In a paper1 published on 7 January in Nature Medicine, researchers describe thetechnology behind the diagnostic aid, a smartphone app called Face2Gene. Itrelies on machine-learning algorithms and brain-like neural networks toclassify distinctive facial features in photos of people with congenital andneurodevelopmental disorders.
Using the patterns that it (the app) infers from the pictures, the modelhomes inon possible diagnoses and provides a list of likely options. Doctors have beenusing the technology as an aid, even though it's not intended to provide definitivediagnoses.
1. model 算法；模型
2. home in on 匹配；定位
Researchers at FDNA, a digital-health company in Boston, Massachusetts, first trained the artificial intelligence (AI) system to distinguish Cornelia de Lange syndrome and Angelman syndrome — two conditions with distinct facial features — from other similar conditions. They also taught the model to classify different genetic forms of a third disorder known as Noonan syndrome.
Then the researchers, led by FDNA chief technology officer Yaron Gurovich, fedthe algorithm more than 17,000 images of diagnosed cases spanning 216 distinctsyndromes. When presented with new images of people’s faces, the app’s bestdiagnostic guess was correct in about 65% of cases. And when considering multiple predictions, Face2Gene's top-ten list contained the right diagnosis about90% of the time.
But the algorithm is only as good as its training data set.
— Silos and Bias
It does raise a number of ethical and legal concerns, say researchers. Theseinclude ethnic bias in training data sets and the commercial fragmentation ofdatabases, both of which could limit the reach of the diagnostic tool.The algorithm is only as good as its training data set — and there’s a risk,especially where rare disorders that affect only small numbers of people worldwideare concerned, that companies and researchers will begin to silo and commodifytheir data sets.And ethnic bias in training data sets that contain mostly Caucasian faces remains aconcern. A 2017 study of children with an intellectual disability found that whereasFace2Gene’s recognition rate for Down syndrome was 80% among white Belgianchildren, it was just 37% for black Congolese children. With a more-diversetraining data set, however, the algorithm’s accuracy for African faces improved,showing that more-equitable representation of diverse populations is achievable.