Deep learning algorithms have been able to diagnose the presence or absence of tuberculosis (TB) in chest x-ray images with astonishing accuracy. Researchers first “trained” the AIs with hundreds of X-ray images of patients without and with tuberculosis. Then, they tested the AIs with 150 new x-rays. The algorithms achieved an impressive 96% accuracy rate — better than many human radiologists — and the researchers believe they can improve upon this with more training cases and more advanced deep learning models. As the study authors note, “Automated detection of pulmonary TB at chest radiography may facilitate screening and evaluation efforts in TB-prevalent areas with limited access to radiologists.”
The world’s leading drug companies are turning to artificial intelligence to improve the hit-and-miss business of finding new medicines, with Glaxo Smith Kline unveiling a new $43 million deal in the field on Sunday.
GSK’s deal with Exscientia will see the Scottish artificial intelligence pioneer discover new small molecules against targets selected by the big pharma. Exscientia is already working with Sanofi, Evotec, Sumitomo Dainippon Pharma and Sunovion across a range of disease areas, and is in talks with others.
The aim is to harness modern supercomputers and machine learning systems to predict how molecules will behave and how likely they are to make a useful drug, thereby saving time and money on unnecessary tests.
AI systems play a central role in other high-tech areas such as the development of driverless cars and facial recognition software.
Lower Cost
Hopkins, who used to work at Pfizer, said Exscientia’s AI system could deliver drug candidates in roughly one-quarter of the time and at one-quarter of the cost of traditional approaches.
The Scotland-based company, which also signed a deal with Sanofi in May, is one of a growing number of start-ups on both sides of the Atlantic that are applying AI to drug research. Others include U.S. firms Berg, Numerate, twoXAR and Atoms, as well as Britain’s BenevolentAI.
“In pharma’s eyes these companies are essentially digital biotechs that they can strike partnerships with and which help feed the pipeline,” said Nooman Haque, head of life sciences at Silicon Valley Bank in London. “If this technology really proves itself, you may start to see M&A with pharma, and closer integration of these AI engines into pharma R&D.”
It is not the first time drug makers has turned to high-tech solutions to boost R&D productivity. The introduction of “high-throughput screening”, using robots to rapidly test millions of compounds, generated mountains of leads in the early 2000s but notably failed to solve inefficiencies in the research process.
When it comes to AI, big pharma is treading cautiously, in the knowledge that the technology has yet to demonstrate it can successfully bring a new molecule from computer screen to the lab to the clinic and finally to market.
