An artificial intelligence (AI) machine developed by a Singapore firm is making the tedious process of diagnosing a brain tumour easier and less prone to human error.
The machine can do in a second what an experienced doctor would take 30 minutes to complete, given the thousands of magnetic resonance imaging (MRI) images that need to be checked.
Cancer cells also have to be identified and a report written.
It is laborious work and it is not uncommon for less experienced doctors to make mistakes.
But Biomind, an "AI doctor" developed by Singapore start-up Hanalytics, eliminates much of that.
The machine will be on the market at the end of this month, said chief executive Raymond Moh.
It was developed at a research centre jointly run by Hanalytics and Beijing Tiantan Hospital, where it has been used on patients over the past two months. China is trying to become a world leader in AI and has been moving aggressively to adopt it in the healthcare sector.
Hanalytics said it is the first AI machine that can diagnose a suite of neurological conditions, from brain tumours to vascular issues.
Biomind can pick up neurological conditions with an accuracy of over 90 per cent, higher than the 60 per cent by top doctors at the Beijing hospital, noted Mr Moh. The machine is linked to the hospital's MRI and CT scanners, so doctors can opt to have it process brain scans.
About Tiantan hospital and Hanalytics
Beijing Tiantan Hospital is known in China for its expertise in neurology and neurosurgery. The hospital performs about 10,000 neurosurgeries each year and houses the country’s only national clinical research centre for neurological diseases. In December last year (2017), the hospital, which was established in 1956, set up a joint AI research centre for neurology with Singapore artificial intelligence (AI) company Hanalytics in Beijing. Both parties said the centre was the first of its kind in the world, and would develop new AI technologies to improve diagnosis, prognosis and patient rehabilitation. Biomind is the first technology the centre has developed. Hanalytics, which began early last year, works with government-linked organisations in Singapore, including the Ministry of Health Holdings (MOHH) and the National Heart Centre. Chief executive Raymond Moh says the company is funded by a group of angel investors. It recently launched a PhD scholarship programme for engineering and computer science with the National University of Singapore, where students will be attached to the company’s research centres.
A single scan contains over 3,000 images. "The AI is able to identify where the condition is, what it is, and churn out a report - all in an instant," said Mr Moh.
The AI makes these decisions after being "trained" on a bank of over two decades' worth of MRI and CT records of the hospital's past diagnoses.
It also learns from its mistakes. If it makes a wrong diagnosis, a senior doctor can correct it and this correction is stored in the database so it does not make the same error in future, said Mr Moh.
The company is so confident of Biomind's capabilities that it is planning to pit the AI computer against top neurologists at a competition at a neurology conference in Beijing at the end of this month.
Globally, tech firms are racing to apply AI to healthcare. DeepMind, a subsidiary of Google's parent company Alphabet, for example, is training an AI algorithm to diagnose eye diseases.
Mr Moh told The Straits Times that Hanalytics decided to focus on the brain even though it is the most complex part of the body because "a big bulk of body scans, CT and MRI, deal with the brain".
Apart from Beijing Tiantan Hospital's renowned neurology expertise, the hospital was picked for the trial because of the sheer number of patients it sees - it does about 10,000 neurosurgery procedures a year.
Dr Gao Peiyi, a neuroradiologist, said Biomind will allow doctors to be more productive. It also provides more consistent diagnoses, he said.
"It does not get tired... Human doctors are affected by emotions, their physical state, time and other external factors, and this can affect the accuracy of their diagnoses."