ABSTRACT
Machine-assisted pathological recognition has been focused on supervised learning (SL) that suffers from a significant annotation bottleneck. We propose a semi-supervised learning (SSL) method based on the mean teacher architecture using 13,111 whole slide images of colorectal cancer from 8803 subjects from 13 independent centers. SSL (~3150 labeled, ~40,950 unlabeled; ~6300 labeled, ~37,800 unlabeled patches) performs significantly better than the SL. No significant difference is found between SSL (~6300 labeled, ~37,800 unlabeled) and SL (~44,100 labeled) at patch-level diagnoses (area under the curve (AUC): 0.980 ± 0.014 vs. 0.987 ± 0.008, P value = 0.134) and patient-level diagnoses (AUC: 0.974 ± 0.013 vs. 0.980 ± 0.010, P value = 0.117), which is close to human pathologists (average AUC: 0.969). The evaluation on 15,000 lung and 294,912 lymph node images also confirm SSL can achieve similar performance as that of SL with massive annotations. SSL dramatically reduces the annotations, which has great potential to effectively build expert-level pathological artificial intelligence platforms in practice.
Subject(s)
Artificial Intelligence/standards , Colorectal Neoplasms/pathology , Deep Learning/standards , Lung Neoplasms/pathology , Supervised Machine Learning/standards , Colorectal Neoplasms/classification , Colorectal Neoplasms/diagnostic imaging , Humans , Lung Neoplasms/classification , Lung Neoplasms/diagnostic imaging , Lymphatic Metastasis , Neural Networks, Computer , ROC CurveABSTRACT
The widespread application of nanomaterials (NMs), which has accompanied advances in nanotechnology, has increased their chances of entering an organism, for example, via the respiratory system, skin absorption or intravenous injection. Although accumulating experimental evidence has indicated the important role of NM-biomembrane interaction in these processes, the underlying mechanisms remain unclear. Computational techniques, as an alternative to experimental efforts, are effective tools to simulate complicated biological behaviors. Computer simulations can investigate NM-biomembrane interactions at the nanoscale, providing fundamental insights into dynamic processes that are challenging to experimental observation. This paper reviews the current understanding of NM-biomembrane interactions, and existing mathematical and numerical modeling methods. We highlight the advantages and limitations of each method, and also discuss the future perspectives in this field. Better understanding of NM-biomembrane interactions can benefit various fields, including nanomedicine and diagnosis.
