Artificial intelligence and convolution neural networks assessing mammographic images: a narrative literature review.

Studies have shown that the use of artificial intelligence can reduce errors in medical image assessment. The diagnosis of breast cancer is an essential task; however, diagnosis can include 'detection' and 'interpretation' errors. Studies to reduce these errors have shown the feasibility of using convolution neural networks (CNNs). This narrative review presents recent studies in diagnosing mammographic malignancy investigating the accuracy and reliability of these CNNs. Databases including ScienceDirect, PubMed, MEDLINE, British Medical Journal and Medscape were searched using the terms 'convolutional neural network or artificial intelligence', 'breast neoplasms [MeSH] or breast cancer or breast carcinoma' and 'mammography [MeSH Terms]'. Articles collected were screened under the inclusion and exclusion criteria, accounting for the publication date and exclusive use of mammography images, and included only literature in English. After extracting data, results were compared and discussed. This review included 33 studies and identified four recurring categories of studies: the differentiation of benign and malignant masses, the localisation of masses, cancer-containing and cancer-free breast tissue differentiation and breast classification based on breast density. CNN's application in detecting malignancy in mammography appears promising but requires further standardised investigations before potentially becoming an integral part of the diagnostic routine in mammography.

Model-Guided Interface Probe Arrangement for Sensitive Protein Detection.

A wide range of analytical techniques in bioanalysis relies on surface-based biomolecular detection, which requires the confinement of probes onto a heterogeneous surface to react with targets. Probe arrangement on the interface is critical for target recognition and determines assay performance. Much effort has been devoted to screen the optimized probe arrangement according to experimental tests. Such a data-driven posteriori pattern faces low efficiency, ambiguous orientation, and possible deviated tested ranges from the best case. Herein, we demonstrate that a model can effectively guide probe arrangement onto the interface to facilitate probe-target recognition, embodied by the assay of human telomerase activity with DNA-conjugated gold nanoparticles (AuNPs). Both theoretical calculation and experimental results indicate that telomerase activity is maximized on the AuNP surface under guidance of the model. The detection limit is at least 1 order of magnitude lower than that of AuNP bearing densely packed DNA, comparable to that of the telomeric repeat amplification protocol (TRAP). The model-guided interface probe arrangement is proved to be highly useful in regulating interface recognition and may offer a new paradigm to promote surface-based biomolecular detection.