"I would love to see a future where looking inside the body becomes as routine as a blood pressure cuff measurement," says Charles Cadieu '04, MEng '05. As president of the medical technology startup Caption Health, he sees that future in reach--with the help of artificial intelligence. Cadieu still remembers the "lightbulb moment" during his postdoctoral research at MIT when he realized that the field of AI would never be the same. He was working in the lab of James DiCarlo (now the Peter de Florez Professor of Neuroscience) on neural networks--AI systems made up of deep-learning algorithms that emulate the dense networks of neurons in the brain. Until then, neural networks had been unable to perform even simple visual tasks that the brain handles with ease.

The FDA has cleared what it describes as the first software that uses AI to guide family doctors, registered nurses and other clinicians in taking cardiac ultrasounds. Developed by Brisbane, California-based Caption Health, the software communicates instructions via prompts on a screen-based interface. The prompts allow non-experts to capture images and videos of diagnostic quality. "This is especially important because it demonstrates the potential for artificial intelligence and machine learning technologies to increase access to safe and effective cardiac diagnostics that can be life-saving for patients," Robert Ochs, a deputy director in the FDA's Center for Devices and Radiological Health, said in a statement. The software is called Caption Guidance and was cleared for use with a diagnostic ultrasound system developed by Teratech Corp., though the software has the potential to be used with other systems, according to the FDA. In granting clearance to the software, the agency said it looked at two independent studies.

Let's assume you are not an expert highly trained in medical imaging. And let's assume you were invited one day to try out a new technology for heart ultrasounds -- diagnostic tools that are notoriously difficult to use because of the chest wall and because some shots must be made while the heart is in motion. When I was given the shot on a recent day, I was able to take the ultrasound in a matter of minutes with the help of software, developed by a San Francisco-based startup called Caption Health. The software told me how to hold the ultrasound probe against the ribs of a model who had been hired for the purpose of my visit and knew on its own when to snap the image. It was a little like having Richard Avedon's knowledge of photography uploaded into the guts of my iPhone camera. You can see the image I took of the parasternal long axis view of the heart pumping at the top of this page.

Edwards Lifesciences is delving deeper into the realm of artificial intelligence through a partnership with San Francisco-based Bay Labs. The goal of the collaboration, which has multiple initiatives is to improve the detection of heart disease. Some of the initiatives include, the development of new AI-powered algorithms in Bay Labs' EchoMD measurement and interpretation software suite; support for ongoing clinical studies at institutions; and the integration of EchoMD algorithms into Edwards Lifesciences' CardioCare quality care navigation platform. Irvine, CA-based Edwards' CardioCare program combines clinical consulting expertise with a cloud-based platform to facilitate in the identification, referral, and care pathway management of patients with structural heart disease. CardioCare can help hospitals improve quality by reducing variability in echocardiography and ensure effective communication between care settings to ensure patients access to care.

For decades, neuroscientists have been trying to design computer networks that can mimic visual skills such as recognizing objects, which the human brain does very accurately and quickly. Until now, no computer model has been able to match the primate brain at visual object recognition during a brief glance. However, a new study from MIT neuroscientists has found that one of the latest generation of these so-called "deep neural networks" matches the primate brain. Because these networks are based on neuroscientists' current understanding of how the brain performs object recognition, the success of the latest networks suggest that neuroscientists have a fairly accurate grasp of how object recognition works, says James DiCarlo, a professor of neuroscience and head of MIT's Department of Brain and Cognitive Sciences and the senior author of a paper describing the study in the Dec. 18 issue of the journal PLoS Computational Biology. "The fact that the models predict the neural responses and the distances of objects in neural population space shows that these models encapsulate our current best understanding as to what is going on in this previously mysterious portion of the brain," says DiCarlo, who is also a member of MIT's McGovern Institute for Brain Research.

Bay Labs Inc. launched out of stealth mode today to use deep learning to help medical professionals in developing countries interpret ultrasounds so they can better treat heart disease. The company also announced a seed round of 2.5 million led by Khosla Ventures. Deep learning experts Yann LeCun, who serves as the director of AI Research at Facebook, and Nicolas Pinto and Jack Culpepper also invested. Bay Labs founder Charles Cadieu said ultrasound imaging presented the perfect problem to be solved with deep learning because it takes years of training to learn how to read ultrasound imaging. As ultrasound devices become cheaper and cheaper, they are increasingly available but especially in developing or rural areas, there may not be people trained to properly use the device or read the images.