The future development of medical AI: three major trends worth paying attention to

When the COVID-19 epidemic is raging, people's mental health is in crisis, medical costs are rising, and the aging population is intertwined with various trends, industry leaders have accelerated the development of medical-specific AI applications. One of the signals from the venture capital market shows that more than 40 startups have raised a large amount of funds (more than 20 million US dollars) to build medical AI solutions, but how is AI used in the medical industry?

A recent report titled "Healthcare AI Survey 2022" surveyed more than 300 respondents from around the world to understand and define the challenges, achievements and use cases of healthcare AI . This is the second year since the survey was launched, and while there are no significant changes in terms of results, some interesting trends do emerge that are indicative of how things may change in the coming years. While some aspects of this evolution are positive (such as the spread of artificial intelligence), others are less exciting (such as an increased attack surface). Let’s take a look at three of them. Trends businesses need to understand.

1. Use no-code tools to make AI easier to use and popularize

According to Gartner estimates, by 2025, 70% of new applications developed by enterprises will use no-code or low-code technology , this figure is higher than less than 25% in 2020. Low-code simplifies programmers' workloads, and no-code solutions that don't require data science intervention will have the greatest impact on enterprises and other fields, which explains why the use of artificial intelligence technology is moving from technical professionals to fields Experts are exciting.

For the medical industry, this means that more than half (61%) of the respondents will regard clinicians as their target users, followed by medical service payers (45%) and medical IT companies (38%), coupled with the rapid development of medical AI applications, large investments, and the popular availability of open source technology, it shows that medical AI is being more widely adopted.

This is important: Putting code into the hands of medical staff, as easy as using common office tools such as Excel or Photoshop, will change AI and make it better. In addition to being easier to use, medical AI can also achieve more accurate and reliable results because it is now used and controlled by medical professionals (rather than software professionals). Of course, these changes won't happen overnight, but for AI, its increasing use by domain experts is a significant step forward.

2. Tools are becoming more complex and texts are becoming more useful

There are other encouraging findings from this survey, such as the continuous development and advancement of AI tools and the desire of users for specific model for in-depth study. When respondents were asked what technologies they planned to adopt by the end of 2022, many technology leaders mentioned data integration (46%), business intelligence (44%), natural language processing (43%) and data annotation ( 38%). Text is currently the data type most likely to be used by AI applications. At the same time, respondents’ emphasis on natural language processing (NLP) and data annotation indicates that more complex AI technologies are on the rise.

These tools provide support for many important usage scenarios, such as supporting clinical decision-making, drug discovery, and medical strategy evaluation, etc. Especially after these two years of the COVID-19 pandemic, as we develop new vaccines and learn how to better support the needs of the medical system after a large-scale event, it is so important to make progress in these technical areas. Through these examples, it is clear that the use of AI in healthcare is very different from other industries, and therefore requires a different approach.

As a result, technology leaders and interviewees from mature organizations cited the availability of healthcare-specific models and algorithms as the most important requirement when evaluating whether to install a software library on-premises or adopt a SaaS solution. Why. Judging from various aspects such as the venture capital landscape, the existing software libraries in the market, and the needs of artificial intelligence users, medical-specific will only continue to grow in the next few years.

3. Security issues are becoming increasingly prominent

In the past year, AI has made many advances, and a series of new attack vectors have also been introduced. When respondents were asked what types of software they use to develop AI applications, the most popular choices were locally installed commercial software (37%) and open source software (35%). Most notably, cloud service usage has decreased by 12% (30%) compared to last year's survey results, most likely due to privacy issues caused by data sharing.

In addition, the majority of respondents (53%) chose to rely on their own data to validate the model rather than using indicators from third parties or software vendors. Respondents from mature organizations (68%) said they preferred to use a model of internal assessment and self-adjustment. And, because there are strict controls and various procedures around the processing of medical data, this also explains why AI users want to handle these issues within the organization as much as possible.

But regardless of preferences for software or how users validate models, escalating medical security threats can have a significant impact. Although other critical infrastructure services also face various challenges, the consequences of medical breaches are no longer just reputational and financial losses. Loss of data or attacks on hospital equipment can be a matter of life and death.

AI is poised for even more significant growth as developers and investors work to put AI technology into the hands of everyday users. But as AI becomes more widely adopted and models and tools continue to improve, safety and ethics will become a key area of ​​focus. How AI technology in the medical industry will develop this year and what it means for the future of the industry are all worth looking forward to.

The above is the detailed content of The future development of medical AI: three major trends worth paying attention to. For more information, please follow other related articles on the PHP Chinese website!