Artificial Intelligence and Management: The Automation–Augmentation Paradox

Academy of Management Review, Journal Year: 2021, Volume and Issue: 46(1), P. 192 - 210

Published: Jan. 1, 2021

Taking three recent business books on artificial intelligence (AI) as a starting point, we explore the automation and augmentation concepts in management domain. Whereas implies that machines take over human task, means humans collaborate closely with to perform task. normative stance, advise organizations prioritize augmentation, which they relate superior performance. Using more comprehensive paradox theory perspective, argue that, domain, cannot be neatly separated from automation. These dual AI applications are interdependent across time space, creating paradoxical tension. Overemphasizing either or fuels reinforcing cycles negative organizational societal outcomes. However, if adopt broader perspective comprising both could deal tension achieve complementarities benefit society. Drawing our insights, conclude scholars need involved research use of organizations. We also substantial change is required how currently conducted order develop meaningful provide practice sound advice.

Language: Английский

---

Deep learning-enabled medical computer vision

npj Digital Medicine, Journal Year: 2021, Volume and Issue: 4(1)

Published: Jan. 8, 2021

Abstract A decade of unprecedented progress in artificial intelligence (AI) has demonstrated the potential for many fields—including medicine—to benefit from insights that AI techniques can extract data. Here we survey recent development modern computer vision techniques—powered by deep learning—for medical applications, focusing on imaging, video, and clinical deployment. We start briefly summarizing a convolutional neural networks, including tasks they enable, context healthcare. Next, discuss several example imaging applications stand to benefit—including cardiology, pathology, dermatology, ophthalmology–and propose new avenues continued work. then expand into general highlighting ways which workflows integrate enhance care. Finally, challenges hurdles required real-world deployment these technologies.

Language: Английский

---

Liquid biopsy enters the clinic — implementation issues and future challenges

Nature Reviews Clinical Oncology, Journal Year: 2021, Volume and Issue: 18(5), P. 297 - 312

Published: Jan. 20, 2021

Language: Английский

---

The state of artificial intelligence-based FDA-approved medical devices and algorithms: an online database

npj Digital Medicine, Journal Year: 2020, Volume and Issue: 3(1)

Published: Sept. 11, 2020

At the beginning of artificial intelligence (AI)/machine learning (ML) era, expectations are high, and experts foresee that AI/ML shows potential for diagnosing, managing treating a wide variety medical conditions. However, obstacles implementation in daily clinical practice numerous, especially regarding regulation these technologies. Therefore, we provide an insight into currently available AI/ML-based devices algorithms have been approved by US Food & Drugs Administration (FDA). We aimed to raise awareness importance regulatory bodies, clearly stating whether device is based or not. Cross-checking validating all approvals, identified 64 based, FDA algorithms. Out those, only 29 (45%) mentioned any AI/ML-related expressions official announcement. The majority (85.9%) was with 510(k) clearance, while 8 (12.5%) received de novo pathway clearance one (1.6%) premarket approval (PMA) clearance. Most technologies, notably 30 (46.9%), 16 (25.0%), 10 (15.6%) were developed fields Radiology, Cardiology Internal Medicine/General Practice respectively. launched first comprehensive open access database strictly technologies FDA. will be constantly updated.

Language: Английский

---

Clinically Applicable AI System for Accurate Diagnosis, Quantitative Measurements, and Prognosis of COVID-19 Pneumonia Using Computed Tomography

Cell, Journal Year: 2020, Volume and Issue: 181(6), P. 1423 - 1433.e11

Published: May 4, 2020

Language: Английский

---

Artificial intelligence versus clinicians: systematic review of design, reporting standards, and claims of deep learning studies

BMJ, Journal Year: 2020, Volume and Issue: unknown, P. m689 - m689

Published: March 25, 2020

Abstract Objective To systematically examine the design, reporting standards, risk of bias, and claims studies comparing performance diagnostic deep learning algorithms for medical imaging with that expert clinicians. Design Systematic review. Data sources Medline, Embase, Cochrane Central Register Controlled Trials, World Health Organization trial registry from 2010 to June 2019. Eligibility criteria selecting Randomised registrations non-randomised a algorithm in contemporary group one or more Medical has seen growing interest research. The main distinguishing feature convolutional neural networks (CNNs) is when CNNs are fed raw data, they develop their own representations needed pattern recognition. learns itself features an image important classification rather than being told by humans which use. selected aimed use predicting absolute existing disease into groups (eg, non-disease). For example, chest radiographs tagged label such as pneumothorax no CNN pixel patterns suggest pneumothorax. Review methods Adherence standards was assessed using CONSORT (consolidated trials) randomised TRIPOD (transparent multivariable prediction model individual prognosis diagnosis) studies. Risk bias tool PROBAST (prediction assessment tool) Results Only 10 records were found clinical trials, two have been published (with low except lack blinding, high adherence standards) eight ongoing. Of 81 trials identified, only nine prospective just six tested real world setting. median number experts comparator four (interquartile range 2-9). Full access all datasets code severely limited (unavailable 95% 93% studies, respectively). overall 58 suboptimal (<50% 12 29 items). 61 stated abstract artificial intelligence at least comparable (or better than) 31 (38%) further required. Conclusions Few exist imaging. Most not prospective, deviate standards. availability lacking most human often small. Future should diminish enhance relevance, improve transparency, appropriately temper conclusions. Study registration PROSPERO CRD42019123605.

Language: Английский

---

Targeting Strategies for Tissue-Specific Drug Delivery

Cell, Journal Year: 2020, Volume and Issue: 181(1), P. 151 - 167

Published: April 1, 2020

Language: Английский

---

Vision, challenges, roles and research issues of Artificial Intelligence in Education

Computers and Education Artificial Intelligence, Journal Year: 2020, Volume and Issue: 1, P. 100001 - 100001

Published: Jan. 1, 2020

The rapid advancement of computing technologies has facilitated the implementation AIED (Artificial Intelligence in Education) applications. refers to use AI Intelligence) or application programs educational settings facilitate teaching, learning, decision making. With help technologies, which simulate human intelligence make inferences, judgments, predictions, computer systems can provide personalized guidance, supports, feedback students as well assisting teachers policymakers making decisions. Although been identified primary research focus field computers and education, interdisciplinary nature presents a unique challenge for researchers with different disciplinary backgrounds. In this paper, we present definition roles studies from perspective needs. We propose framework show considerations implementing learning teaching settings. structure guide both education backgrounds conducting studies. outline 10 potential topics that are particular interest journal. Finally, describe type articles like solicit management submissions.

Language: Английский

---

The false hope of current approaches to explainable artificial intelligence in health care

The Lancet Digital Health, Journal Year: 2021, Volume and Issue: 3(11), P. e745 - e750

Published: Oct. 25, 2021

The black-box nature of current artificial intelligence (AI) has caused some to question whether AI must be explainable used in high-stakes scenarios such as medicine. It been argued that will engender trust with the health-care workforce, provide transparency into decision making process, and potentially mitigate various kinds bias. In this Viewpoint, we argue argument represents a false hope for explainability methods are unlikely achieve these goals patient-level support. We an overview techniques highlight how failure cases can cause problems individual patients. absence suitable methods, advocate rigorous internal external validation models more direct means achieving often associated explainability, caution against having requirement clinically deployed models.

Language: Английский

---

Artificial intelligence in healthcare: transforming the practice of medicine

Future Healthcare Journal, Journal Year: 2021, Volume and Issue: 8(2), P. e188 - e194

Published: July 1, 2021

Artificial intelligence (AI) is a powerful and disruptive area of computer science, with the potential to fundamentally transform practice medicine delivery healthcare. In this review article, we outline recent breakthroughs in application AI healthcare, describe roadmap building effective, reliable safe systems, discuss possible future direction augmented healthcare systems.

Language: Английский

---