Ultrasound AI: A look into AI medical imaging

With its ability to solve complex challenges, Artificial Intelligence (AI) is one of the key areas that’s revolutionizing medical practice. Across the US, applications of AI in healthcare, particularly AI in ultrasound, are increasing at a rapid pace and according to some estimates, the investment in healthcare AI is expected to reach a staggering $6.6 billion by 2021.^1,2

Amongst the most well-known challenges found within healthcare we can include; high costs, wastage of investments (25% of US healthcare spending is wasted),³ operational inefficiencies, and growing pressure to manage an ever-increasing volume of patient and staff information while also ensuring data security standards.⁴ Yet, at the center of all this, there is one vital concern – how can patient care be optimized and patient outcome be improved?

This is where the use of AI in ultrasound comes into play. Here, AI-based solutions are being implemented to improve patient outcomes through increasing accuracy of diagnoses and supporting in decision making. The use of deep learning in medical imaging, for instance, helps the patient get a seamless, integrated and personalized experience in their treatment journey.

Here, we’ll explore in detail the ways in which AI medical imaging and ultrasound innovations can improve both staff’s workflow and daily routine as well as – and most importantly – patient diagnosis and outcome.

Improving patient diagnosis through AI medical imaging

At present, automated medical image diagnosis is arguably the most successful use of AI in the healthcare sector.⁵ In fact, AI medical imaging or ultrasound AI is one of the areas where deep learning has already started to shape the healthcare landscape.

Why? Because deep learning medical imaging has been paramount in allowing ultrasound solutions to become more accessible, regardless of location and user's knowledge. This means that in an emergency setting, AI medical imaging allows ultrasound users with basic scanning skills to produce precise diagnoses.⁶

A notable example of ultrasound AI innovation being used to improve patient diagnosis and care can be found in cardiology ultrasound. 3D Auto RV – right ventricular quantification is an application that combines proven TOMTEC experience in RV analysis software with Philips machine learning artificial intelligence.⁷ The benefit? Its AI-based auto-segmentation mode reduces the need for manual image adjustment. As a result, assessments become faster, more reproducible and more accurate.

AI ultrasound tackling efficiency demands

Optimization is one of the key motivations for the widespread use of AI medical imaging. Where time is of the essence, unsurprisingly, the main role of artificial intelligence in ultrasound imaging becomes that of supporting ultrasound users by automating time-consuming tasks.

The use of AI in ultrasound helps improve efficiency issues ⁸ sonographers meet in their daily routine through the:

Facilitation of patient scheduling.
Prevention of unnecessary intervention through automating accurate data input.
Reliable curation of data.⁹
Effective normalization and annotation of curated data.
Prognosis and patient outcome prediction based on the patient’s medical data history¹⁰ curated consistently and stored safely.
Reduction of time spent on administrative tasks allowing the most cost-time-efficient use of medical staff.
Accessibility and usability of medical equipment. (With AI, equipment and software can be used easily and quickly without depending on a specialized operator.

As these and other AI healthcare applications develop, there will be legal issues to consider around patient privacy. So how do we deal with those?

Artificial intelligence in healthcare applications and legal issues

Despite the potential for much-needed efficiency, there is still some anxiety surrounding artificial intelligence in healthcare applications and their legal issues and repercussions. Cybersecurity and safe data management are a source of concern, especially with the present lack of regulation and monitoring surrounding AI usage.

However, it’s important to note that as support for AI-based applications in healthcare, especially in AI ultrasound, grows exponentially, medical regulatory bodies increasingly factor in artificial intelligence in their regulations. Suppliers, in turn, work alongside them, designing more secure systems able to protect data effectively. IntelliSpace Portal, for instance, is powered with a robust architecture precisely designed to protect patient data.

Looking forward, there is a bright future where ultrasound AI solutions can continue to improve patient care worldwide as healthcare providers, regulatory bodies, and suppliers come together to ensure that artificial intelligence is employed safely and effectively.

Sources

¹https://www.forbes.com/sites/insights-intelai/2019/02/11/ai-and-healthcare-a-giant-opportunity/#6edf9b574c68. Last accessed Oct 2019.

²https://benhamouglobalventures.com/2018/08/02/digital-transformation-of-healthcare-state-of-the-union/ Last accessed Oct 2019

³“How Artificial Intelligence is increasing in Radiology,” 2018, Philips Healthcare podcast.

⁴https://www.usa.philips.com/healthcare/nobounds/5-tips-for-managing-healthcare-cybersecurity-risk Last accessed Oct 2019

⁵ Yu, K.-H., Beam, A. L., & Kohane, I. S. Artificial intelligence in healthcare. Nature Biomedical Engineering, 2(10), 719–731 (2018).

⁶ Artificial Intelligence within Ultrasound. https://www.signifyresearch.net/medical-imaging/artificial-intelligence-within-ultrasound/. (2018) Last accessed Oct 2019

⁷ https://www.philips.com/c-dam/b2bhc/master/landing-pages/cvx/epiq-cvx-3d-auto-rv-white-paper.pdf

⁸ Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56. doi:10.1038/s41591-018-0300-7

⁹ “How Artificial Intelligence is increasing in Radiology,” 2018, Philips Healthcare podcast.

¹⁰ https://svn.bmj.com/content/2/4/230