YY/T 1833.2-2022 Artificial intelligence medical device - Quality requirements and evaluation - Part 2: General requirements for datasets
1 Scope
This document specifies the general quality requirements and evaluation methods for datasets used in the full life cycle of artificial intelligence medical devices.
It is applicable to the development and evaluation of datasets used in the research and development, production, testing, quality control and other aspects of artificial intelligence medical devices.
2 Normative references
The following documents contain provisions which, through reference in this text, constitute provisions of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.
GB/T 2828.4 Sampling procedures for inspection by attributes - Part 4: Procedures for assessment of declared quality levels
GB/T 2828.11 Sampling procedures for inspection by attributes - Part 11: Procedures for assessment of declared quality levels for small population
GB/T 6378.4 Sampling procedures for inspection by variables - Part 4: Procedures for assessment of declared quality levels for mean
YY/T 1833.1 Artificial intelligence medical device - Quality requirements and evaluation - Part 1: Terminology
3 Terms and definitions
For the purposes of this document, the terms and definitions given in YY/T 1833.1 as well as the followings apply.
3.1
inspection by attributes
inspection relating to a specified requirement or a set of specified requirements that either classifies a unit product as acceptable or rejected, or only counts the number of rejected ones in unit products
[Source: GB/T 2828.1-2012, 3.1.3]
3.2
variables quality characteristic
quality characteristic of the unit product to be inspected that can be measured on a continuous scale
[Source: GB/T 8054-2008, 3.1.3]
3.3
sampling inspection by variables
process of randomly taking a certain number of unit products from the lot according to the specified sampling scheme, and then obtaining their quality characteristic values by measurement, test or other methods, comparing these values with the quality requirements, and finally judging whether the lot of products can be accepted
[Source: GB/T 8054-2008, 3.1.4]
3.4
lot
a definite part of a population composed under basically the same conditions according to sampling purposes
[Source: GB/T 10111-2008, 3.1.4]
3.5
accuracy
a measure that the data is correct in content and valid in form
[Source: GB/T 11457-2006, 2.22, modified]
3.6
precision
degree of accuracy or difference with respect to the quantities stated, such as 2-digit decimal number versus 5-digit decimal number
[Source: GB/T 11457-2006, 2.1160]
Foreword I
Introduction II
1 Scope
2 Normative references
3 Terms and definitions
4 Dataset description requirements
5 Dataset quality requirements
6 Dataset quality conformity evaluation
Annex A (Normative) Dataset type description
Annex B (Informative) Data screening and cleaning description
Bibliography
