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

General General

Evaluation of computer aided detection during colonoscopy in the community (AI-SEE): a multicenter randomized clinical trial.

In The American journal of gastroenterology

BACKGROUND : There has been increasing interest in artificial intelligence in gastroenterology. To reduce miss rate during colonoscopy, there has been significant exploration in computer aided detection (CADe) devices. In this study, we evaluate the use of CADe in colonoscopy in community-based, non-academic practices.

METHODS : Between September 28, 2020 and September 24, 2021, a randomized controlled trial (AI-SEE) was performed evaluating the impact of CADe on polyp detection in 4 community-based endoscopy centers in the USA. Patients were block randomized to undergoing colonoscopy with or without CADe (EndoVigilant). Primary outcomes measured were adenomas per colonoscopy (APC) and adenomas per extraction (APE; the percentage of polyps removed that are adenomas). Secondary endpoints included serrated polyps per colonoscopy, non-adenomatous, non-serrated polyps per colonoscopy, adenoma and serrated polyp detection rate, and procedural time.

RESULTS : A total of 769 patients were enrolled (387 with CADe), with similar patient demographics between the two groups. There was no significant difference in adenomas per colonoscopy in the CADe and non-CADe groups (0.73 vs 0.67, p=0.496). While use of CADe did not improve identification of serrated polyps per colonoscopy (0.08 vs 0.08, p=0.965), use of CADe increased identification of non-adenomatous, non-serrated polyps per colonoscopy (0.90 vs 0.51, p<0.0001), resulting in a lower APE in the CADe group. Adenoma detection rate (35.9 vs 37.2%, p=0.774) and serrated polyp detection rate (6.5 vs 6.3%, p=1.000) were similar in the CADe and non-CADe group. Mean withdrawal time was longer in the CADe compared to non-CADe group (11.7 vs 10.7 minutes, p=0.003). However, when no polyps were identified, there was similar mean withdrawal time (9.1 vs 8.8 minutes, p=0.288). There were no adverse events.

CONCLUSIONS : Use of CADe did not result in a statistically significant difference in the number of adenomas detected. Additional studies are needed to better understand why some endoscopists derive substantial benefits from CADe and others do not. ClinicalTrials.gov number, NCT04555135.

Wei Mike T, Shankar Uday, Parvin Russell, Hasan Abbas Syed, Chaudhary Sushant, Friedlander Yishai, Friedland Shai

2023-Mar-09

Surgery Surgery

Mechanism of Injury and Age Predict Operative Intervention in Pediatric Perineal Injury.

In Pediatric emergency care

OBJECTIVES : Literature characterizing pediatric perineal trauma is sparse and generally limited to females. The purpose of this study was to characterize pediatric perineal injuries with specific focus on patient demographics, mechanisms of injury, and care patterns at a regional level 1 pediatric trauma center.

METHODS : Retrospective review of children aged younger than 18 years evaluated at a level 1 pediatric trauma center from 2006 to 2017. Patients were identified by International Classification of Diseases-9 and 10 codes. Extracted data included demographics, injury mechanism, diagnostic studies, hospital course, and structures injured. The χ2 and t tests were used to examine differences between subgroups. Machine learning was used to predict variable importance in determining the need for operative interventions.

RESULTS : One hundred ninety-seven patients met inclusion criteria. Mean age was 8.5 years. A total of 50.8% were girls. Blunt trauma accounted for 83.8% of injuries. Motor vehicle collisions and foreign bodies were more common in patients aged 12 years and older, whereas falls and bicycle-related injuries were more common in those younger than 12 years (P < 0.01). Patients younger than 12 years were more likely to sustain blunt trauma with isolated external genital injuries (P < 0.01). Patients aged 12 and older had a higher incidence of pelvic fractures, bladder/urethral injuries, and colorectal injuries, suggesting more severe injury patterns (P < 0.01). Half of patients required operative intervention. Children aged 3 years or younger and older than 12 years had longer mean hospital stays compared with children aged 4 to 11 years (P < 0.01). Mechanism of injury and age constituted more than 75% of the variable importance in predicting operative intervention.

CONCLUSIONS : Perineal trauma in children varies by age, sex, and mechanism. Blunt mechanisms are the most common, with patients frequently requiring surgical intervention. Mechanism of injury and age may be important in deciding which patients will require operative intervention. This study describes injury patterns in pediatric perineal trauma that can be used to guide future practice and inform injury prevention efforts.

McLaughlin Christopher J, Martin Kathryn L

2023-Mar-07

General General

Machine learning quantitatively characterizes the deformation and destruction of explosive molecules.

In Physical chemistry chemical physics : PCCP

Although explosives have been widely used in mines, road development, old building demolishing, and munition explosions; currently, how chemical bonds between atoms break and recombine, how the molecular structure is deformed and destroyed, how the reaction product molecules are formed, and the details for this rapid change process in explosive reactions are not yet fully understood, which limits the full use of explosive energy and safer use of explosives. This paper presents a quantitative model of molecular structure deformation using machine learning algorithms as well as a qualitative model of its relationship with molecular structure destruction, based on a molecular dynamics simulation and detailed analysis of the shock-loaded ε-CL-20, providing new perspectives for explosive community research. Specifically, the quantitative model of molecular structure deformation establishes the quantitative relationship between the molecular volume change and molecular position change, and between molecular distance change and molecular volume change using the machine learning algorithms such as Delaunay triangulation, clustering, and gradient descent. We find that the molecular spacing in explosives is strongly compressed after being shocked, and the peripheral structure can shrink inward, which is beneficial to keep the cage structure stable. When the peripheral structure is compressed to a certain extent, the cage structure volume begins to expand and is then destroyed. In addition, hydrogen atom transfer occurs within the explosive molecule. This study amplifies the structural changes and the chemical reaction process for explosive molecules after being strongly compressed by a shock wave, which can enrich the knowledge of the real detonation reaction process. The analysis method based on quantitative characterization using machine learning proposed in this study can also be used to analyze the microscopic reaction mechanism in other materials.

Zhang Kaining, Chen Lang, Zhang Teng, Lu Jianying, Liu Danyang, Wu Junying

2023-Mar-09

Dermatology Dermatology

Attitudes Towards Artificial Intelligence Among Dermatologists Working in Saudi Arabia.

In Dermatology practical & conceptual

INTRODUCTION : Artificial intelligence (AI) and its applications are among the most discussed modern technologies today. Despite the rapidly expanding use of AI in medicine, and specifically in dermatology, only a few studies have studied the attitude of physicians toward AI.

OBJECTIVE : To recognize the attitudes towards AI among dermatologists in the Kingdom of Saudi Arabia.

METHODS : A cross-sectional survey was done among dermatologists in Saudi Arabia. Questionnaires were distributed through several online channels.

RESULTS : Overall, 103 dermatologists filled out the survey. The majority saw very strong or strong potential for AI in the automated detection of skin diseases based on dermatological clinical images (50.9%), dermoscopic images (66.6%) and within dermatopathology (66.6%). In regard to results of attitudes towards AI, 56.6% and 52. 8% agreed that AI will revolutionize medicine and dermatology, respectively. However, many of the respondents disagreed that AI will replace physicians (41.5%) and human dermatologists (39.6%) in the future. Age did not impact the overall attitude of dermatologists.

CONCLUSION : Dermatologists in Saudi Arabia showed an optimistic attitude towards AI in dermatology and medicine. However, dermatologists believe that AI will not replace humans in the future.

Al-Ali Fatima, Polesie Sam, Paoli John, Aljasser Mohammed, Salah Louai A

2023-Jan-01

Radiology Radiology

Preoperative prediction of tumor deposits in rectal cancer with clinical-magnetic resonance deep learning-based radiomic models.

In Frontiers in oncology

BACKGROUND : This study aimed to establish an effective model for preoperative prediction of tumor deposits (TDs) in patients with rectal cancer (RC).

METHODS : In 500 patients, radiomic features were extracted from magnetic resonance imaging (MRI) using modalities such as high-resolution T2-weighted (HRT2) imaging and diffusion-weighted imaging (DWI). Machine learning (ML)-based and deep learning (DL)-based radiomic models were developed and integrated with clinical characteristics for TD prediction. The performance of the models was assessed using the area under the curve (AUC) over five-fold cross-validation.

RESULTS : A total of 564 radiomic features that quantified the intensity, shape, orientation, and texture of the tumor were extracted for each patient. The HRT2-ML, DWI-ML, Merged-ML, HRT2-DL, DWI-DL, and Merged-DL models demonstrated AUCs of 0.62 ± 0.02, 0.64 ± 0.08, 0.69 ± 0.04, 0.57 ± 0.06, 0.68 ± 0.03, and 0.59 ± 0.04, respectively. The clinical-ML, clinical-HRT2-ML, clinical-DWI-ML, clinical-Merged-ML, clinical-DL, clinical-HRT2-DL, clinical-DWI-DL, and clinical-Merged-DL models demonstrated AUCs of 0.81 ± 0.06, 0.79 ± 0.02, 0.81 ± 0.02, 0.83 ± 0.01, 0.81 ± 0.04, 0.83 ± 0.04, 0.90 ± 0.04, and 0.83 ± 0.05, respectively. The clinical-DWI-DL model achieved the best predictive performance (accuracy 0.84 ± 0.05, sensitivity 0.94 ± 0. 13, specificity 0.79 ± 0.04).

CONCLUSIONS : A comprehensive model combining MRI radiomic features and clinical characteristics achieved promising performance in TD prediction for RC patients. This approach has the potential to assist clinicians in preoperative stage evaluation and personalized treatment of RC patients.

Fu Chunlong, Shao Tingting, Hou Min, Qu Jiali, Li Ping, Yang Zebin, Shan Kangfei, Wu Meikang, Li Weida, Wang Xuan, Zhang Jingfeng, Luo Fanghong, Zhou Long, Sun Jihong, Zhao Fenhua

2023

deep learning, diffusion-weighted imaging, magnetic resonance imaging, rectal cancer, tumor deposit

Radiology Radiology

COVID-19 imaging, where do we go from here? Bibliometric analysis of medical imaging in COVID-19.

In European radiology ; h5-index 62.0

OBJECTIVES : We conducted a systematic and comprehensive bibliometric analysis of COVID-19-related medical imaging to determine the current status and indicate possible future directions.

METHODS : This research provides an analysis of Web of Science Core Collection (WoSCC) indexed articles on COVID-19 and medical imaging published between 1 January 2020 and 30 June 2022, using the search terms "COVID-19" and medical imaging terms (such as "X-ray" or "CT"). Publications based solely on COVID-19 themes or medical image themes were excluded. CiteSpace was used to identify the predominant topics and generate a visual map of countries, institutions, authors, and keyword networks.

RESULTS : The search included 4444 publications. The journal with the most publications was European Radiology, and the most co-cited journal was Radiology. China was the most frequently cited country in terms of co-authorship, with the Huazhong University of Science and Technology being the institution contributing with the highest number of relevant co-authorships. Research trends and leading topics included: assessment of initial COVID-19-related clinical imaging features, differential diagnosis using artificial intelligence (AI) technology and model interpretability, diagnosis systems construction, COVID-19 vaccination, complications, and predicting prognosis.

CONCLUSIONS : This bibliometric analysis of COVID-19-related medical imaging helps clarify the current research situation and developmental trends. Subsequent trends in COVID-19 imaging are likely to shift from lung structure to function, from lung tissue to other related organs, and from COVID-19 to the impact of COVID-19 on the diagnosis and treatment of other diseases. Key Points • We conducted a systematic and comprehensive bibliometric analysis of COVID-19-related medical imaging from 1 January 2020 to 30 June 2022. • Research trends and leading topics included assessment of initial COVID-19-related clinical imaging features, differential diagnosis using AI technology and model interpretability, diagnosis systems construction, COVID-19 vaccination, complications, and predicting prognosis. • Future trends in COVID-19-related imaging are likely to involve a shift from lung structure to function, from lung tissue to other related organs, and from COVID-19 to the impact of COVID-19 on the diagnosis and treatment of other diseases.

Wen Ru, Zhang Mudan, Xu Rui, Gao Yingming, Liu Lin, Chen Hui, Wang Xingang, Zhu Wenyan, Lin Huafang, Liu Chen, Zeng Xianchun

2023-Mar-09

Bibliometrics, COVID-19, CiteSpace, Medical imaging, Treads