TY - JOUR
T1 - Mitigating Bias in Radiology Machine Learning
T2 - 3. Performance Metrics
AU - Faghani, Shahriar
AU - Khosravi, Bardia
AU - Zhang, Kuan
AU - Moassefi, Mana
AU - Jagtap, Jaidip Manikrao
AU - Nugen, Fred
AU - Vahdati, Sanaz
AU - Kuanar, Shiba P.
AU - Rassoulinejad-Mousavi, Seyed Moein
AU - Singh, Yashbir
AU - Vera Garcia, Diana V.
AU - Rouzrokh, Pouria
AU - Erickson, Bradley J.
N1 - Publisher Copyright:
© RSNA, 2022.
PY - 2022/9
Y1 - 2022/9
N2 - The increasing use of machine learning (ML) algorithms in clinical settings raises concerns about bias in ML models. Bias can arise at any step of ML creation, including data handling, model development, and performance evaluation. Potential biases in the ML model can be minimized by implementing these steps correctly. This report focuses on performance evaluation and discusses model fitness, as well as a set of performance evaluation toolboxes: namely, performance metrics, performance interpretation maps, and uncertainty quantification. By discussing the strengths and limitations of each toolbox, our report highlights strategies and considerations to mitigate and detect biases during performance evaluations of radiology artificial intelligence models.
AB - The increasing use of machine learning (ML) algorithms in clinical settings raises concerns about bias in ML models. Bias can arise at any step of ML creation, including data handling, model development, and performance evaluation. Potential biases in the ML model can be minimized by implementing these steps correctly. This report focuses on performance evaluation and discusses model fitness, as well as a set of performance evaluation toolboxes: namely, performance metrics, performance interpretation maps, and uncertainty quantification. By discussing the strengths and limitations of each toolbox, our report highlights strategies and considerations to mitigate and detect biases during performance evaluations of radiology artificial intelligence models.
KW - Convolutional Neural Network (CNN)
KW - Diagnosis
KW - Segmentation
UR - http://www.scopus.com/inward/record.url?scp=85139163075&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85139163075&partnerID=8YFLogxK
U2 - 10.1148/ryai.220061
DO - 10.1148/ryai.220061
M3 - Article
AN - SCOPUS:85139163075
SN - 2638-6100
VL - 4
JO - Radiology: Artificial Intelligence
JF - Radiology: Artificial Intelligence
IS - 5
M1 - e220061
ER -