAID), which since 2014 has allowed police and other law enforcement agencies to search seized computers and other devices for indecent images of children quickly against a record of 14 million images to help identify victims. The investment will be used to consider whether adding aspects of artificial intelligence (AI) to the system to analyse voices and estimate ages would help in tracking down child abusers.
Burnout has become a popular buzzword in today's business world, meant to describe prolonged periods of stress in the workplace leading to feelings of depression and dissatisfaction with one's occupation. The topic has become so pervasive that the World Health Organization (WHO) addressed it at its 2019 World Health Assembly in Geneva in May, adding burnout to the 11th revision of the International Classification of Diseases (ICD-11) -- although classifying it as an "occupational phenomenon" rather than a medical condition. Healthcare itself is not immune to burnout, and a recent study in Journal of the American College of Radiology demonstrates it is taking a toll on pediatric radiologists in particular. The study surveyed Society of Pediatric Radiology (SPR) members and found nearly two-thirds expressed at least one symptom of burnout. While burnout is a complicated phenomenon and no two people experience it the same way, a commentary on the study suggests artificial intelligence (AI) could help alleviate some of the difficulties that can lead to burnout.
The Radiological Society of North America (RSNA) is organizing a challenge intended to show the application of machine learning and artificial intelligence on medical imaging and the ways in which these emerging tools and methodologies may improve diagnostic care. The RSNA Pediatric Bone Age Machine Learning Challenge addresses a familiar image analysis activity for pediatric radiologists: assessment of bone age from hand radiographs of pediatric patients used to evaluate growth and diagnose developmental disorders. The Challenge uses a dataset of hand radiographs provided by a consortium of leading research institutions -- Stanford University, the University of California, Los Angeles and the University of Colorado -- that have associated bone age assessments provided by multiple expert observers. Participants in the challenge will be judged by how well the bone age evaluations produced by their algorithms accord with the expert observers' evaluations. Participants will have the opportunity to directly compare their algorithms in a structured way using this carefully curated dataset.
Babies can cry because they're sick or in pain, but sometimes -- okay, a lot of the time -- they cry because they're hungry, cranky, or just feel like stretching their developing vocal cords. Now, researchers from Northern Illinois University have found a way to use artificial intelligence to decipher between the two types of vocalizations -- and not only could the AI save new parents from a lot of unnecessary worry, it could also save sick babies' lives. In a study published in IEEE/CAA Journal of Automatica Sinica, the researchers detail how they designed an algorithm based on automatic speech recognition technology that could distinguish between "normal" and "abnormal" cries, the latter being the kind often caused by a medical problem. They gathered their training data for the algorithm from 26 infants in a hospital's neonatal intensive care unit. Then they asked experienced nurses and caregivers to parse the probable reason for each cry -- "hungry," "diaper," "attention," "sleepy," or "discomfort" -- with the first four reasons considered "normal" and the fifth "abnormal."
Background: While machine learning (ML) models are rapidly emerging as promising screening tools in critical care medicine, the identification of homogeneous subphenotypes within populations with heterogeneous conditions such as pediatric sepsis may facilitate attainment of high-predictive performance of these prognostic algorithms. This study is aimed to identify subphenotypes of pediatric sepsis and demonstrate the potential value of partitioned data/subtyping-based training. Methods: This was a retrospective study of clinical data extracted from medical records of 6,446 pediatric patients that were admitted at a major hospital system in the DC area. Vitals and labs associated with patients meeting the diagnostic criteria for sepsis were used to perform latent profile analysis. Modern ML algorithms were used to explore the predictive performance benefits of reduced training data heterogeneity via label profiling. Results: In total 134 (2.1%) patients met the diagnostic criteria for sepsis in this cohort and latent profile analysis identified four profiles/subphenotypes of pediatric sepsis. Profiles 1 and 3 had the lowest mortality and included pediatric patients from different age groups. Profile 2 were characterized by respiratory dysfunction; profile 4 by neurological dysfunction and highest mortality rate (22.2%). Machine learning experiments comparing the predictive performance of models derived without training data profiling against profile targeted models suggest statistically significant improved performance of prediction can be obtained. For example, area under ROC curve (AUC) obtained to predict profile 4 with 24-hour data (AUC = .998, p < .0001) compared favorably with the AUC obtained from the model considering all profiles as a single homogeneous group (AUC = .918) with 24-hour data.
Pediatric asthma is the most prevalent chronic childhood illness, afflicting about 6.2 million children in the United States. However, asthma could be better managed by identifying and avoiding triggers, educating about medications and proper disease management strategies. This research utilizes deep learning methodologies to predict asthma-related emergency department (ED) visit within 3 months using Medicaid claims data. We compare prediction results against traditional statistical classification model - penalized Lasso logistic regression, which we trained and have deployed since 2015. The results have indicated that deep learning model Artificial Neural Networks (ANN) slightly outperforms (with AUC = 0.845) the Lasso logistic regression (with AUC = 0.842). The reason may come from the nonlinear nature of ANN.
The Infant Mortality Rate (IMR) is the number of infants per 1000 that do not survive until their first birthday. It is an important metric providing information about infant health but it also measures the society's general health status. Despite the high level of prosperity in the U.S.A., the country's IMR is higher than that of many other developed countries. Additionally, the U.S.A. exhibits persistent inequalities in the IMR across different racial and ethnic groups. In this paper, we study the infant mortality prediction using features extracted from birth certificates. We are interested in training classification models to decide whether an infant will survive or not. We focus on exploring and understanding the importance of features in subsets of the population; we compare models trained for individual races to general models. Our evaluation shows that our methodology outperforms standard classification methods used by epidemiology researchers.