By Sujal Shah On Jul 6, 2024

“The only constant in life is change.” Heraclitus.

This is especially true in science and technology. From the advent of steam engines, which transitioned human muscle power to machine power, to the development of computers and sensors that have taken over cognitive functions of memory, computing and perception, the progression has been remarkable. Artificial Intelligence (AI) has further advanced analytical, decision-making, and predictive capabilities. The next frontier is Generative AI (GenAI), which aims to emulate human creativity, making its output indistinguishable from human-generated content. A testament to its potential is ChatGPT, which amassed over one hundred million users within two months of its launch.

GenAI creates new data, images, video, audio or text by learning patterns from existing examples, simulating human creativity and imagination through algorithms. It’s revolutionising industries like art, design, and content creation by autonomously generating novel content. GenAI, a versatile tool, holds potential across various sectors. In this article, we delve into its transformative impact on the medical device industry.

GenAI for medical devices

Personalised healthcare

Healthcare should be personalised as everyone’s needs differ. The prevailing “one size fits all” approach in healthcare is set for a revolution. Standardised dosages of medicine and replacement parts (eg, knee) could soon be replaced by personalised solutions designed by GenAI. By analysing a patient’s medical history and current scans, GenAI can design personalised parts and customise medication, reducing the risk of complications and recovery time. When combined with 3D printing, GenAI can help manufacturers design personalised parts tailored to individual anatomical and physiological requirements.

Device enhancement

GenAI not only aids patients but also enables medical practitioners to diagnose better. For instance, a stethoscope could transmit data to a doctor’s mobile device, where GenAI could assist in diagnosis by considering the patient’s medical history. Similarly, radiologists could send scans to GenAI engines to generate draft reports, saving doctors considerable time.

Software as a medical device (SaMD)

As SaMD usage grows, manufacturers may pivot towards software-based solutions rather than continuing to focus on physical devices. For instance, GenAI-powered software on a doctor’s smartphone could interpret patient interactions, provide historical data, and simultaneously create meeting notes. I expect that in the not-too-distant future, the “manufacturers” (software developers) of SaMD, will monetise the software rather than a physical device.

24/7 Healthcare

Wearables allow real-time data collection from patients anywhere. This data is sent to the cloud, where GenAI can predict critical illnesses even before they occur, enabling remote patient monitoring and diagnosis. Unlike current practices, patients do not have to visit clinics for monitoring or diagnosis.

Workflow enhancement

GenAI can make significant contributions across different workflows, such as:

• In the operating room, a GenAI assistant can serve as a virtual aide to the surgeon, providing relevant patient data and recording surgical notes.
• When it comes to patient interaction, the GenAI-powered assistant can engage in natural language conversations tailored to each patient, considering their medical history, medications, reports, and doctor’s comments.
• Device manufacturers can streamline crucial manufacturing and administrative tasks such as generating the necessary documentation for product certification and drawing from product specifications, validation reports, and government guidelines.

Challenges of GenAI

The application of GenAI is not without its challenges. One such challenge is the potential for bias in the model, which can stem from the dataset used for training. For instance, a model designed to suggest dietary needs, if trained on a non-vegetarian diet, may falter when asked to provide recommendations for a vegetarian diet. This is just one example of the many biases that can arise due to the quality of data used.

The next challenge is accountability for the outcome. Who should be accountable when a GenAI model produces an incorrect result? Ensuring that the generated outcomes align with regulatory requirements is another hurdle. Moreover, it is crucial to guarantee that ethical standards are not compromised when a model generates responses.

While GenAI models can produce innovative results, they can also lead to undesirable outcomes. Therefore, any content produced by these algorithms must be overseen by qualified professionals.

Lastly, akin to any software, these systems must incorporate necessary security measures to ensure the safe and effective use of GenAI in various applications.

GenAI: A new era for medical domain

In summary, we see that GenAI can benefit the medical domain in several ways, such as:

• Providing extreme personalisation in customer interactions, medications, and devices.
• Enhancing product design and facilitating new product development.
• Streamlining the production, testing, and compliance of medical devices.
• Improving the efficacy of healthcare professionals by handling diverse data sources.
• Aiding in diagnosis and image analysis.
• Improving the efficiency of medical procedures.

The potential extends beyond patient care. Addressing challenges in areas such as administration and operations of organisations serves as an ideal launchpad for the integration of GenAI, due to their relative feasibility and lower risk. As organisations navigate this technological journey, they will gradually amass valuable experience and confidence, paving the way for the development of advanced clinical applications.

However, the full realisation of GenAI’s potential requires a collaborative approach. Policymakers, technologists, and healthcare professionals must join forces to ensure the ethical, unbiased, and secure implementation of GenAI. This collective effort is crucial in establishing the necessary safeguards for a better world, where technology serves humanity responsibly and effectively.