Rational and design of ST-segment elevation not associated with acute cardiac necrosis (LESTONNAC). A prospective registry for validation of a deep learning system assisted by artificial intelligence

Background and aims: The accurate differentiation between T1a and T1b Barrett’s cancer has both therapeutic and prognostic implications but is challenging even for experienced physicians. We trained an Artificial Intelligence (AI) system on the basis of deep artificial neural networks (deep learning) to differentiate between T1a and T1b Barrett’s cancer white-light images. Methods: Endoscopic images from three tertiary care centres in Germany were collected retrospectively. A deep learning system was trained and tested using the principles of cross-validation. A total of 230 white-light endoscopic images (108 T1a and 122 T1b) was evaluated with the AI-system. For comparison, the images were also classified by experts specialized in endoscopic diagnosis and treatment of Barrett’s cancer. Results: The sensitivity, specificity, F1 and accuracy of the AI-system in the differentiation between T1a and T1b cancer lesions was 0.77, 0.64, 0.73 and 0.71, respectively. There was no statistically significant difference between the performance of the AI-system and that of human experts with sensitivity, specificity, F1 and accuracy of 0.63, 0.78, 0.67 and 0.70 respectively. Conclusion: This pilot study demonstrates the first multicenter application of an AI-based system in the prediction of submucosal invasion in endoscopic images of Barrett’s cancer. AI scored equal to international experts in the field, but more work is necessary to improve the system and apply it to video sequences and in a real-life setting. Nevertheless, the correct prediction of submucosal invasion in Barret´s cancer remains challenging for both experts and AI.

Download Full-text

Project Management’s Triple Constraints Mapping on Augmented Reality based Learning System for Improved Quality

10.20944/preprints201912.0121.v1 ◽

2019 ◽

Author(s):

Mehreen Sirshar ◽

Syeda Hafsa Ali ◽

Haleema Sadia Baig

Keyword(s):

Artificial Intelligence ◽

Computer Vision ◽

Deep Learning ◽

Project Management ◽

Augmented Reality ◽

Learning System ◽

It Professionals ◽

Successful Completion ◽

Project Objectives ◽

Work Done

Over the last few decades there has been an exponential growth in IT, motivating IT professionals and scientists to explore new dimensions resulting in the advancement of artificial intelligence and its subcategories like computer vision, deep learning and augmented reality. AR is comparatively a new area which was initially explored for gaming but recently a lot of work has been done in education using AR. Most of this focuses on improving students understanding and motivation. Like any other project, the performance of an AR based project is determined by the customer satisfaction which is usually affected by the theory of triple constraints; cost, time and scope. many studies have shown that most of the projects are under development because they are unable to overcome these constraints and meet project objectives. We were unable to find any notable work done regarding project management for augmented reality systems and application. Therefore, in this paper, we propose a system for management of AR applications which mainly focuses on catering triple constraints to meet desired objectives. Each variable is further divided into subprocesses and by following these processes successful completion of the project can be achieved.

Download Full-text

Artificial intelligence in architecture: Generating conceptual design via deep learning

International Journal of Architectural Computing ◽

10.1177/1478077118800982 ◽

2018 ◽

Vol 16 (4) ◽

pp. 306-327 ◽

Cited By ~ 8

Author(s):

Imdat As ◽

Siddharth Pal ◽

Prithwish Basu

Keyword(s):

Artificial Intelligence ◽

Machine Learning ◽

Deep Learning ◽

Conceptual Design ◽

Dimensional Space ◽

Three Dimensional ◽

Field Research ◽

Building Blocks ◽

Learning System ◽

Generative Adversarial Networks

Artificial intelligence, and in particular machine learning, is a fast-emerging field. Research on artificial intelligence focuses mainly on image-, text- and voice-based applications, leading to breakthrough developments in self-driving cars, voice recognition algorithms and recommendation systems. In this article, we present the research of an alternative graph-based machine learning system that deals with three-dimensional space, which is more structured and combinatorial than images, text or voice. Specifically, we present a function-driven deep learning approach to generate conceptual design. We trained and used deep neural networks to evaluate existing designs encoded as graphs, extract significant building blocks as subgraphs and merge them into new compositions. Finally, we explored the application of generative adversarial networks to generate entirely new and unique designs.

Download Full-text

Understanding Human Impressions of Artificial Intelligence

10.31234/osf.io/5ursp ◽

2021 ◽

Author(s):

Kevin Robert McKee ◽

Xuechunzi Bai ◽

Susan Fiske

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Cognitive Processes ◽

Intelligent Systems ◽

Social Cognitive ◽

Learning System ◽

Human Interactions ◽

Behavioral Studies ◽

And Training ◽

Willingness To Cooperate

Artificial intelligence increasingly suffuses everyday life. However, people are frequently reluctant to interact with A.I. systems. This challenges both the deployment of beneficial A.I. technology and the development of deep learning systems that depend on humans for oversight, direction, and training. Previously neglected but fundamental, social-cognitive processes guide human interactions with A.I. systems. In five behavioral studies (N = 3,099), warmth and competence feature prominently in participants’ impressions of artificially intelligent systems. Judgments of warmth and competence systematically depend on human-A.I. interdependence. In particular, participants perceive systems that optimize interests aligned with human interests as warmer and systems that operate independently from human direction as more competent. Finally, a prisoner’s dilemma game shows that warmth and competence judgments predict participants’ willingness to cooperate with a deep learning system. These results demonstrate the generality of intent detection to interactions with technological actors. Researchers and developers should carefully consider the degree and alignment of interdependence between humans and new artificial intelligence systems.

Download Full-text

TCT CONNECT-18 Frontline Strategies in ST-Segment Elevation Myocardial Infarction Management: Creating Artificial Intelligence Algorithms for ECG Diagnosis

Journal of the American College of Cardiology ◽

10.1016/j.jacc.2020.09.049 ◽

2020 ◽

Vol 76 (17) ◽

pp. B8-B9

Author(s):

Sameer Mehta ◽

Francisco Fernandez ◽

Carlos Villagran ◽

Daniel Vieira ◽

Gladys Pinto ◽

...

Keyword(s):

Artificial Intelligence ◽

Myocardial Infarction ◽

St Segment Elevation ◽

Segment Elevation Myocardial Infarction ◽

St Segment

Download Full-text

Classification of Coffee and Wine with a Microwave Resonator and Deep Learning Machine Technique

10.26434/chemrxiv.12458252 ◽

2020 ◽

Author(s):

Hyo Bong Hong ◽

Jae-Chan Jeong ◽

Hans Joachim Krause

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

High Accuracy ◽

Complex Compounds ◽

Learning System ◽

Microwave Resonator ◽

Microwave Resonance ◽

Peak Assignment ◽

Learning Machine

In this study, coffee and wine were measured using an microwave resonator, and a deep learning system was trained using the acquired data, and then tested to see if the deep leaning system could distinguish these samples. We tested 6 kinds of wine, 6 kinds of cold brew coffee and 6 kinds of bottled coffee. The microwave resonance spectra of all samples were graphically displayed. The graphical images were processed by an artificial intelligence (AI) technique. By applying deep learning machine technique instead of the peak assignment for complex compounds in general, it was possible to facilitate the classification of coffee or wine with high accuracy.

Download Full-text

Detecting Digoxin Toxicity by Artificial Intelligence-Assisted Electrocardiography

International Journal of Environmental Research and Public Health ◽

10.3390/ijerph18073839 ◽

2021 ◽

Vol 18 (7) ◽

pp. 3839

Author(s):

Da-Wei Chang ◽

Chin-Sheng Lin ◽

Tien-Ping Tsao ◽

Chia-Cheng Lee ◽

Jiann-Torng Chen ◽

...

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Learning Algorithm ◽

Characteristic Curve ◽

Delayed Diagnosis ◽

Learning System ◽

Digoxin Toxicity ◽

Therapeutic Window ◽

Deep Learning Algorithm ◽

Stratified Analysis

Although digoxin is important in heart rate control, the utilization of digoxin is declining due to its narrow therapeutic window. Misdiagnosis or delayed diagnosis of digoxin toxicity is common due to the lack of awareness and the time-consuming laboratory work that is involved. Electrocardiography (ECG) may be able to detect potential digoxin toxicity based on characteristic presentations. Our study attempted to develop a deep learning model to detect digoxin toxicity based on ECG manifestations. This study included 61 ECGs from patients with digoxin toxicity and 177,066 ECGs from patients in the emergency room from November 2011 to February 2019. The deep learning algorithm was trained using approximately 80% of ECGs. The other 20% of ECGs were used to validate the performance of the Artificial Intelligence (AI) system and to conduct a human-machine competition. Area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were used to evaluate the performance of ECG interpretation between humans and our deep learning system. The AUCs of our deep learning system for identifying digoxin toxicity were 0.912 and 0.929 in the validation cohort and the human-machine competition, respectively, which reached 84.6% of sensitivity and 94.6% of specificity. Interestingly, the deep learning system using only lead I (AUC = 0.960) was not worse than using complete 12 leads (0.912). Stratified analysis showed that our deep learning system was more applicable to patients with heart failure (HF) and without atrial fibrillation (AF) than those without HF and with AF. Our ECG-based deep learning system provides a high-accuracy, economical, rapid, and accessible way to detect digoxin toxicity, which can be applied as a promising decision supportive system for diagnosing digoxin toxicity in clinical practice.

Download Full-text

Detection and Classification of Cardiac Arrhythmias by a Challenge-Best Deep Learning Neural Network Model

10.1101/766022 ◽

2019 ◽

Author(s):

Tsai-Min Chen ◽

Chih-Han Huang ◽

Edward S. C. Shih ◽

Yu-Feng Hu ◽

Ming-Jing Hwang

Keyword(s):

Neural Network ◽

Deep Learning ◽

Cardiac Arrhythmia ◽

Heart Diseases ◽

Clinical Applications ◽

Modeling Tools ◽

Bundle Branch Block ◽

St Segment Elevation ◽

St Segment ◽

Depth Analysis

AbstractBackgroundElectrocardiogram (ECG) is widely used to detect cardiac arrhythmia (CA) and heart diseases. The development of deep learning modeling tools and publicly available large ECG data in recent years has made accurate machine diagnosis of CA an attractive task to showcase the power of artificial intelligence (AI) in clinical applications.Methods and FindingsWe have developed a convolution neural network (CNN)-based model to detect and classify nine types of heart rhythms using a large 12-lead ECG dataset (6877 recordings) provided by the China Physiological Signal Challenge (CPSC) 2018. Our model achieved a median overall F1-score of 0.84 for the 9-type classification on CPSC2018’s hidden test set (2954 ECG recordings), which ranked first in this latest AI competition of ECG-based CA diagnosis challenge. Further analysis showed that concurrent CAs observed in the same patient were adequately predicted for the 476 patients diagnosed with multiple CA types in the dataset. Analysis also showed that the performances of using only single lead data were only slightly worse than using the full 12 lead data, with leads aVR and V1 being the most prominent. These results are extensively discussed in the context of their agreement with and relevance to clinical observations.ConclusionsAn AI model for automatic CA diagnosis achieving state-of-the-art accuracy was developed as the result of a community-based AI challenge advocating open-source research. In- depth analysis further reveals the model’s ability for concurrent CA diagnosis and potential use of certain single leads such as aVR in clinical applications.AbbreviationsCA, cardiac arrhythmia; AF, Atrial fibrillation; I-AVB, first-degree atrioventricular block; LBBB, left bundle branch block; RBBB, right bundle branch block; PAC, premature atrial contraction; PVC, premature ventricular contraction; STD, ST-segment depression; STE, ST-segment elevation.

Download Full-text

Primed Prediction: A Critical Examination of the Consequences of Exclusion of the Ontological Now in AI Protocol

10.16997/book55.k ◽

2021 ◽

pp. 183-201

Author(s):

Carrie O’Connell ◽

Chad Van de Wiele

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Theoretical Foundation ◽

Black Box ◽

Learning System ◽

Social Effects ◽

Critical Examination ◽

Technological Innovations ◽

Prediction And Simulation ◽

Material Plane

Revisiting Norbert Wiener’s cybernetic prediction as the theoretical foundation of AI this chapter makes a plea how we need to uncover the black box of what is behind prediction and simulation. It explores the shortcomings of cybernetic prediction, the theoretical foundation of Artificial Intelligence, through the lens of Jean Baudrillard’s simulacra and simulation. Specifically, what prediction excludes – namely, an accounting for the ontological now – is what Baudrillard warned against in his analysis of the role technological innovations play in untethering reality from the material plane, leading to a crisis of simulacrum of experience. From this perspective, any deep-learning system rooted in the Wiener’s view of cybernetic feedback loops risks creating behaviour more so than predicting it. As this chapter will argue, such prediction is a narrow, self-referential system of feedback that ultimately becomes a self-fulfilling prophecy girded by the psycho-social effects of the very chaos it seeks to rationalise.

Download Full-text

Clinical application of artificial intelligence-assisted diagnosis using anteroposterior pelvic radiographs in children with developmental dysplasia of the hip

The Bone & Joint Journal ◽

10.1302/0301-620x.102b11.bjj-2020-0712.r2 ◽

2020 ◽

Vol 102-B (11) ◽

pp. 1574-1581

Author(s):

Si-Cheng Zhang ◽

Jun Sun ◽

Chuan-Bin Liu ◽

Ji-Hong Fang ◽

Hong-Tao Xie ◽

...

Keyword(s):

Artificial Intelligence ◽

Deep Learning ◽

Receiver Operating Characteristic Curve ◽

Operating Characteristic ◽

Characteristic Curve ◽

Developmental Dysplasia ◽

Learning System ◽

Pelvic Radiographs ◽

Operating Characteristic Curve ◽

Dysplasia Of The Hip

Aims The diagnosis of developmental dysplasia of the hip (DDH) is challenging owing to extensive variation in paediatric pelvic anatomy. Artificial intelligence (AI) may represent an effective diagnostic tool for DDH. Here, we aimed to develop an anteroposterior pelvic radiograph deep learning system for diagnosing DDH in children and analyze the feasibility of its application. Methods In total, 10,219 anteroposterior pelvic radiographs were retrospectively collected from April 2014 to December 2018. Clinicians labelled each radiograph using a uniform standard method. Radiographs were grouped according to age and into ‘dislocation’ (dislocation and subluxation) and ‘non-dislocation’ (normal cases and those with dysplasia of the acetabulum) groups based on clinical diagnosis. The deep learning system was trained and optimized using 9,081 radiographs; 1,138 test radiographs were then used to compare the diagnoses made by deep learning system and clinicians. The accuracy of the deep learning system was determined using a receiver operating characteristic curve, and the consistency of acetabular index measurements was evaluated using Bland-Altman plots. Results In all, 1,138 patients (242 males; 896 females; mean age 1.5 years (SD 1.79; 0 to 10) were included in this study. The area under the receiver operating characteristic curve, sensitivity, and specificity of the deep learning system for diagnosing hip dislocation were 0.975, 276/289 (95.5%), and 1,978/1,987 (99.5%), respectively. Compared with clinical diagnoses, the Bland-Altman 95% limits of agreement for acetabular index, as determined by the deep learning system from the radiographs of non-dislocated and dislocated hips, were -3.27° - 2.94° and -7.36° - 5.36°, respectively (p < 0.001). Conclusion The deep learning system was highly consistent, more convenient, and more effective for diagnosing DDH compared with clinician-led diagnoses. Deep learning systems should be considered for analysis of anteroposterior pelvic radiographs when diagnosing DDH. The deep learning system will improve the current artificially complicated screening referral process. Cite this article: Bone Joint J 2020;102-B(11):1574–1581.

Download Full-text