Home>Article>Technology peripherals> Three major trends in future medical AI worth paying attention to
Between the coronavirus outbreak, the mental health crisis, rising healthcare costs and an aging population, industry leaders are rushing to develop artificial intelligence (AI) applications specific to healthcare. One signal is coming from the venture capital market: more than 40 startups have raised significant funding ($20 million or more) to build AI solutions for the industry. But how is artificial intelligence actually used in healthcare?
The 2022 AI in Healthcare Survey asked more than 300 respondents from around the world to better understand the challenges, achievements and applications that define AI in healthcare case. The results did not change significantly in the second year, but they did show some interesting trends that are indicative of how the next few years will unfold. While some aspects of this evolution are positive (democratization of AI), other aspects are less exciting (larger attack surface). Here are three trends businesses need to know about:
Gartner estimates that by 2025, 70% of new applications developed by enterprises will use no-code or low-code technologies (up from less than 25% in 2020). While low-code simplifies a programmer’s workload, it’s no-code solutions that require no data science intervention that will have the greatest impact on businesses and beyond. That’s why it’s exciting to see the leap in the use of AI from technical job titles to domain experts.
For healthcare, this means that more than half (61%) of respondents to the Artificial Intelligence in Healthcare survey identified clinicians as their target users, This is followed by healthcare payers (45%) and healthcare IT companies (38%). This, coupled with significant development and investment in healthcare-specific AI applications and the availability of open source technology, points to broader industry applications.
This is significant: Putting code into the hands of health care workers, just like common office tools like Excel or Photoshop, will make AI better. In addition to making the technology easier to use, it also enables more accurate and reliable results because it can be in the hands of medical professionals (not software professionals). These changes won't happen overnight, but the growth of AI with its core users being domain experts is certainly a big step forward.
Other encouraging findings involve the advancement of artificial intelligence tools and users The desire to delve deeper into a specific model. When asked what technologies they plan to adopt by the end of 2022, technology leaders in the survey cited data integration (46%), business intelligence (44%), NLP (43%) and data annotation (38%). Text is now the data type most likely to be used in AI applications, and an emphasis on natural language processing (NLP) and data annotation suggests that more sophisticated AI techniques are on the rise.
These tools can support important activities such as clinical decision-making, drug discovery and medical policy evaluation. After two years of COVID-19, the importance of progress in these areas is clear as we develop new vaccines and discover how to better support the needs of healthcare systems after large-scale events. We can also see from these examples that the use of artificial intelligence in the medical industry is very different from other industries and requires a different approach.
Consequently, both technology leaders and interviewees from established enterprises cited the availability of healthcare-specific models and algorithms as the most important factor when evaluating locally installed software libraries or SaaS solutions. requirements, which is no surprise. Judging by the venture capital landscape, existing libraries on the market, and demand from AI users, healthcare-specific models will only grow increasingly in the coming years.
With all the progress artificial intelligence has made in the past year, it has also opened up A new set of attack vectors. When asked what types of software respondents use to build their AI applications, the most popular choices are locally installed commercial software (37%) and open source software (35%). Most notably, usage of cloud services fell by 12% (30%) compared to last year's survey, most likely due to privacy concerns over data sharing.
Additionally, the majority of respondents (53%) choose to rely on their own data to validate models rather than rely on third-party or software vendor metrics. Respondents from established companies (68%) indicated a clear preference for using internal assessments and self-tuning models. Likewise, due to the strict controls and procedures around medical data processing, it is clear that AI users will want to keep operations in-house whenever possible.
But regardless of software preferences or how users validate models, escalating healthcare security threats can have a significant impact. While other critical infrastructure services face challenges, the impact of breaches in healthcare extends beyond reputational and financial losses. Data loss or tampering with hospital equipment can mean the difference between life and death.
Artificial intelligence is poised for even more significant growth as developers and investors work to put the technology into the hands of everyday users. However, as AI becomes more widely available and models and tools improve, safety, security and ethics will become a critical focus. It will be interesting to see how AI in these healthcare fields develops this year and what it means for the future of the industry.
Source: www.cio.com
The above is the detailed content of Three major trends in future medical AI worth paying attention to. For more information, please follow other related articles on the PHP Chinese website!