VIDEO: Big Data, self-diagnosis to play role in future of RA, expert says

SANDESTIN, FLA. – Big data informing patient treatment, computer algorithms reading imaging instead of humans, and even accurate patient self-diagnosis could emerge over the next 10 years in the treatment of rheumatoid arthritis, an expert said at the annual Congress of Clinical Rheumatology.

Gerd Burmester, MD, director of rheumatology and clinical immunology at Charité University in Berlin, trotted out staggering numbers on future medical data collection on patients. Data analytics companies project that more than 1,000 terabytes of data per lifetime is expected to be gathered, with just 10% expected to be clinical information and 30% in the form of “-omics,” such as proteomics and genomics, he said. The other 60% is expected to come from sensors and wearables that patients essentially collect themselves with their own devices, he said.

“We will have to use data in the interest of the patient,” he said. “This is the real secret. In order to do this, we need cognitive computing, which assesses structured and unstructured data and is self-learning.”

The days of images being read by human radiologists could be numbered, he said.

“There will be a revolution in imaging scoring,” he said, with computer algorithms generating scores, more quickly separating the normal scans from those that need clinical attention.

He described a possible scenario in which patients get genetic analyses, blood biomarker testing, and imaging performed at kiosks about town, producing a diagnosis without a single physician visit. It might seem fanciful, but when he asked the audience how many thought it was impossible over the next decade, no one raised a hand.

With advances such as the self-rheumatoid arthritis examination tool Rheuma-Check and the decline in cost for whole genome sequencing – along with wait times to see rheumatologists sometimes as long as 6 months – such a scenario might not be far fetched, Dr. Burmester said. It is possible, he said, because patient histories that used to sit in charts, images that used to be on film only, and genetic data that used to be unavailable, are all now in structured, digital form.

Referring to a recent commentary in the New England Journal of Medicine, Dr. Burmester said physicians have to accept the coming role of computer algorithms.

“If medicine wishes to stay in control of its own future,” he said, “physicians will not only have to embrace algorithms, they will also have to excel at developing and evaluating them, bringing machine-learning methods into the medical domain.”

SOURCE: Burmester, G. CCR 2018.

SANDESTIN, FLA. – Big data informing patient treatment, computer algorithms reading imaging instead of humans, and even accurate patient self-diagnosis could emerge over the next 10 years in the treatment of rheumatoid arthritis, an expert said at the annual Congress of Clinical Rheumatology.

Gerd Burmester, MD, director of rheumatology and clinical immunology at Charité University in Berlin, trotted out staggering numbers on future medical data collection on patients. Data analytics companies project that more than 1,000 terabytes of data per lifetime is expected to be gathered, with just 10% expected to be clinical information and 30% in the form of “-omics,” such as proteomics and genomics, he said. The other 60% is expected to come from sensors and wearables that patients essentially collect themselves with their own devices, he said.

“We will have to use data in the interest of the patient,” he said. “This is the real secret. In order to do this, we need cognitive computing, which assesses structured and unstructured data and is self-learning.”

The days of images being read by human radiologists could be numbered, he said.

“There will be a revolution in imaging scoring,” he said, with computer algorithms generating scores, more quickly separating the normal scans from those that need clinical attention.

He described a possible scenario in which patients get genetic analyses, blood biomarker testing, and imaging performed at kiosks about town, producing a diagnosis without a single physician visit. It might seem fanciful, but when he asked the audience how many thought it was impossible over the next decade, no one raised a hand.

With advances such as the self-rheumatoid arthritis examination tool Rheuma-Check and the decline in cost for whole genome sequencing – along with wait times to see rheumatologists sometimes as long as 6 months – such a scenario might not be far fetched, Dr. Burmester said. It is possible, he said, because patient histories that used to sit in charts, images that used to be on film only, and genetic data that used to be unavailable, are all now in structured, digital form.

Referring to a recent commentary in the New England Journal of Medicine, Dr. Burmester said physicians have to accept the coming role of computer algorithms.

“If medicine wishes to stay in control of its own future,” he said, “physicians will not only have to embrace algorithms, they will also have to excel at developing and evaluating them, bringing machine-learning methods into the medical domain.”

SOURCE: Burmester, G. CCR 2018.

SANDESTIN, FLA. – Big data informing patient treatment, computer algorithms reading imaging instead of humans, and even accurate patient self-diagnosis could emerge over the next 10 years in the treatment of rheumatoid arthritis, an expert said at the annual Congress of Clinical Rheumatology.

Gerd Burmester, MD, director of rheumatology and clinical immunology at Charité University in Berlin, trotted out staggering numbers on future medical data collection on patients. Data analytics companies project that more than 1,000 terabytes of data per lifetime is expected to be gathered, with just 10% expected to be clinical information and 30% in the form of “-omics,” such as proteomics and genomics, he said. The other 60% is expected to come from sensors and wearables that patients essentially collect themselves with their own devices, he said.

“We will have to use data in the interest of the patient,” he said. “This is the real secret. In order to do this, we need cognitive computing, which assesses structured and unstructured data and is self-learning.”

The days of images being read by human radiologists could be numbered, he said.

“There will be a revolution in imaging scoring,” he said, with computer algorithms generating scores, more quickly separating the normal scans from those that need clinical attention.

He described a possible scenario in which patients get genetic analyses, blood biomarker testing, and imaging performed at kiosks about town, producing a diagnosis without a single physician visit. It might seem fanciful, but when he asked the audience how many thought it was impossible over the next decade, no one raised a hand.

With advances such as the self-rheumatoid arthritis examination tool Rheuma-Check and the decline in cost for whole genome sequencing – along with wait times to see rheumatologists sometimes as long as 6 months – such a scenario might not be far fetched, Dr. Burmester said. It is possible, he said, because patient histories that used to sit in charts, images that used to be on film only, and genetic data that used to be unavailable, are all now in structured, digital form.

Referring to a recent commentary in the New England Journal of Medicine, Dr. Burmester said physicians have to accept the coming role of computer algorithms.

“If medicine wishes to stay in control of its own future,” he said, “physicians will not only have to embrace algorithms, they will also have to excel at developing and evaluating them, bringing machine-learning methods into the medical domain.”

SOURCE: Burmester, G. CCR 2018.