AI making its way to improve results and efficiency in Cardiovascular Imaging

Artificial intelligence is showing reassuring results in cardiology, especially in the area of cardiovascular imaging. Ever wondered how it works? 

Machine learning algorithms (a subdivision of AI) are making it possible for cardiologists to find out new opportunities and delve into new discoveries difficult to be noticed using conventional techniques. This offers newer gateways helpful in medical decision-making.

Key features of AI in Cardiovascular Imaging:

  • AI can help improve performance at low cost thereby facilitating decision making, interpretation, and precise image acquisition of anatomical structures as well as diagnosis.
  • The big data obtained using imaging will be helpful in personalizing medical treatments and keeping an electronic record of patient data, health records, and outcome data.
  • It is believed that this will help physicians work more efficiently on core issues while computers will handle the technical part.

 

Different kinds of imagining possibilities using AI

Echocardiography – This is the most commonly used imaging technique in cardiology, but it is highly user-dependent. It is also important to undergo serious training in order to interpret the results accurately. AI can be a low-cost alternative to provide a standardized analysis of echocardiographic images. It has already shown great success in this area.

Computed Tomography – AI has been greatly appreciated in the field of cardiac CT as it has helped in noise reduction, and image optimization thereby preventing invasive coronary angiography (ICA) in the identification of severe stenosis.

Magnetic Resonance Imaging – This includes anatomical images of various aspects of the heart including flow imaging, contractile function, perfusion imaging, and myocardial characterization. As is the case with Echocardiography, MRI is also highly user-dependent. Reports have shown that implementing computer-aided detection in the clinical setup can increase accuracy and simplify the analysis.

Nuclear Imaging of the heart – This is performed to assess any perfusion defects within the myocardial lining. AI-based models can improve the clinical value of the results obtained. AI-based models have been highly successful at detecting abnormal myocardium in CAD, and their efficiency is at par with manual analysis images received.

Conclusion  

Technology is not new to humans. We’re getting more and more comfortable with the idea of relying on machines for safer and more accurate conduct of our day to day lives. In the case of cardiovascular imaging, AI has proved itself promising in various ways. Here are the reasons why the medical industry is ready to adopt computer-aided detection and diagnosis in Cardiology:

  • Detection and diagnosis of disease
  • Interpretation of data
  • Collection and comparison of data for future studies
  • Clinical decision making
  • Accurate image acquisition of images
  • Reducing health care expenditure
  • Reducing the workload for physicians

 

AI seems to have a lot of pros for the medical industry, but it needs to be made perfect with more testing and re-testing. It will definitely be a great tool for cardiologists as it is capable of recognizing patterns that are otherwise difficult to assess for the human brain.

World’s first AI University at Abu Dhabi offers 100% scholarships

Abu Dhabi has just launched the “world’s first AI university” in the city of in Masdar City known as the Mohammad Bin Zayed University of Artifical Intelligence (MBZUAI). Abu Dhabi is also one of the first nations to appoint a minister for artificial intelligence in 2017.

MBZUAI’s website mentions that “a graduate-level, research-based academic institution that offers specialized degree programs for local and international students in the field of Artificial Intelligence.” Shaikh Mohammad Bin Zayed Al Nahyan, Abu Dhabi’s crown prince tweeted “Launching of the world’s first graduate-level artificial intelligence university in Abu Dhabi echoes the UAE’s pioneering spirit, and paves the way towards a new era of innovation and technological advancement that benefits the UAE and a world.”

Reference Link: https://www.theweek.in/news/sci-tech/2019/10/16/uae-launches-worlds-first-ai-university-with-100-scholarships.html

AI and scientific literature work in sync

When scholars choose a topic to work on their research, they need more sources or materials to review literature and add more value to their findings. According to Canadian science publishing’s article from last year, 2.5 million research papers are published annually while another unidentified source suggests that new researches are published around the world; approximately 1 million each year! Which is equal to one every 30 seconds. With the overload of new papers in each field and more growing every year it is practically impossible for scholars to keep with the information that is put out in each paper. Christian Berger’s team from the University of Gothenburg in Sweden, found a staggering number of papers on the subject; more than 10,000 in the same subject. Fortunately, the team had the support of an AI system, a writing investigation tool called Iris.ai.

Iris.ai is an AI, a tool developed for scholars to make writing research papers easier. It is a Berlin-based company that claims to save 90% of time with 85% precision of data matching, has more than 70 m open access papers. Iris.ai is programmed to learn about the topic provided and perform an elaborate frequency analysis over the text. Then it read the words for which it needs to find results and additional material that could be helpful for the paper. It uses a 500-word description of the researcher’s issue, or the link of their paper and the AI restores a guide to thousands of coordinating reports. As the website suggests, it is a scientific writing assistant.

According to Berger, it was “a quick and nevertheless precise overview of what should be relevant to a certain research question”. Iris.ai is one among many of the new AI-based tools offering targeted results of the knowledge landscape. One such tool is called Semantic Scholar, produced by the Allen Institute for Artificial Intelligence in Seattle, Washington, and Microsoft Academic.

Although every instrument is different from each other and gives different output, they all provide researchers with a different look at the scientific literature than conventional tools such as PubMed and Google Scholar. Semantic Scholar is a browser-based search tool that mimics the engines like Google and it is free. But it is more informative than Google Scholar in terms of specific results required by researchers. Doug Raymond, Semantic Scholar’s general manager, says that one million individuals utilize their service every month. It uses natural language processing or NLP to extract data while building connections to determine if the information is relevant and reputable or not.

Artificial intelligence is saving a lot of time and making it easier and quicker to automate some procedures. In the academic publishing industry, the Al-based innovations are being produced and implemented to help both authors and publishers for peer reviews, searching published content, detecting plagiarism, and identifying data fabrication. AI could be costly, but it can accelerate a researchers’ access to new fields. More and more such AI tools are being developed to cater to various requisites of writing a paper, such as filtering topics for relevance, keyword search, etc.

Experts who need more assistance for their specific concerns might consider free Al­ tool such as Microsoft Academic or Semantic Scholar. While AI is easing so many burdens and saving time for a researcher, let’s not forget that it is still machine intelligence and may require human intervention here and there to make a paper more presentable and precise.