Overview
Artificial Intelligent/Machine Learning based Software as Medical Device (SaMD) is increasingly recognized for its potential to enhance the welfare of humanity. To ensure the safety, performance, and quality of medical device software utilizing AI/ML technology, TFDA has established guidance informed by regulations from the United States, Japan, and Korea, as well as the International Medical Device Regulatory Forum (IMDRF). This guidance was first announced in September 2020 and was updated in August 2021.
Latest Requirements
The latest requirements are summarized in the table below. However, additional documents may still be required during the review process, depending on the specific characteristics of each SaMD.
| Description | Functional Types |
|---|
| Defining the software as: | Artificial Intelligence, AI |
| Machine Learning, ML |
| Deep Learning |
| Description of the software function: | |
|---|
| Functional types, including |
| - Computer Assisted Detection (CADE) |
| - Computer Aided Diagnostics (CADx) |
| - Computer Aided Triage |
| Algorithm framework | Design |
|---|
| Training method |
| Testing principle and algorithm framework |
| Data limitation: | Shall include details of the training method, framework, process, data relevancy, and quality maintenance. |
|---|
| Shall describe the training module database, including population, clinical meaning, output type, output method, and other additional clinical significances. |
| There are three kinds of data applied in SaMD: data for training, validation, or testing. |
| Testing data must be rigorously separated from training or validation data to avoid bias in the results. |
| User Environmental and Data Management Safety: | Shall list limitations of operating environment and personnel. |
|---|
| If the SaMD can connect to the internet or has wireless transmission functions, refer to the "Cybersecurity Guidelines for Medical Device Manufacturers" and provide related documents. |
| Functional Verification and Validation: | Refer to the "Software Medical Device Guidelines", including: |
|---|
| - Level of concern |
| - Software description |
| - Device hazard analysis |
| - Software requirements specifications |
| - Architecture design chart |
| - Software design specification |
| - Software development environment verification and validation documentation |
| - Revision level history |
| - Unresolved anomalies/bugs or defects |
| Furthermore, the output results must comply with the intended use of the SaMD, including how the output will be presented, its limitations, and subsequent clinical suggestions. |
| Clinical significance: | Scientific evidence shall be provided to establish the applicability and suitability of the software specifications. It is recommended to include the following in the study protocol: |
|---|
| - Intended use |
| - Study objectives |
| - Patient population, e.g., age, ethnicity, race... |
| - Number of clinicians and qualifications |
| - Methodology for gathering clinical information |
| - Statistical methods for analyzing data study results |
Additional Information
New link: TFDA Guidance on SaMD - https://www.fda.gov.tw/TC/newsContent.aspx?cid=3&id=27247
Type Approval impacted: MEDICAL
Date published: 04/27/22
