The data quality in China's health care sector needs to be improved in order to apply artificial intelligence (AI) or deep learning technologies, said an expert at a panel discussion here on November 14.
China has a large amount of data in the health sector but the quality was not good enough for the AI technology, said Wang Fei, assistant professor on health data mining and machine learning with Weill Cornell Medicine of Cornell University, adding that "the ground is still rough."
For instance, different coding systems like International Classification of Disease (ICD)-9, ICD-10, and even self-defined ones are used in diagnosing diabetes among different Chinese hospitals, Wang said at the panel discussion on China's health care revolution organized by China Institute.
"Unlike other domains where the data are clean and well-structured, health care data are highly heterogeneous, ambiguous, noisy and incomplete." noted a paper by Wang and others published in May 2017 in the scientific journal Briefings in Bioinformatics.
The application of AI technologies in health care industry is in an infant stage compared with that in automatic driving, according to Wang.
Wang said that a lot of caution shall be paid as AI technologies can't do well everywhere though they have huge potential in a lot of places.
Deep learning could be used in portfolio imaging, lung scanning, drug designing and other fields in health care sector, said Wang.
China is seen enjoying an advantage in developing AI technologies due to the abundance of as well as easier and cheaper access to data.
(Xinhua News Agency November 15, 2018)