Your browser doesn't support javascript.
loading
Mitigating Bias in Radiology Machine Learning: 2. Model Development.
Zhang, Kuan; Khosravi, Bardia; Vahdati, Sanaz; Faghani, Shahriar; Nugen, Fred; Rassoulinejad-Mousavi, Seyed Moein; Moassefi, Mana; Jagtap, Jaidip Manikrao M; Singh, Yashbir; Rouzrokh, Pouria; Erickson, Bradley J.
Afiliação
  • Zhang K; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Khosravi B; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Vahdati S; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Faghani S; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Nugen F; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Rassoulinejad-Mousavi SM; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Moassefi M; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Jagtap JMM; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Singh Y; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Rouzrokh P; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
  • Erickson BJ; Radiology Informatics Laboratory, Department of Radiology, Mayo Clinic, 200 1st St SW, Rochester, MN 55905.
Radiol Artif Intell ; 4(5): e220010, 2022 Sep.
Article em En | MEDLINE | ID: mdl-36204532
ABSTRACT
There are increasing concerns about the bias and fairness of artificial intelligence (AI) models as they are put into clinical practice. Among the steps for implementing machine learning tools into clinical workflow, model development is an important stage where different types of biases can occur. This report focuses on four aspects of model development where such bias may arise data augmentation, model and loss function, optimizers, and transfer learning. This report emphasizes appropriate considerations and practices that can mitigate biases in radiology AI studies. Keywords Model, Bias, Machine Learning, Deep Learning, Radiology © RSNA, 2022.
Palavras-chave

Texto completo: 1 Base de dados: MEDLINE Idioma: En Ano de publicação: 2022 Tipo de documento: Article

Texto completo: 1 Base de dados: MEDLINE Idioma: En Ano de publicação: 2022 Tipo de documento: Article