Eye on AI - October 29th, 2021
Welcome to Aigora's "Eye on AI" series, where we round up exciting news at the intersection of consumer science and artificial intelligence!
Keeping in the spirit of Halloween, our focus this week is on seeing the unseeable, including a recently developed algorithm that may help spot the hidden signs of potential heart failure and a new technology that allows oceanographers to observe microscopic plankton in the deep ocean.
AI New Study Attempts to Spot Humanity’s Most Deadly Killer Before It Strikes
Humanity’s most deadly killer isn’t some ghoul lurking in the night. It’s heart disease, which, according to the World Health Organization, accounted for roughly 16% of human deaths since 2000. But a newly developed heart failure detection method developed by Mount Sinai School of Medicine researchers, as detailed in the Science Daily article “Scientists show how AI may spot unseen signs of heart failure”, may soon change that.
Until now, doctors relied heavily on an imaging technique called an echocardiogram to detect heart failure in patients. While helpful, echocardiograms are labor-intensive procedures only offered in select hospitals. However, recent breakthroughs in AI have suggested that electrocardiograms (ECG), a more widely available electrical recording device, may be a fast and readily available alternative.
“Although appealing, traditionally it has been challenging for physicians to use ECGs to diagnose heart failure,” says Girish N. Nadkarni, Associate Professor of Medicine at the Icahn School of Medicine at Mount Sinai and lead researcher in the study. “... This study represents an exciting step forward in finding information hidden within the ECG data which can lead to better screening and treatment paradigms using a relatively simple and widely available test.”
The study showed how a new AI algorithm was able to read the ECG images and written reports of patients together to assess the strength of their left and the right ventricles with a 94% accuracy. Though promising, this new method isn’t yet ready for widespread use. There were a few instances where the results weren’t as promising. But overall, the study suggests that this new algorithm may be a useful tool for helping clinical practitioners combat heart failure, which would make for a much healthier treat than the candy being handed out on the 31st.
New Technology Helps Oceanographers Observe Plankton in Deep Ocean
To conclude, let’s address another difficult-to-see phenomenon that new technology is shedding more light on: plankton! According to the article “Researchers develop AI-based technology for clearer plankton observation in deep ocean”, researchers at the Chinese Academy of Sciences have developed a novel image super-resolution method for digitally restoring resolutions of the hard-to-see deep ocean in situ plankton images.
“The researchers trained a deep learning model called Enhanced Deep Residual Network (EDSR) with a large-scale, real-world dataset called IsPlanktonSR,” writes the Chinese Academy of Sciences. “During the training, they tried different loss functions and compared the model performance by using traditional downsampled and IsPlanktonSR data sets. Through extensive experiments, the team has demonstrated that the model trained on real data through the contextual loss has delivered the best visual and quantitative SR performance.”
Plankton, which is microscopic sea organisms, are considered the base of the entire marine food web. They feed innumerable sea species, from fish to whales, and are considered by many scientists the keystone species of our waters. Recently, global warming has forced many plankton species to begin to relocate to colder waters, which could have a significant impact on marine life. The new imaging techniques developed by the Chinese Academy of Sciences will help researchers understand deep-sea plankton and better predict how their changes will affect sea life.
That's it for now. If you'd like to receive email updates from Aigora, including weekly video recaps of our blog activity, click on the button below to join our email list. Thanks for stopping by!