Artificial Intelligence

Dr. Robot, please report to the OR

Trembling hands? Error message. Robot-assisted systems with artificial intelligence support medical treatment that physicians cannot do on their own: minimally invasive procedures, precise navigation within the body – and even diagnoses.

12 Jan. 2017
Dr. Roboter AI
Dr. Robot, please report to the OR. (Photo: iStock)

People's health is a serious business, where the tolerance for error should approach zero. This is certainly one reason that robotic assistance systems have found their way into more and more operating rooms worldwide in recent years, because robots and artificial intelligence can help doctors, nurses and care providers perform at their very best.

The most commonly used surgical system using robotic engineering today is DaVinci, which allows doctors to conduct minimally invasive surgeries from a computer screen. DaVinci eliminates unintended movements, such as a tremble of the surgeon's hand. This surgical robot cannot be programmed, however. It does not carry out autonomous movements – let alone think for itself.

Computer as controller

But this might change in the future. Particularly in connection with computer-generated animations and 3D images, it would be possible at least in theory for robots to independently conduct various procedures or support the surgeon in their work. Example: A CT scan measures an organ being attacked by a tumor.

The calculated three-dimensional image serves as a guide during the procedure, but also as a control template. Cameras in the OR then monitor the exact position of the surgical instruments, and sensors in the instruments convey continuous data to a central computer. If the surgeon sets up an incision incorrectly with the robot arm, the system immediately halts . And patients are automatically protected from errors.

Artificial intelligence versus scars

Healthcare not only seeks better results from surgeries. Rapid convalescence is also important. The less intrusion on the body, the faster the healing. "Operations in cavities, in the stomach or chest area, are gradually no longer being done with big incisions, but using minimally invasive robot-assisted systems," says Klaus-Peter Jünemann , Director of Kiel University Hospital, where robot-assisted surgery is already practiced. When patients are not being “cut open” as much, wound healing becomes much less of a problem, and it limits blood loss and boosts precision, says Jünemann.

At the same time, doctors and other experts are quick to point out that robots will always remain mere assistants, and people always oversee the operations. This is also and especially true for diagnosis. But that could also change in the future. Assessing clinical symptoms is only evidence-based – derived from known cases – about 20 percent of the time. That means that one in five diagnoses is wrong or incomplete. There are nearly 1.5 million medication errors made annually in the United States alone.

Robots as superior physicians?

Artificial intelligence using big data is able – unlike the human practitioner – to compare huge volumes of information and draw conclusions from the results. This is how the Watson super computer was able to diagnose a rare form of blood cancer in a patient , which had not yet been identified by her physicians. Watson reached its result after comparing the woman's generic information with data sets from 20 million clinical cancer studies.

Doctors' ability to make correct diagnoses could improve significantly with increasing networking and the growth of artificial intelligence applications. This would make it possible to combine current examination and diagnostic imaging results with data from all available sources: treatment guidelines, electronic medical records, notes from doctors and care providers, research results, clinical studies and articles in medical journals.

Watson is now used by a clinic in Marburg to assist physicians in diagnosing rare or unknown diseases. Why? More than 3,000 scientific studies are published every month around the world in the field of medicine. No doctor can know them all – but Watson can. The system uses this data to generate a list of possible diagnoses, each assigned a value corresponding to the likelihood of the hypothesis. The final decision lies with the doctor, of course.

How is digitization impacting the healthcare sector? What policy decisions are being made? Can global networking make it possible to access the know-how of specialists from around the world? What do doctors need to be able to use the new technologies? These are just some of the questions being explored at the CeBIT Digital Summit "d!conomy Healthcare" on 22 March. Along with keynote presentations, opportunities, ideas, challenges, knowledge-sharing and best practices are also explored in smaller workshops.

Artificial Intelligence CEBIT RSS Feed