TY - JOUR
T1 - Cancer omic data based explainable AI drug recommendation inference
T2 - A traceability perspective for explainability
AU - Xi, Jianing
AU - Wang, Dan
AU - Yang, Xuebing
AU - Zhang, Wensheng
AU - Huang, Qinghua
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2023/1
Y1 - 2023/1
N2 - The application of Artificial Intelligence (AI) on cancer drug recommendation can prompt the development of personalized cancer therapy. However, most of the current AI drug recommendations cannot give explainable inferences, where their prediction procedures are black boxes, and are difficult to earn the trust of doctors or patients. In explainable inference, the key steps during the recommendation procedures can be located easily, facilitating model adjustment for wrong predictions and model generalization for new drugs/samples. In this paper, we analyze the necessity of developing explainable AI drug recommendation, and propose an evaluation metric called traceability rate. The traceability rate is calculated as the proportion of correct predictions that are traceable along the knowledge graph in all the ground truths. We further conduct an experiment on a benchmark drug response dataset to apply the traceability rate as evaluation metric, where the results show a trade-off between model performance and explainability. Therefore, the explainable AI drug recommendation still demands for further improvement to meet the requirement of clinical personalized therapy.
AB - The application of Artificial Intelligence (AI) on cancer drug recommendation can prompt the development of personalized cancer therapy. However, most of the current AI drug recommendations cannot give explainable inferences, where their prediction procedures are black boxes, and are difficult to earn the trust of doctors or patients. In explainable inference, the key steps during the recommendation procedures can be located easily, facilitating model adjustment for wrong predictions and model generalization for new drugs/samples. In this paper, we analyze the necessity of developing explainable AI drug recommendation, and propose an evaluation metric called traceability rate. The traceability rate is calculated as the proportion of correct predictions that are traceable along the knowledge graph in all the ground truths. We further conduct an experiment on a benchmark drug response dataset to apply the traceability rate as evaluation metric, where the results show a trade-off between model performance and explainability. Therefore, the explainable AI drug recommendation still demands for further improvement to meet the requirement of clinical personalized therapy.
KW - Drug recommendation
KW - Explainability
KW - Omic data
KW - Traceability
UR - http://www.scopus.com/inward/record.url?scp=85138470504&partnerID=8YFLogxK
U2 - 10.1016/j.bspc.2022.104144
DO - 10.1016/j.bspc.2022.104144
M3 - 文章
AN - SCOPUS:85138470504
SN - 1746-8094
VL - 79
JO - Biomedical Signal Processing and Control
JF - Biomedical Signal Processing and Control
M1 - 104144
ER -