Artificial Intelligence-driven Real-time Monitoring of Cardiovascular Conditions with Wearable Devices: A Scoping Review (Preprint) DOI
Ali Abedi, Anshul Verma, Dherya Jain

et al.

Published: March 18, 2025

BACKGROUND Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide, accounting for 18 million deaths annually. Early detection and prediction cardiovascular conditions are essential timely intervention improved patient outcomes. Wearable devices offer a promising, non-invasive solution continuous monitoring signals, vital signs, physical activity. However, large data volumes generated by these rapid fluctuations in signals necessitate advanced Artificial Intelligence (AI) techniques real-time analysis effective clinical decision-making. OBJECTIVE The objective this scoping review is to identify main challenges AI-driven platforms condition with wearable explore potential solutions. Additionally, aims examine how AI algorithms deployment pipelines optimized enable monitoring. METHODS A comprehensive search was conducted following electronic databases: MEDLINE(R) ALL (Ovid), Embase Cochrane Central Register Controlled Trials Web Science Core Collection (Clarivate), IEEE Xplore, ACM Digital Library, yielding 2,385 unique records. Inclusion criteria focused on studies that utilized participant collection applied detect or predict events diseases. After title abstract screening, 153 articles remained, full-text review, 19 met inclusion criteria. RESULTS findings indicate despite promise devices, research remains limited lacks validation. Most relied publicly available datasets rather than real-world validation recruited participants community settings. Studies deployed frequently failed report operational characteristics challenges. ECG-based sensors were most used primarily hospital variety techniques, ranging from traditional machine learning lightweight deep algorithms, either via cloud-based processing. CONCLUSIONS Robust, interdisciplinary needed harness full AI-driven, health management using devices. This includes development scalable solutions community-based deployment. Furthermore, such as compliance, hardware connectivity constraints, model optimization must be carefully addressed.

Language: Английский

The Smart deep learning based Model for Early Detection and Diagnosis of Melanoma DOI
Abdul Sajid Mohammed,

Anuteja Reddy Neravetla,

Sana Samreen

et al.

Published: Feb. 23, 2024

Melanoma, the deadliest form of skin cancer, is one most rapidly increasing cancer types. Early detection and diagnosis melanoma can significantly reduce impact disease on patient outcomes. However, traditional diagnostic approaches often only detect in later stages, which limits efficacy treatment. In this context, a smart deep learning-based model for early proposed. The consists convolutional neural network (CNN) visualization network. CNN extract morphological features from dermoscopy images moles, while provides pixel-level accuracy sensitivity potential malignancy. trained validated clinical datasets. has achieved superior results than existing methods, providing viable solution melanoma. showed great aiding dermatologists melanoma, leading to better outcomes reduced mortality rates. use learning technology allowed efficient processing large amounts data, fast accurate diagnoses. Additionally, was able learn improve over time, making it valuable tool detecting diagnosing

Language: Английский

Citations

3
Why digital innovation may not reduce healthcare’s environmental footprint DOI
Gabrielle Samuel, Geoffrey M. Anderson, Federica Lucivero

et al.

BMJ, Journal Year: 2024, Volume and Issue: unknown, P. e078303 - e078303

Published: June 3, 2024

Digital innovations come with their own environmental cost and should not be seen as a simple fix for healthcare emissions, argue Gabrielle Samuel colleagues Healthcare is becoming increasingly digitalised through in information communication technologies well advances machine learning artificial intelligence (AI).1 Advocates enthuse that this digitalisation—including monitoring devices, streaming, data storage—will improve key aspects of delivery such safety, accessibility, quality care, effectiveness, efficiency.2 Others debate whether these promises can met because complex social, cultural, economic, political implementation challenges.3 More recently, digital innovation has been promoted means to reduce the harms associated delivery.4 systems contribute roughly 5% country's total greenhouse gas figure often being higher high income countries.5 Although digitalisation harms, could also implemented ways do lead reductions. Indeed, given paradoxical increase energy use introduction saving technologies—the so called rebound effect—digital may resource little change health outcomes. have potential decrease harm from several (box 1). First, are expected help emissions existing facilities by improving efficiency. In UK NHS predicted carbon savings realtime monitoring, including intelligence, better control buildings (eg, lights, heating, cooling) forecast allocation more effectively.6 Use predict electricity water consumption across various allowed hospital managers identify variation usage deal causes.7 Box 1 ### How might #### Improving operational efficiency infrastructureRETURN TO TEXT

Language: Английский

Citations

3
Exploring IoT Architectures in Healthcare: A Systematic Mapping Study DOI

Fatima Bendaouch,

Hayat Zaydi,

Brahim El Bhiri

et al.

Advances in Science, Technology & Innovation/Advances in science, technology & innovation, Journal Year: 2025, Volume and Issue: unknown, P. 149 - 158

Published: Jan. 1, 2025

Language: Английский

Citations

0
Design and Implementation of an IoT-Enabled ECG Sensor System for Real-Time Cardiac Monitoring DOI

B. Nagarjun Singh,

Puneet Sharma, Deepak Kumar Verma

et al.

Lecture notes in electrical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 153 - 164

Published: Jan. 1, 2025

Language: Английский

Citations

0
Artificial Intelligence-driven Real-time Monitoring of Cardiovascular Conditions with Wearable Devices: A Scoping Review (Preprint) DOI
Ali Abedi, Anshul Verma, Dherya Jain

et al.

Published: March 18, 2025

BACKGROUND Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide, accounting for 18 million deaths annually. Early detection and prediction cardiovascular conditions are essential timely intervention improved patient outcomes. Wearable devices offer a promising, non-invasive solution continuous monitoring signals, vital signs, physical activity. However, large data volumes generated by these rapid fluctuations in signals necessitate advanced Artificial Intelligence (AI) techniques real-time analysis effective clinical decision-making. OBJECTIVE The objective this scoping review is to identify main challenges AI-driven platforms condition with wearable explore potential solutions. Additionally, aims examine how AI algorithms deployment pipelines optimized enable monitoring. METHODS A comprehensive search was conducted following electronic databases: MEDLINE(R) ALL (Ovid), Embase Cochrane Central Register Controlled Trials Web Science Core Collection (Clarivate), IEEE Xplore, ACM Digital Library, yielding 2,385 unique records. Inclusion criteria focused on studies that utilized participant collection applied detect or predict events diseases. After title abstract screening, 153 articles remained, full-text review, 19 met inclusion criteria. RESULTS findings indicate despite promise devices, research remains limited lacks validation. Most relied publicly available datasets rather than real-world validation recruited participants community settings. Studies deployed frequently failed report operational characteristics challenges. ECG-based sensors were most used primarily hospital variety techniques, ranging from traditional machine learning lightweight deep algorithms, either via cloud-based processing. CONCLUSIONS Robust, interdisciplinary needed harness full AI-driven, health management using devices. This includes development scalable solutions community-based deployment. Furthermore, such as compliance, hardware connectivity constraints, model optimization must be carefully addressed.

Language: Английский

Citations

0