Scientists Design AI That Can Forecast Lifespan of Individuals

Scientists Design AI That Can Forecast Lifespan of Individuals

Scientists have designed a novel AI system that can forecast how long an individual will be alive just by seeing at the pictures of their organs.

The system, designed by experts in Australia from University of Adelaide, analyzed the medical imaging of chests of 48 patients and was successful to forecast which of them would expire in a timeframe of 5 years, with 69% correctness.

This is analogous to ‘manual’ forecasts by clinicians, said researchers.

The report, published in the Scientific Reports journal, has suggestions for the premature analysis of medical intervention and serious illness.

Scientists Design AI That Can Forecast Lifespan of Individuals

“Forecasting the potential of a patient is useful since it may allow doctors to modify treatments to the person,” claimed PhD student at the University of Adelaide, Luke Oakden-Rayner.

“The precise evaluation of biological age and the forecast of longevity of patient have thus far been restricted by inability of doctors to see within the body and calculate the health of every organ,” stated Oakden-Rayner.

“Our study has examined the employment of ‘deep learning’, a method where computer systems can study how to analyze and understand images,” he claimed.

“Even though for this research only a tiny sample of patients was utilized, our study proposes that the computer has learnt to recognize the composite imaging emergences of diseases, something that needs broad training for human specialists,” he said.

While the experts could not recognize precisely what the computer system was looking in the images to generate its forecasts, the most certain forecasts were made for patients with brutal chronic diseases such as congestive heart failure and emphysema.

“Rather than aiming on diagnosis of diseases, the automatic systems can forecast medical results in such a way, for which the doctors are not skilled to do, by detecting subtle patterns and fitting in huge volumes of data,” Oakden-Rayner stated.

“Our study opens new gates for the application of AI technology in analysis of medical image, and can provide new ray of hope for the premature detection of sober illness, needing particular medical involvements,” he said.

The researchers expect to relate the same methods to forecast other significant medical circumstances, such as the beginning of heart attacks.

Be the first to comment

Leave a Reply

Your email address will not be published.


*