Survey of Explainable AI Techniques in Healthcare

Sensors, Journal Year: 2023, Volume and Issue: 23(2), P. 634 - 634

Published: Jan. 5, 2023

Artificial intelligence (AI) with deep learning models has been widely applied in numerous domains, including medical imaging and healthcare tasks. In the field, any judgment or decision is fraught risk. A doctor will carefully judge whether a patient sick before forming reasonable explanation based on patient's symptoms and/or an examination. Therefore, to be viable accepted tool, AI needs mimic human interpretation skills. Specifically, explainable (XAI) aims explain information behind black-box model of that reveals how decisions are made. This paper provides survey most recent XAI techniques used related applications. We summarize categorize types, highlight algorithms increase interpretability topics. addition, we focus challenging problems applications provide guidelines develop better interpretations using concepts image text analysis. Furthermore, this future directions guide developers researchers for prospective investigations clinical topics, particularly imaging.

Language: Английский

---

Electroencephalography Signal Processing: A Comprehensive Review and Analysis of Methods and Techniques

Sensors, Journal Year: 2023, Volume and Issue: 23(14), P. 6434 - 6434

Published: July 16, 2023

The electroencephalography (EEG) signal is a noninvasive and complex that has numerous applications in biomedical fields, including sleep the brain–computer interface. Given its complexity, researchers have proposed several advanced preprocessing feature extraction methods to analyze EEG signals. In this study, we comprehensive review of articles related processing. We searched major scientific engineering databases summarized results our findings. Our survey encompassed entire process processing, from acquisition pretreatment (denoising) extraction, classification, application. present detailed discussion comparison various techniques used for Additionally, identify current limitations these their future development trends. conclude by offering some suggestions research field

Language: Английский

---

A Tutorial on Federated Learning from Theory to Practice: Foundations, Software Frameworks, Exemplary Use Cases, and Selected Trends

IEEE/CAA Journal of Automatica Sinica, Journal Year: 2024, Volume and Issue: 11(4), P. 824 - 850

Published: March 20, 2024

When data privacy is imposed as a necessity, Federated learning (FL) emerges relevant artificial intelligence field for developing machine (ML) models in distributed and decentralized environment. FL allows ML to be trained on local devices without any need centralized transfer, thereby reducing both the exposure of sensitive possibility interception by malicious third parties. This paradigm has gained momentum last few years, spurred plethora real-world applications that have leveraged its ability improve efficiency accommodate numerous participants with their sources. By virtue FL, can learned from all such sources while preserving privacy. The aim this paper provide practical tutorial including short methodology systematic analysis existing software frameworks. Furthermore, our provides exemplary cases study three complementary perspectives: i) Foundations describing main components key elements categories; ii) Implementation guidelines study, systematically examining functionalities provided frameworks deployment, devising design scenario, providing source code different approaches; iii) Trends, shortly reviewing non-exhaustive list research directions are under active investigation current landscape. ultimate purpose work establish itself referential researchers, developers, scientists willing explore capabilities applications.

Language: Английский

---

Advancements in MRI-Based Radiomics and Artificial Intelligence for Prostate Cancer: A Comprehensive Review and Future Prospects

Cancers, Journal Year: 2023, Volume and Issue: 15(15), P. 3839 - 3839

Published: July 28, 2023

The use of multiparametric magnetic resonance imaging (mpMRI) has become a common technique used in guiding biopsy and developing treatment plans for prostate lesions. While this is effective, non-invasive methods such as radiomics have gained popularity extracting features to develop predictive models clinical tasks. aim minimize invasive processes improved management cancer (PCa). This study reviews recent research progress MRI-based PCa, including the pipeline potential factors affecting personalized diagnosis. integration artificial intelligence (AI) with medical also discussed, line development trend radiogenomics multi-omics. survey highlights need more data from multiple institutions avoid bias generalize model. AI-based model considered promising tool good prospects application.

Language: Английский

---

Federated Learning for Healthcare Applications

IEEE Internet of Things Journal, Journal Year: 2023, Volume and Issue: 11(5), P. 7339 - 7358

Published: Oct. 19, 2023

Due to the fast advancement of artificial intelligence (AI), centralized-based models have become critical for healthcare tasks like in medical image analysis and human behavior recognition. Although these exhibit suitable performance, they are frequently constrained by privacy concerns. To attenuate this, a centralized learning strategy cannot be used cases where there is risk data breach, particularly centers. Federated (FL) technique that allows training global model without sharing distributed local aggregating them. By implementing FL throughout process, we can obtain with comparable generalization abilities while maintaining privacy. This survey provides an introduction fundamental concepts categories FL, highlights limitations model, discusses how address constraints. We also provide detailed overview applications using models, along commonly evaluation metrics public sets. In this context, implemented case study demonstrate applied field. Furthermore, outline key challenges future trends FL.

Language: Английский

---

Vertical federated learning based on data subset representation for healthcare application

Computer Methods and Programs in Biomedicine, Journal Year: 2025, Volume and Issue: 263, P. 108623 - 108623

Published: Feb. 12, 2025

Language: Английский

---

Validation of a Deep Learning Chest X-ray Interpretation Model: Integrating Large-Scale AI and Large Language Models for Comparative Analysis with ChatGPT

Diagnostics, Journal Year: 2023, Volume and Issue: 14(1), P. 90 - 90

Published: Dec. 30, 2023

This study evaluates the diagnostic accuracy and clinical utility of two artificial intelligence (AI) techniques: Kakao Brain Artificial Neural Network for Chest X-ray Reading (KARA-CXR), an assistive technology developed using large-scale AI large language models (LLMs), ChatGPT, a well-known LLM. The was conducted to validate performance technologies in chest reading explore their potential applications medical imaging diagnosis domain. methodology consisted randomly selecting 2000 images from single institution's patient database, radiologists evaluated readings provided by KARA-CXR ChatGPT. used five qualitative factors evaluate generated each model: accuracy, false findings, location inaccuracies, count hallucinations. Statistical analysis showed that achieved significantly higher compared In 'Acceptable' category, rated at 70.50% 68.00% observers, while ChatGPT 40.50% 47.00%. Interobserver agreement moderate both systems, with KARA 0.74 GPT4 0.73. For 'False Findings', scored 68.50%, 37.00% high interobserver agreements 0.96 0.97 GPT4. 'Location Inaccuracy' 'Hallucinations', outperformed significant margins. demonstrated non-hallucination rate 75%, which is than ChatGPT's 38%. (0.91) (0.85) hallucination category. conclusion, this demonstrates diagnostics. It also shows domain, has relatively

Language: Английский

---

The Power of AI-Assisted Diagnosis

ICST Transactions on e-Education and e-Learning, Journal Year: 2023, Volume and Issue: 8(4), P. e3 - e3

Published: Sept. 6, 2023

The rapid advancements in artificial intelligence (AI) have unleashed a wave of transformative technologies, and one area that has witnessed significant progress is AI-assisted diagnosis healthcare. With the ability to analyze vast amounts medical data, learn from patterns, make accurate predictions, AI systems hold immense potential revolutionize diagnostic process, enabling earlier detection, improved accuracy, personalized treatment recommendations. This review aims explore impact healthcare, specifically focusing on its role assisting physicians with diagnosis, highlighting benefits, challenges, ethical considerations associated integration into clinical practice. Through utilization AI's capabilities, enhancement patient outcomes, optimization resource allocation, reshaping professionals' approaches can be achieved.

Language: Английский

---

Federated Learning of XAI Models in Healthcare: A Case Study on Parkinson’s Disease

Cognitive Computation, Journal Year: 2024, Volume and Issue: 16(6), P. 3051 - 3076

Published: Aug. 28, 2024

Abstract Artificial intelligence (AI) systems are increasingly used in healthcare applications, although some challenges have not been completely overcome to make them fully trustworthy and compliant with modern regulations societal needs. First of all, sensitive health data, essential train AI systems, typically stored managed several separate medical centers cannot be shared due privacy constraints, thus hindering the use all available information learning models. Further, transparency explainability such becoming urgent, especially at a time when “opaque” or “black-box” models commonly used. Recently, technological algorithmic solutions these investigated: on one hand, federated (FL) has proposed as paradigm for collaborative model training among multiple parties without any disclosure private raw data; other research eXplainable (XAI) aims enhance either through interpretable by-design approaches post-hoc explanation techniques. In this paper, we focus case study, namely predicting progression Parkinson’s disease, assume that data originate from different collection centralized is precluded limitations. We aim investigate how FL XAI can allow achieving good level accuracy trustworthiness. Cognitive biologically inspired adopted our analysis: an fuzzy rule-based system neural network explained using version SHAP technique. analyze accuracy, interpretability, two approaches, also varying degree heterogeneity across distribution scenarios. Although generally more accurate, results show achieves competitive performance setting presents desirable properties terms interpretability transparency.

Language: Английский

---

Generalizable and Explainable Deep Learning for Medical Image Computing: An Overview

Current Opinion in Biomedical Engineering, Journal Year: 2024, Volume and Issue: 33, P. 100567 - 100567

Published: Nov. 14, 2024

Language: Английский

---