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

Ophthalmology Ophthalmology

Artificial Intelligence in Various Medical Fields With Emphasis on Radiology: Statistical Evaluation of the Literature.

In Cureus

Background Artificial intelligence (AI) has significantly impacted numerous medical specialties with high emphasis on radiology. Associated novel diagnostic methods have become a rapidly emerging hot topic, and it is essential to provide insights into quantitative analysis of the growing literature. Purpose The purpose of this study is to highlight future academic trends, identify potential research gaps, and analyze scientific landscape of AI in the field of medicine. The main aim is to explore comprehensive dataset over a 46-year period in terms of publication type, publication citation, country of origin, institution, and medical specialty. Material and Methods The Web of Science database was searched from 1975 to 2020, and publications on AI were explored. Both original research reports and review articles were included in comprehensive bibliometric analysis. Descriptive statistics were calculated, and numerous variables were applied, namely year of publication, institution, type of publication, specialty area, country of origin, and citation numbers, and the Kruskal-Wallis analysis of variance was used. Results A total of 117,974 relevant citations were retrieved, of which 83,979 original research and review articles were retained for analysis. Not surprisingly, the largest proportion of citations were from the United States (23%, n = 19,180) followed by China, Spain, England, and Germany. The number of citations was relatively consistent during the 1970s and emerging gradually during the 1980s. However, ongoing scientific trend positively evolved, and the numbers started to grow significantly in the 1990s and demonstrated continuous increasing wave since then. The most frequently represented key medical specialties were oncology, radiology, neuroradiology, and ophthalmology. Overall, no major statistical difference was found between these four domains (p = 0.753). Conclusions In summary, research on AI-powered technologies in the medical domain was at early stage in the 1970s. However, associated deep learning algorithms significantly attracted and revolutionized the scientific community with subsequent evolution of research and exponential growth of multidisciplinary publications since that time. Work in this field has impacted radiology as an area of predominant interest and has been led by institutions in the United States, Spain, France, China, and England. The bibliometric study reported herein can provide a broad overview and valuable guidance to help medical researchers gain insights into key points and trace the global trends regarding the status of AI research in medicine, particularly in radiology and other relevant multispecialty areas.

Pakdemirli Emre, Wegner Urszula


artificial intelligence, bibliometric analysis, machine learning, medicine, radiology

General General

Analysis of pedestrian activity before and during COVID-19 lockdown, using webcam time-lapse from Cracow and machine learning.

In PeerJ

At the turn of February and March 2020, COVID-19 pandemic reached Europe. Many countries, including Poland imposed lockdown as a method of securing social distance between potentially infected. Stay-at-home orders and movement control within public space not only affected the touristm industry, but also the everyday life of the inhabitants. The hourly time-lapse from four HD webcams in Cracow (Poland) are used in this study to estimate how pedestrian activity changed during COVID-19 lockdown. The collected data covers the period from 9 June 2016 to 19 April 2020 and comes from various urban zones. One zone is tourist, one is residential and two are mixed. In the first stage of the analysis, a state-of-the-art machine learning algorithm (YOLOv3) is used to detect people. Additionally, a non-standard application of the YOLO method is proposed, oriented to the images from HD webcams. This approach (YOLOtiled) is less prone to pedestrian detection errors with the only drawback being the longer computation time. Splitting the HD image into smaller tiles increases the number of detected pedestrians by over 50%. In the second stage, the analysis of pedestrian activity before and during the COVID-19 lockdown is conducted for hourly, daily and weekly averages. Depending on the type of urban zone, the number of pedestrians decreased from 33% in residential zones to 85% in tourist zones located in the Old Town. The presented method allows for more efficient detection and counting of pedestrians from HD time-lapse webcam images compared to SSD, YOLOv3 and Faster R-CNN. The result of the research is a published database with the detected number of pedestrians from the four-year observation period for four locations in Cracow.

Szczepanek Robert


COVID-19, Cracow, Data science, Database, OpenCV, Pedestrian counting, People detection, Webcam, YOLOv3

Radiology Radiology

Can Additional Patient Information Improve the Diagnostic Performance of Deep Learning for the Interpretation of Knee Osteoarthritis Severity.

In Journal of clinical medicine

The study compares the diagnostic performance of deep learning (DL) with that of the former radiologist reading of the Kellgren-Lawrence (KL) grade and evaluates whether additional patient data can improve the diagnostic performance of DL. From March 2003 to February 2017, 3000 patients with 4366 knee AP radiographs were randomly selected. DL was trained using knee images and clinical information in two stages. In the first stage, DL was trained only with images and then in the second stage, it was trained with image data and clinical information. In the test set of image data, the areas under the receiver operating characteristic curve (AUC)s of the DL algorithm in diagnosing KL 0 to KL 4 were 0.91 (95% confidence interval (CI), 0.88-0.95), 0.80 (95% CI, 0.76-0.84), 0.69 (95% CI, 0.64-0.73), 0.86 (95% CI, 0.83-0.89), and 0.96 (95% CI, 0.94-0.98), respectively. In the test set with image data and additional patient information, the AUCs of the DL algorithm in diagnosing KL 0 to KL 4 were 0.97 (95% confidence interval (CI), 0.71-0.74), 0.85 (95% CI, 0.80-0.86), 0.75 (95% CI, 0.66-0.73), 0.86 (95% CI, 0.79-0.85), and 0.95 (95% CI, 0.91-0.97), respectively. The diagnostic performance of image data with additional patient information showed a statistically significantly higher AUC than image data alone in diagnosing KL 0, 1, and 2 (p-values were 0.008, 0.020, and 0.027, respectively).The diagnostic performance of DL was comparable to that of the former radiologist reading of the knee osteoarthritis KL grade. Additional patient information improved DL diagnosis in interpreting early knee osteoarthritis.

Kim Dong Hyun, Lee Kyong Joon, Choi Dongjun, Lee Jae Ik, Choi Han Gyeol, Lee Yong Seuk


deep learning, diagnosing, knee, osteoarthritis, performance

Pathology Pathology

Transcriptional and proteomic insights into the host response in fatal COVID-19 cases.

In Proceedings of the National Academy of Sciences of the United States of America

Coronavirus disease 2019 (COVID-19), the global pandemic caused by SARS-CoV-2, has resulted thus far in greater than 933,000 deaths worldwide; yet disease pathogenesis remains unclear. Clinical and immunological features of patients with COVID-19 have highlighted a potential role for changes in immune activity in regulating disease severity. However, little is known about the responses in human lung tissue, the primary site of infection. Here we show that pathways related to neutrophil activation and pulmonary fibrosis are among the major up-regulated transcriptional signatures in lung tissue obtained from patients who died of COVID-19 in Wuhan, China. Strikingly, the viral burden was low in all samples, which suggests that the patient deaths may be related to the host response rather than an active fulminant infection. Examination of the colonic transcriptome of these patients suggested that SARS-CoV-2 impacted host responses even at a site with no obvious pathogenesis. Further proteomics analysis validated our transcriptome findings and identified several key proteins, such as the SARS-CoV-2 entry-associated protease cathepsins B and L and the inflammatory response modulator S100A8/A9, that are highly expressed in fatal cases, revealing potential drug targets for COVID-19.

Wu Meng, Chen Yaobing, Xia Han, Wang Changli, Tan Chin Yee, Cai Xunhui, Liu Yufeng, Ji Fenghu, Xiong Peng, Liu Ran, Guan Yuanlin, Duan Yaqi, Kuang Dong, Xu Sanpeng, Cai Hanghang, Xia Qin, Yang Dehua, Wang Ming-Wei, Chiu Isaac M, Cheng Chao, Ahern Philip P, Liu Liang, Wang Guoping, Surana Neeraj K, Xia Tian, Kasper Dennis L


COVID-19, NETosis, SARS-CoV-2, fibrosis, neutrophil

General General

Integrating multiple microarray dataset analysis and machine learning methods to reveal the key genes and regulatory mechanisms underlying human intervertebral disc degeneration.

In PeerJ

Intervertebral disc degeneration (IDD), a major cause of lower back pain, has multiple contributing factors including genetics, environment, age, and loading history. Bioinformatics analysis has been extensively used to identify diagnostic biomarkers and therapeutic targets for IDD diagnosis and treatment. However, multiple microarray dataset analysis and machine learning methods have not been integrated. In this study, we downloaded the mRNA, microRNA (miRNA), long noncoding RNA (lncRNA), and circular RNA (circRNA) expression profiles (GSE34095, GSE15227, GSE63492 GSE116726, GSE56081 and GSE67566) associated with IDD from the GEO database. Using differential expression analysis and recursive feature elimination, we extracted four optimal feature genes. We then used the support vector machine (SVM) to make a classification model with the four optimal feature genes. The ROC curve was used to evaluate the model's performance, and the expression profiles (GSE63492, GSE116726, GSE56081, and GSE67566) were used to construct a competitive endogenous RNA (ceRNA) regulatory network and explore the underlying mechanisms of the feature genes. We found that three miRNAs (hsa-miR-4728-5p, hsa-miR-5196-5p, and hsa-miR-185-5p) and three circRNAs (hsa_circRNA_100723, hsa_circRNA_104471, and hsa_circRNA_100750) were important regulators with more interactions than the other RNAs across the whole network. The expression level analysis of the three datasets revealed that BCAS4 and SCRG1 were key genes involved in IDD development. Ultimately, our study proposes a novel approach to determining reliable and effective targets in IDD diagnosis and treatment.

Chang Hongze, Yang Xiaolong, You Kemin, Jiang Mingwei, Cai Feng, Zhang Yan, Liu Liang, Liu Hui, Liu Xiaodong


Integrated analysis, Intervertebral disc degeneration, Key genes, Machine learning methods

General General

Enrichment of beneficial cucumber rhizosphere microbes mediated by organic acid secretion.

In Horticulture research

Resistant cultivars have played important roles in controlling Fusarium wilt disease, but the roles of rhizosphere interactions among different levels of resistant cultivars are still unknown. Here, two phenotypes of cucumber, one resistant and one with increased susceptibility to Fusarium oxysporum f.sp. cucumerinum (Foc), were grown in the soil and hydroponically, and then 16S rRNA gene sequencing and nontargeted metabolomics techniques were used to investigate rhizosphere microflora and root exudate profiles. Relatively high microbial community evenness for the Foc-susceptible cultivar was detected, and the relative abundances of Comamonadaceae and Xanthomonadaceae were higher for the Foc-susceptible cultivar than for the other cultivar. FishTaco analysis revealed that specific functional traits, such as protein synthesis and secretion, bacterial chemotaxis, and small organic acid metabolism pathways, were significantly upregulated in the rhizobacterial community of the Foc-susceptible cultivar. A machine-learning approach in conjunction with FishTaco plus metabolic pathway analysis revealed that four organic acids (citric acid, pyruvate acid, succinic acid, and fumarate) were released at higher abundance by the Foc-susceptible cultivar compared with the resistant cultivar, which may be responsible for the recruitment of Comamonadaceae, a potential beneficial microbial group. Further validation demonstrated that Comamonadaceae can be "cultured" by these organic acids. Together, compared with the resistant cultivar, the susceptible cucumber tends to assemble beneficial microbes by secreting more organic acids.

Wen Tao, Yuan Jun, He Xiaoming, Lin Yue, Huang Qiwei, Shen Qirong


Secondary metabolism, Soil microbiology