In the last decade, artificial intelligence (AI) has gained popularity and transformed industries such as digital healthcare, manufacturing, and finance. In digital healthcare, AI drives innovations and improves patient outcomes by making processes more efficient, from detecting and diagnosing diseases to ensuring quality end-of-life care. AI is a broad category and typically integrates with its subfields — machine learning (ML), deep learning (DL), natural language processing (NLP), and computer vision (CV).
As AI continues to grow in healthcare there are more diverse and transformative applications being developed. Here are five key examples of areas significantly impacted by AI.
1) Predictive analysis
Predictive analysis leverages large amounts of data from electronic health records (EHRs), medical scans, and genetic data to forecast current and future health outcomes. During an outbreak, the analysis helps detect a disease early or can track its progression, enabling for timely management of optimized treatment plans.
2) Enhancement of medical imaging
Radiologists can improve their interpretation of medical imaging by enhancing the accuracy of image analysis, abnormality detection, and patient outcome prediction. Training AI algorithms to read medical images from an X-ray, magnetic resonance imaging (MRI), and computed tomography (CT) scans speeds up the process and reduces the likelihood of human errors.
3) Transformation of medical records
In the past, healthcare practitioners often relied on handwritten records or unstructured data entered by various healthcare practitioners. Today, records are converted into structured data through natural language processing algorithms. This process enables the integration of various health systems to facilitate advanced research and improve operational efficiency.
4) Advancement in telemedicine
AI in telemedicine improves medical assistance accessibility and optimizes healthcare operations. Through telemedicine, patients can interact with AI-powered chatbots that provide near real-time support. The AI algorithm can suggest treatments based on the patient’s symptoms and medical records while also analyzing results against population trends and data collected from electronic health records and remote monitoring devices.
5) Acceleration in drug discovery and development
AI analyzes the biological data pool to identify, screen, and optimize the drug identification process. AI also helps enhance drug development by using historical trial data to design more efficient clinical trials.
The list above is not exhaustive; AI plays a significant role in many other medical applications. AI is a powerful tool that spans numerous domains and offers numerous benefits. However, these benefits come with challenges and growing concerns around data privacy and security, biases, ethicality, accuracy, and reliability.
The challenges
Governmental organizations are seeing the immense economic and societal benefits of AI. Comprehensive guidelines and regulatory frameworks are in active development to ensure AI technologies are developed and deployed in a safe, secure, and trustworthy manner. The White House issued an executive order in October 2022 that covers a comprehensive strategy for responsible innovations. The order addresses the risks by including components such as promoting innovation, ethical and responsible use of AI, privacy and security, public trust, and workforce development.
The United States Food and Drug Administration (FDA) has also been actively developing guidelines for using AI and ML in medical devices, emphasizing machine learning-enabled device software functions (ML-DSFs). The guidelines recommend what information should be included in a Predetermined Change Control Plan (PCCP) for a marketing submission for an ML-DSF.
To boost AI development across the economy, on May 21, 2024, the European Union (EU) Council approved the EU Artificial Intelligence Regulation, which establishes a common regulatory and legal framework for AI within the EU and aims to foster the development and update of innovative, safe, and trustworthy AI systems. The EU AI Act is applicable to various entities, including providers, deployers, importers, and distributors of AI systems.
Even with governmental organizations backing innovation and the use of AI in medical devices, the negative perception of AI in the medical world persists. Healthcare decisions have traditionally been made by humans. The assistance of smart tools raises concerns about integration, accountability, transparency, and privacy.
In addition, the concerns are multifaceted and must be addressed carefully. The use of AI in healthcare involves a significant amount of sensitive patient data and health records. Protecting patient data in a foolproof way is an essential prerequisite to maintaining trust and complying with regulations like the Health Insurance Portability and Accountability Act (HIPAA). There have been discussions about the lack of transparency in the inclusivity of training data across diverse demographics, the decision-making process of AI algorithms, and the accountability for errors made by AI systems or undetected errors.
Healthcare facilities typically delay the implementation of new tools due to the concern about interoperability. Different healthcare systems use varied formats and standards, making it difficult for AI systems to integrate seamlessly into the existing platform. Even the current clinical workflows will need careful planning and customization to incorporate AI tools without disrupting the day-to-day operations of healthcare practitioners.
Overcoming negative perceptions
These concerns and challenges highlight the need for a more diverse and comprehensive testing approach. AI tools used in life-critical systems, especially healthcare systems, should undergo rigorous testing and obtain regulatory approval before implementation. Testing ensures the safety and efficacy of these technologies. Technology companies and healthcare innovators must take a multifaceted approach to test the tool by following the guidelines set by health authorities and governments.
An AI tool’s power depends on the quality of the data it receives. Preparing the data for testing and addressing biases within the dataset is important. Innovators must start with high-quality, varied, and representative datasets for training and validation. The caveat to being successful is that innovators should always seek new data sources and perform continuous testing.
The clean and preprocessed data develops and trains the AI tool with a set of algorithms for simulation and benchmarking. Subsequently, innovators will need to complete the clinical trials to ensure the algorithm’s efficacy compared to standard care or alternative tools. Some innovators will choose to use automated software to help with the iterative algorithm testing to ensure sufficient coverage to validate the products, especially if it is testing a software application.
Regulatory compliance testing on AI medical tools is crucial before deployment in clinical settings. There are different regulatory bodies and standards that regulate and supervise the standards management for medical devices, such as the Food and Drug Administration (FDA) in the United States, the European Medicines Agency (EMA) in Europe, the Pharmaceuticals and Medical Devices Agency (PMDA) in Japan, and the National Medical Products Administration (NMPA) in China. Innovators must adhere to the relevant guidelines for medical device quality management systems.
Data privacy and security laws safeguard patient data when testing AI medical tools. Innovators must comply with regulations governing personal health data, such as the Health Insurance Portability and Accountability Act (HIPAA) in the United States and the General Data Protection Regulation (GDPR) in Europe. Implementing tight security measures is important to protect data from unauthorized access and breaches. The innovators can employ strategies such as data anonymization, de-identification, encryption, and access controls to protect sensitive health information.
The healthcare industry can foster a more positive outlook and encourage wider adoption of AI technologies by addressing concerns head-on and continuously showcasing AI’s tangible benefits. All parties play a role in working together to harness the benefit of AI for improved patient care — from governmental organizations, data scientists, healthcare professionals, and innovators.
Janet Ooi
Janet Ooi is Digital Healthcare Solutions Manager at Keysight Technologies.