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

General General

Adsorption-Based Separation of Near-Azeotropic Mixtures-A Challenging Example for High-Throughput Development of Adsorbents.

In The journal of physical chemistry. B

Adsorption of gas mixtures is central to adsorption-based gas separations, and the number of adsorbate mixture/adsorbent systems that exist is staggering. Because examples of machine learning (ML) models predicting single-component adsorption of arbitrary molecules in large libraries of crystalline adsorbents have been developed, it is interesting to determine whether these models can accurately predict mixture adsorption. Here, we use molecular simulations to generate mixture adsorption data with a set of 12 near-azeotropic molecules in a diverse set of MOFs. These data provide a challenging example for any method to rapidly predict mixture adsorption in MOFs. We combine a previous ML single-component isotherm model with ideal adsorbed solution theory (IAST) to make predictions that can be compared directly with molecular simulation data for these adsorbed mixtures. This combination of ML and IAST illustrates the scope that is available with these methods, but the accuracy of the resulting predictions is disappointing. By examining the same examples with IAST based on minimal molecular simulation data for single-component isotherms, we show that having an accurate description of adsorption in the dilute loading limit is critical to being able to accurately predict mixture adsorption. This observation points to a useful direction for future work developing robust ML models of adsorption isotherms for diverse collections of molecules and adsorbents.

Tang Dai, Gharagheizi Farhad, Sholl David S


Public Health Public Health

High-Resolution Spatiotemporal Modeling for Ambient PM2.5 Exposure Assessment in China from 2013 to 2019.

In Environmental science & technology ; h5-index 132.0

Exposure to fine particulate matter (PM2.5) has become a major global health concern. Although modeling exposure to PM2.5 has been examined in China, accurate long-term assessment of PM2.5 exposure with high spatiotemporal resolution at the national scale is still challenging. We aimed to establish a hybrid spatiotemporal modeling framework for PM2.5 in China that incorporated extensive predictor variables (satellite, chemical transport model, geographic, and meteorological data) and advanced machine learning methods to support long-term and short-term health studies. The modeling framework included three stages: (1) filling satellite aerosol optical depth (AOD) missing values; (2) modeling 1 km × 1 km daily PM2.5 concentrations at a national scale using extensive covariates; and (3) downscaling daily PM2.5 predictions to 100-m resolution at a city scale. We achieved good model performances with spatial cross-validation (CV) R2 of 0.92 and temporal CV R2 of 0.85 at the air quality sites across the country. We then estimated daily PM2.5 concentrations in China from 2013 to 2019 at 1 km × 1 km grid cells. The downscaled predictions at 100 m resolution greatly improved the spatial variation of PM2.5 concentrations at the city scale. The framework and data set generated in this study could be useful to PM2.5 exposure assessment and epidemiological studies.

Huang Conghong, Hu Jianlin, Xue Tao, Xu Hao, Wang Meng


Pathology Pathology

Artificial intelligence in upper GI endoscopy - current status, challenges and future promise.

In Journal of gastroenterology and hepatology ; h5-index 51.0

White-light endoscopy with biopsy is the current gold standard modality for detecting and diagnosing upper gastrointestinal (GI) pathology. However, missed lesions remain a challenge. To overcome interobserver variability and learning curve issues, artificial intelligence (AI) has recently been introduced to assist endoscopists in the detection and diagnosis of upper GI neoplasia. In contrast to AI in colonoscopy, current AI studies for upper GI endoscopy are smaller pilot studies. Researchers currently lack large volume, well-annotated, high-quality datasets in gastric cancer, dysplasia in Barrett's esophagus and early esophageal squamous cell cancer. This review will look at the latest studies of AI in upper GI endoscopy, discuss some of the challenges facing researchers, and predict what the future may hold in this rapidly changing field.

Yu Honggang, Singh Rajvinder, Shin Seon Ho, Ho Khek Yu


Endoscopy, Esophagus, Stomach, Upper GI

General General

Artificial intelligence in pancreaticobiliary endoscopy.

In Journal of gastroenterology and hepatology ; h5-index 51.0

Artificial intelligence (AI) applications in health care have exponentially increased in recent years, and a few of these are related to pancreatobiliary disorders. AI-based methods were applied to extract information, in prognostication, to guide clinical treatment decisions and in pancreatobiliary endoscopy to characterize lesions. AI applications in endoscopy are expected to reduce inter-operator variability, improve the accuracy of diagnosis, and assist in therapeutic decision-making in real time. AI-based literature must however be interpreted with caution given the limited external validation. A multidisciplinary approach combining clinical and imaging or endoscopy data will better utilize AI-based technologies to further improve patient care.

Akshintala Venkata S, Khashab Mouen A


Biliary, Endoscopy, Neoplasms < gastroenterology, Pancreas < gastroenterology, Pancreatobiliary (ERCP) < gastroenterology

General General

Artificial intelligence in small bowel capsule endoscopy - current status, challenges and future promise.

In Journal of gastroenterology and hepatology ; h5-index 51.0

Neural network-based solutions are under development to alleviate physicians from the tedious task of small-bowel capsule endoscopy reviewing. Computer-assisted detection is a critical step, aiming to reduce reading times while maintaining accuracy. Weakly supervised solutions have shown promising results; however, video-level evaluations are scarce, and no prospective studies have been conducted yet. Automated characterization (in terms of diagnosis and pertinence) by supervised machine learning solutions is the next step. It relies on large, thoroughly labeled databases, for which preliminary "ground truth" definitions by experts are of tremendous importance. Other developments are under ways, to assist physicians in localizing anatomical landmarks and findings in the small bowel, in measuring lesions, and in rating bowel cleanliness. It is still questioned whether artificial intelligence will enter the market with proprietary, built-in or plug-in software, or with a universal cloud-based service, and how it will be accepted by physicians and patients.

Dray Xavier, Iakovidis Dimitris, Houdeville Charles, Jover Rodrigo, Diamantis Dimitris, Histace Aymeric, Koulaouzidis Anastasios


algorithms, artificial intelligence, capsule endoscopy, deep learning, neural networks, small bowel

Cardiology Cardiology

A recurrent neural network using historical data to predict time series indoor PM2.5 concentrations for residential buildings.

In Indoor air

Due to the severe outdoor PM2.5 pollution in China, many people have installed air-cleaning systems in homes. To make the systems run automatically and intelligently, we developed a recurrent neural network (RNN) that uses historical data to predict the future indoor PM2.5 concentration. The RNN architecture includes an autoencoder and a recurrent part. We used data measured in an apartment over the course of an entire year to train and test the RNN. The data include indoor/outdoor PM2.5 concentration, environmental parameters and time of day. By comparing three different input strategies, we found that a strategy employing historical PM2.5 and time of day as inputs performed best. With this strategy, the model can be applied to predict the relatively stable trend of indoor PM2.5 concentration in advance. When the input length is 2 h and the prediction horizon is 30 min, the median prediction error is 8.3 µg/m3 for the whole test set. For times with indoor PM2.5 concentrations between (20,50] µg/m3 and (50,100] µg/m3 , the median prediction error is 8.3 and 9.2 µg/m3 , respectively. The low prediction error between the ground-truth and predicted values shows that the RNN can predict indoor PM2.5 concentrations with satisfactory performance.

Dai Xilei, Liu Junjie, Li Yongle


artificial intelligence, deep learning, indoor PM2.5, outdoor parameters, recurrent neural network, time series model