The Future of AI-powered Healthcare

What is artificial intelligence (AI)? Is it the evoking computer from sci-fi aware of its own existence and determined to destroy humanity? Is it a robot that does our job for us while we kick our feet up? Right now, maybe it is neither, it can be defined as “A System that mimics human intelligence to perform complex tasks using advanced learning algorithms that capture underlying patterns and relationships from the data they collect.”  The tasks and benefits from such a system can be many but generally serve as three main use categories: accuracy improvement, automation of tasks, or a recommendations engine.

In developing a SAMD (Software As a Medical Device) product consider both the regulatory guidelines and best practices.  The FDA is partnering with industry to develop regulations in this emerging field, they recently released a guidance on Clinical Decision Support Software describing the criterion in which software is considered a medical device by the agency. And, during software life-cycle development, ISO 62304 outlines the processes of risk management, maintenance, configuration management, and problem resolution.

Developers should build in systems on the front end for data mining whether in the form of document capturing tools, video data collection, speech recognition, or otherwise.  And, comprehensive cybersecurity around these data sources in addition to the access, analysis, and output systems.

Lastly, algorithms should take bias into account.  This is already present in the diagnosis making process today, clinicians can jump to conclusions based on early information and stick to their guns even as new information becomes available (premature closure / anchoring). The algorithms themselves can have bias, in how data is fitted when machine learning is automated.

  • Automation Bias: Tendency of people to show deference to automated output, maybe due to person’s lack of confidence/experience, or assumption that the automation designed to make the correct determination.
  • Fitting Bias: Over Fitting- Automation has been overly relying on the trained data and does not provide correct responses when given new information.  Under Fitting - Machine is under trained and doesn’t correctly identify relationships between the variables.

Widespread AI use is in its infancy, its currently being leveraged across several surgical products currently on the market including surgery planners, guidance systems, AR, blood loss monitoring, and predictive analytics. The future holds many opportunities for AI to burn down existing healthcare challenges.

Accuracy Improvement:

  • Comprehensive Patient Medical Information
  • Summarization and Highlighting of Patient Case History
  • Accurate Encoding of procedures and diagnosis for insurance
  • Accurate diagnosis from medical images
  • Risk-aware decision making –using predictive analysis of surgical outcome, implant choice, length of hospital stay, risk of re-hospitalization
  • Post op x-ray, feedback loop, feedback to surgeon on trending accuracy stats, predictive risks
  • Physician burnout - make less errors during diagnosis
  • Physician shortage – making fewer surgeons more efficient

Automation Enabled Improvements:

  • Improved surgical planning / operation
  • AI-assisted surgical robotics
  • Supply chain automation
  • Reduced non-conformances, out-of-commission instrument sets
  • Reduced waste, reprocessing costs
  • Smart intra-op assistant / training

Recommendations Engine:

  • Patient/procedure/surgeon customized device on demand
  • Fair surgeon success ratings based on predictive risk/outcomes
  • Informing consumers on surgeon/facility for their condition to maximize outcomes

It probably won’t be too far into the future before some of these AI-enabled improvements become mainstream practice in the healthcare domain. The recent advances in ChatGPT have shown how complex knowledge intensive tasks such as text summarization, essay generation, intelligent Q&A (Question and Answer), etc. can be accomplished by current language models.  Convolutional Neural Networks (CNN)-based deep learning models are showing promise for automatic detection and classification of tumors in medical imaging. Advanced Machine Learning (ML), Rule-based modeling, and Embedded-AI can help with addressing other opportunities such as risk prediction, improved surgical planning, AI-assisted robotic devices, supply chain automation, and customized recommendations

AI will help in bringing consistency in the process, improve overall efficiency, reduce cost of operations while adhering and improving the regulatory compliance.

Interested in implementing AI/ML technology into your business?

Kaleidoscope uses advanced learning algorithms to capture patterns and relationships within your data to help you better understand the data collected and provide both exploratory and predictive analytics based on findings. Contact Matt Suits: [email protected]

Back to Insights + News

Authors

  • Eric Kennedy

    Eric Kennedy

    Principal Engineer | [email protected]

    Eric Kennedy is an engineer at Kaleidoscope Innovation based in Cincinnati, Ohio, and has over 15 years of global medical device experience leading large- and medium-scale concept-to-launch orthopedic, micro-surgical, cardiovascular and ophthalmic devices.

  • Dr. Ravi Nandigam

    Dr. Ravi Nandigam

    Principal Consultant

    Dr. Ravi Nandigam is a Principal Consultant in the Advanced Engineering Group at Infosys. He has 15 years of experience applying Artificial Intelligence, Machine Learning, and Software-based solutions in diverse Engineering domains. Dr.Nandigam is an inventor of a patent and author of many technical articles in peer-reviewed international journals on topics of AI/ML-based applications in Engineering.

  • Dr. Ravi Kumar G. V. V.

    Dr. Ravi Kumar G. V. V.

    Vice President and Head Advanced Engineering Group (AEG)

    Dr. Ravi Kumar is Vice President and Head Advanced Engineering Group (AEG) of Engineering Services, Infosys. He led numerous innovations and applied research projects for more than 26 years. His areas of expertise include mechanical structures and

AI as Intelligent Design? Not Yet, But It’s Coming.

From art generators to chatbots, AI seems to be having its zeitgeist moment in popular culture. But for those of us who work in design, the near-term and future applications of AI have been lively discussion points in strategic planning meetings for quite some time. There is no doubt that AI will be an instrumental part of our world’s future. It will allow us to rapidly synthesize all the data being collected via our phones, cameras, computers, smart devices, and much more, giving us the ability to decipher and understand that data in illuminating, meaningful, and likely, world-changing ways.  

What does this mean for the design industry? Though it may be a long time before AI is able to design a product from the ground up, the potential is clearly there. In fact, we believe AI is a tool that designers should be adding to their arsenal sooner rather than later. 

 

Putting AI to Work 

To put our money where our industry-informed opinions are, the Kaleidoscope Innovation team recently embarked on a studio project to design a high-end lighting fixture that could mimic lighting patterns found in nature. The project would enable our team to flex our aesthetic skills while using the full range of our design toolbox. One of those tools is Midjourney, a proprietary artificial intelligence program produced by an independent research lab by the same name. Though still in the open beta phase, Midjourney proved to be a useful partner in our mission. The collaboration between AI and the guiding hand of our expert design team delivered intriguing results. 

One important distinction about the AI portion of the project: We were not setting out to produce real-world functionality, and in fact, we had no expectation or need for the AI to produce fleshed-out ideas or even design sketches. This experiment was about exploring new territories in aesthetics and applying them to materials and manufacturability considerations. 

Our first step was to gather a team to collaborate on the search terms that would help visually articulate the aesthetic aspirations for our new fixture. Midjourney works by inputting text-based prompts, which the AI algorithm uses to generate new images using vast databases of existing images. The terms we fed the algorithm included chandelier, lighting, brilliant, elegant light, airy, crystalline patterns of light, dancing, photorealistic detailed plants, greenery, daytime, bright, modern, beautiful, natural colors, garden, and greenery. The team also used technical inputs alongside these qualitative descriptors to determine the aspect ratio and resolution while also guiding the algorithm to reference certain lighting styles and rendering approaches.  

Digesting these descriptive words, Midjourney searched vast amounts of data across the internet to create original—albeit amalgamated—artwork. The images it produced reflected the algorithm’s interpretation of the inputs the team provided. From there, we tweaked specific inputs to alter the color, lighting, tone, and subject matter, continuing to iterate until we had collected a series of AI-generated lighting fixtures that could inspire the team.

How Did AI Do?  

Based on the text inputs the team provided, Midjourney was able to identify design elements that could produce the effect of light shining through leaves. The images it produced looked organic, almost surreal in the way they were able to capture the kind of nature-made glow and transparency that is elusive in real-world lighting solutions. The various iterations of artwork then became mood boards that set up our team to brainstorm ways in which the effect could conceivably be produced.  

The algorithm’s interesting use of materials, colors, lighting effects, and overall mood inspired us to apply those attributes to a holistic design. In other words, instead of our team scratching their heads visualizing how the light should transmit, AI provided us with ideas that enabled us to focus on materials, manufacturability, technical requirements, and more. Rather than spending hours scouring the internet for inspirational imagery, the team was able to craft that inspiration imagery ourselves through AI in a fraction of the time—imagery that exactly aligned with our design vision. 

concept board

Without question, Midjourney served as a highly effective springboard that sparked ideas our team would probably not have come up with starting from a blank sheet of paper and pen. In this sense, AI provides an upfront efficiency that can move a project farther down the road faster than it might otherwise have gone. Perhaps more than that, a significant strength of AI in this application is that it can cast a wide net in terms of inspiration and exploration. It’s an open mind, and designers should be willing—and eager—to go down the rabbit holes, teasing out new possibilities. Once an intriguing direction is established, the designer can take over to turn the AI-generated inspiration into an actual product.  

The key to a successful AI collaboration is plugging in the right words or phrases to best draw out the AI. And so, crafting prompts could be viewed more as art than science. Further, with a program like Midjourney, there is an element of unpredictability: You don’t have much control over what you’re going to get out of it. There is a lot of trial and error and shooting in the dark. Therefore, if you already have a set idea in mind, using AI to design it will probably be more frustrating than productive.  

The inherent aspect of exploration and discovery is a factor to consider as well. Our team felt excited about experimenting with this technology specifically because the lighting fixture was an internal project. Had we been designing for a client, we would have been more hesitant to use AI while balancing product requirements, timeline, budget, and resources.  

Lastly, because this was a purely aesthetic exercise, we weren’t trying to solve any mechanical problems through AI—that’s skill is not in its wheelhouse at this point. This limitation provides a real barrier to the widespread adoption of AI, but as the algorithms improve over time, AI may be able to help us solve even our stickiest mechanical problems. 

Beyond leveraging AI for creative exploration, Kaleidoscope has also put it to use in some of our research work. As part of our insights and user experience programs, we often do ethnography or time-and-motion studies in which we observe individuals interacting with a tool or experience. Typically, one of our team members is responsible for reviewing videos to log data, tracking everything from how often someone does something to the amount of time it takes them to do it. It’s a time-consuming process that has led us to start dabbling with programming AI to analyze video recordings for certain elements and then export the data quickly and effectively. Using AI to track the frequency and duration of actions for time-and-motion studies shows tremendous potential to save time and reduce costs while freeing our team members to focus on more creative assignments. 

The Verdict 

The Kaleidoscope team came away with an appreciation for where AI can support our design efforts today, particularly as a powerful aid in producing aesthetic inspiration and as a tool to sort and output raw data. Both help the design process in productive ways and serve as a small window to what may someday be an AI-driven design future.

This was written for IDSA, if you'd like to see the INNOVATION Magazine article, please check out idsa.org/news-publications/innovation-magazine/spring-2023/

Back to Insights + News

Authors

  • Tony Siebel

    Tony Siebel

    Director of Design | [email protected]

    Tony Siebel is director of design at Kaleidoscope Innovation, delivering a user-centered mindset to products and experiences.

  • Tom Gernetzke

    Tom Gernetzke

    Senior Industrial Designer | [email protected]

    Tom Gernetzke is a senior lead industrial designer at Kaleidoscope Innovation and has spent the last 12 years creatively bringing new product ideas to life.

  • Caterina Rizzoni

    Caterina Rizzoni

    Lead Industrial Designer | [email protected]

    Caterina Rizzoni is a lead industrial designer at Kaleidoscope Innovation and is the Director-at-Large of Conferences for IDSA.

Infosys Medical Devices and Engineering Services x Kaleidoscope Innovation

The Healthcare and Medical devices industry is undergoing a revolutionary transformation in the way solutions and devices are being formulated and developed. Medical devices are becoming more connected than ever and remote patient monitoring with data analytics is becoming a norm.

It is imperative for the medical device companies to adopt a strategic approach to stay ahead of the innovation curve by leveraging technology advancements in multiple areas such as mobility, wireless, cloud, and analytics to drive innovation that addresses market needs and challenges of longer device development cycles, optimization of development processes, and high production costs.

At Infosys, we help our clients in designing customized devices, end-to-end product development, maintenance, manufacturing support, regulatory documentation, and product compliance and certifications. We also help optimize R&D cost and improve supply chain efficiencies by leveraging new technologies and partner ecosystems. This is to bring innovative medical devices and Software as a Medical Device applications into the market with the objective of improving patient care while reducing the cost of care.

Our ISO 13485 certified processes and Quality Management System ensures high-quality product development which enables our client to meet their regulatory needs and objectives. With our recent acquisition of product design and development firm, Kaleidoscope Innovation, we plan to redefine patient treatment and consumer health across the globe.

Full article can be found on Infosys.com

Back to Insights + News