The new article describes in detail special regulatory requirements concerning software medical devices utilising Artificial Intelligence technology to ensure the specific risks associated with it are duly addressed.

HSA Guidance on Labeling for Medical Devices: Implantable Devices and IVDs

The Health Sciences Authority (HSA), Singapore’s regulatory agency in healthcare products, has published a guidance document dedicated to software medical devices in the context of a life cycle approach.

The document provides an overview of the applicable regulatory requirements and additional clarifications and recommendations to be considered by medical device manufacturers (software developers) to ensure compliance with them.

At the same time, the authority reserves the right to change the guidance and recommendations provided therein, should such changes be reasonably necessary to reflect corresponding amendments to the underlying legislation. 

The scope of the guidance covers, among other things, the aspects related to the regulatory framework and considerations for Artificial Intelligence Medical Devices (AI-MDs), emphasizing the unique challenges and requirements posed by integrating AI technology into medical devices.

The increasing use of AI technologies in medical devices introduces a new dimension to medical device regulation, necessitating specific considerations for the responsible development and deployment of AI MDs.

This includes compliance with existing legislation and guidelines pertinent to healthcare, such as the Personal Data Protection Act, Human Biomedical Research Act, and Healthcare Services Act. Developers and implementers must ensure AI-MDs are developed with a keen eye on safety, efficacy, and privacy.

Regulatory Framework for AI-MDs

First, the authority mentions that AI-MDs, while subject to the general regulatory principles applicable to medical devices, demand additional considerations due to their unique features, such as continuous learning capabilities, the extent of human intervention, and the dynamics of model training and retraining.

An ISO 13485-based Quality Management System (QMS) is fundamental for managing the lifecycle of AI-MDs, from design and development to deployment and post-market surveillance.

FDA on assessing credibility of computational modelling2

Pre-Market Registration Requirements

The pre-market phase for AI-MDs entails a detailed submission process covering several key areas:

  • Dataset: Specifications for input data, including its selection, pre-processing, and rationale, are critical. This ensures the AI model’s output is reliable and based on appropriately curated data.
  • AI Model: Information on the AI model used, including its selection rationale, limitations, and evaluation metrics, is necessary to establish its suitability for its intended medical purpose.
  • Performance and Clinical Evaluation: Test protocols, reports, and clinical relevance of the AI-MD’s outputs must be documented, showcasing the device’s diagnostic accuracy and clinical utility.
  • Deployment: Details on the device workflow, human intervention levels, and software versioning are vital for ensuring safe and effective deployment in healthcare settings.

Considerations for AI-MDs with Continuous Learning

According to the guidance, AI-MDs that adapt and learn from new data post-deployment require stringent control measures to ensure their learning processes do not compromise the device’s specified performance or safety.

This includes quality checks, validation strategies, and software version control to manage continuous learning effectively and responsibly.

Post-Market Monitoring and Changes

The lifecycle of an AI-MD extends into the post-market phase, where active monitoring and tuning of the device are crucial to maintaining its performance and safety in the real world. Developers and distributors must submit periodic reports to regulatory bodies to facilitate oversight and timely intervention if needed.

Additionally, any changes to the AI-MD, whether in response to learning outcomes or other factors, necessitate a structured Change Notification process to ensure ongoing compliance and safety.


The regulatory landscape for AI-MDs reflects the complexity and dynamic nature of integrating artificial intelligence into medical devices. From pre-market considerations to post-market surveillance, each step is designed to ensure that AI-MDs comply with existing regulatory standards and adapt to the unique challenges posed by AI technology. By adhering to these guidelines, developers can pave the way for innovative, safe, and effective medical devices that leverage the full potential of AI.

With the implementation of the approach described in the guidance, the authority intends to ensure the proper balance between facilitating further development of novel medical devices utilising innovative technologies and ensuring the safety of patients.

How Can RegDesk Help?

RegDesk is a holistic Regulatory Information Management System that provides medical device and pharma companies with regulatory intelligence for over 120 markets worldwide. It can help you prepare and publish global applications, manage standards, run change assessments, and obtain real-time alerts on regulatory changes through a centralized platform. Our clients also have access to our network of over 4000 compliance experts worldwide to obtain verification on critical questions. Global expansion has never been this simple.