相关实验视频

Updated: Sep 10, 2025

Spontaneous Murine Model of Anaplastic Thyroid Cancer

Spontaneous Murine Model of Anaplastic Thyroid Cancer

Published on: February 3, 2023

甲状腺癌人工智能模型中的代表性和偏差:系统性审查

Rashi Ramchandani^1,2, Eddie Guo³, Sanaz G Biglou¹

¹Faculty of Medicine, University of Ottawa, Ottawa, Ontario, Canada.

Thyroid : official journal of the American Thyroid Association

|August 27, 2025

概括

在甲状腺癌护理中的人工智能 (AI) 是有前途的,但由于偏见的数据集而面临挑战. 这项审查发现人工智能模型中的人口代表性不足,需要更公平的数据来改善全球患者的结果.

科学领域:

医疗信息学
癌症学
人工智能

背景情况:

人工智能对甲状腺癌治疗越来越感兴趣,旨在改善诊断,预测结果和个性化治疗.
在甲状腺癌算法中潜在的AI偏差,导致代表性不足的人群的护理差异.
需要对甲状腺癌的人口代表性和偏见进行系统评估.

研究的目的:

系统地审查用于甲状腺癌治疗的AI模型.
评估人口代表性,并确定这些AI模型中的潜在偏差.
评估人工智能模型数据与全球甲状腺癌流行病学的一致性.

主要方法:

在EMBASE,PubMed和Google Scholar上进行系统的文献搜索,直到2024年1月.
包括涉及甲状腺癌管理的人工智能研究与人口统计数据.
两位在PROSPERO注册的独立审查员进行数据提取和偏差风险评估.

主要成果:

审查了197项研究,主要关注诊断 (133) 和预测/预后 (47).
主要参与者来自中国 (124) 和美国 (26);女性 (12,410) 多于男性 (4222).
与全球患病率相比,东亚人显著不足,白人和黑人参与者过多;社会经济因素经常被遗漏.

结论:

目前的甲状腺癌人工智能模型在数据多样性和代表性方面存在重大差距.
人工智能模型与一般流行病学趋势保持一致,但缺乏全面的人口纳入.
考虑到不同的人口结构和社会文化背景,开发更公平的人工智能模型对于改善全球甲状腺癌护理至关重要.

关键词:

人工智能偏见多样性股权资本加入甲状腺癌

更多相关视频

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

Computer-Aided Three-Dimensional Visualization in the Treatment of Locally Advanced Thyroid Cancer

Computer-Aided Three-Dimensional Visualization in the Treatment of Locally Advanced Thyroid Cancer

Published on: June 9, 2023

相关概念视频

Synthesis and Regulation of Thyroid Hormones

Synthesis and Regulation of Thyroid Hormones

Low blood levels of the thyroid hormones — triiodothyronine (T3) and thyroxine (T4) — signal the hypothalamus to release the thyrotropin-releasing hormone (TRH). TRH then reaches the pituitary gland and stimulates the release of thyroid-stimulating hormone(TSH) into the bloodstream.
Upon reaching the thyroid gland, TSH stimulates the follicular cells' active uptake of iodide ions from the blood. The ions diffuse to the apical surface of the cells and are oxidized to iodine. The...

Bias in Epidemiological Studies

Bias in Epidemiological Studies

Biases can arise at various stages of research, from study design and data collection to analysis and interpretation. Recognizing and addressing these biases is essential to ensure the validity and reliability of epidemiological findings.Broadly speaking, biases in epidemiology fall into three main categories: selection bias, information bias, and confounding. A more detailed description of possible biases is: