Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Machine Learning and Transformers for Thyroid Carcinoma Diagnosis: A Review

Yassine Habchi, Hamza Kheddar, Yassine Himeur, Mohamed Chahine Ghanem

TL;DR

This review addresses the problem of thyroid carcinoma diagnosis using machine learning and Transformer-based approaches, synthesizing supervised, unsupervised, and deep learning methods with a focus on ViT and LLM applications. It introduces a taxonomy for AI TC methods, analyzes standardized assessment criteria and public datasets, and assesses current models from preprocessing to feature extraction and classification, including case studies. The authors highlight the rising role of Transformers and LLMs for imaging, histopathology, and clinical text, while discussing limitations around data quality, privacy, and generalizability, and proposing future directions such as explainable AI, federated learning, and multi-modal data fusion. Overall, the paper surveys state-of-the-art AI-enabled TC diagnostics, offering guidance for researchers and clinicians on deploying accurate, privacy-preserving, and clinically actionable AI tools.

Abstract

The growing interest in developing smart diagnostic systems to help medical experts process extensive data for treating incurable diseases has been notable. In particular, the challenge of identifying thyroid cancer (TC) has seen progress with the use of machine learning (ML) and big data analysis, incorporating Transformers to evaluate TC prognosis and determine the risk of malignancy in individuals. This review article presents a summary of various studies on AI-based approaches, especially those employing Transformers, for diagnosing TC. It introduces a new categorization system for these methods based on artificial intelligence (AI) algorithms, the goals of the framework, and the computing environments used. Additionally, it scrutinizes and contrasts the available TC datasets by their features. The paper highlights the importance of AI instruments in aiding the diagnosis and treatment of TC through supervised, unsupervised, or mixed approaches, with a special focus on the ongoing importance of Transformers and large language models (LLMs) in medical diagnostics and disease management. It further discusses the progress made and the continuing obstacles in this area. Lastly, it explores future directions and focuses within this research field.

Machine Learning and Transformers for Thyroid Carcinoma Diagnosis: A Review

TL;DR

This review addresses the problem of thyroid carcinoma diagnosis using machine learning and Transformer-based approaches, synthesizing supervised, unsupervised, and deep learning methods with a focus on ViT and LLM applications. It introduces a taxonomy for AI TC methods, analyzes standardized assessment criteria and public datasets, and assesses current models from preprocessing to feature extraction and classification, including case studies. The authors highlight the rising role of Transformers and LLMs for imaging, histopathology, and clinical text, while discussing limitations around data quality, privacy, and generalizability, and proposing future directions such as explainable AI, federated learning, and multi-modal data fusion. Overall, the paper surveys state-of-the-art AI-enabled TC diagnostics, offering guidance for researchers and clinicians on deploying accurate, privacy-preserving, and clinically actionable AI tools.

Abstract

The growing interest in developing smart diagnostic systems to help medical experts process extensive data for treating incurable diseases has been notable. In particular, the challenge of identifying thyroid cancer (TC) has seen progress with the use of machine learning (ML) and big data analysis, incorporating Transformers to evaluate TC prognosis and determine the risk of malignancy in individuals. This review article presents a summary of various studies on AI-based approaches, especially those employing Transformers, for diagnosing TC. It introduces a new categorization system for these methods based on artificial intelligence (AI) algorithms, the goals of the framework, and the computing environments used. Additionally, it scrutinizes and contrasts the available TC datasets by their features. The paper highlights the importance of AI instruments in aiding the diagnosis and treatment of TC through supervised, unsupervised, or mixed approaches, with a special focus on the ongoing importance of Transformers and large language models (LLMs) in medical diagnostics and disease management. It further discusses the progress made and the continuing obstacles in this area. Lastly, it explores future directions and focuses within this research field.
Paper Structure (24 sections, 12 figures, 6 tables)

This paper contains 24 sections, 12 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Contribution of the paper
  3. Bibliometric analysis
  4. Roadmap
  5. Standardized assessment criteria and commonly used datasets
  6. Metrics
  7. Datasets for TC
  8. Summary of current models and methods
  9. Purpose of AI-driven examination
  10. Pre-processing
  11. ML features selection and extraction
  12. Selection methods
  13. Extraction methods
  14. SL-based TCD classification
  15. USL-based TCD classification
  16. ...and 9 more sections

Figures (12)

  • Figure S1: Approaches for identifying TC.
  • Figure S2: Bibliometric analysis in terms of: (a) documents by author; (b) documents by year; (c) documents by country; (d) documents by type.
  • Figure S3: Classification of TCD strategies utilizing AI.
  • Figure S4: Synopsis of CNN-driven research in TC diagnosis with percentages for accuracy, sensitivity, and specificity mahurkar2017normalization, and canton2021automaticpeng2021deepguan2019deepteixeira2017learningliu2019automatedqiao2021deepzhang2020detectiontekchandani2021severity.
  • Figure S5: Performance assessment of TC frameworks in percentages (%) for private TCD mehta2019high, guo2019xgboost, zhang2020detection, tran2023video, gu2019predictioncolakoglu2019diagnosticpark2021combining.
  • ...and 7 more figures