PSTCNN: Explainable COVID-19 diagnosis using PSO-guided self-tuning CNN

Since 2019, the coronavirus disease-19 (COVID-19) has been spreading rapidly worldwide, posing an unignorable threat to the global economy and human health. It is a disease caused by severe acute respiratory syndrome coronavirus 2, a single-stranded RNA virus of the genus Betacoronavirus. This virus is highly infectious and relies on its angiotensin-converting enzyme 2-receptor to enter cells. With the increase in the number of confirmed COVID-19 diagnoses, the difficulty of diagnosis due to the lack of global healthcare resources becomes increasingly apparent. Deep learning-based computer-aided diagnosis models with high generalisability can effectively alleviate this pressure. Hyperparameter tuning is essential in training such models and significantly impacts their final performance and training speed. However, traditional hyperparameter tuning methods are usually time-consuming and unstable. To solve this issue, we introduce Particle Swarm Optimisation to build a PSO-guided Self-Tuning Convolution Neural Network (PSTCNN), allowing the model to tune hyperparameters automatically. Therefore, the proposed approach can reduce human involvement. Also, the optimisation algorithm can select the combination of hyperparameters in a targeted manner, thus stably achieving a solution closer to the global optimum. Experimentally, the PSTCNN can obtain quite excellent results, with a sensitivity of 93.65% ± 1.86%, a specificity of 94.32% ± 2.07%, a precision of 94.30% ± 2.04%, an accuracy of 93.99% ± 1.78%, an F1-score of 93.97% ± 1.78%, Matthews Correlation Coefficient of 87.99% ± 3.56%, and Fowlkes-Mallows Index of 93.97% ± 1.78%. Our experiments demonstrate that compared to traditional methods, hyperparameter tuning of the model using an optimisation algorithm is faster and more effective.     

<i>In vitro</i> polymerization of the dopamine-borate melanin precursor: A proof-of-concept regarding boron neutron-capture therapy for melanoma

The ¹⁰boron neutron-capture therapy (BNCT) is an emerging antitumoral method that shows increasing biomedical interest. BNCT is based on the selective accumulation of the ¹⁰boron isotope within the tumor, which is then irradiated with low-energy thermal neutrons, generating nuclear fission that produces ⁷lithium, ⁴helium, and γ rays. Simple catechol-borate esters have been rather overlooked as precursors of melanin biosynthesis, and therefore, a proof-of-concept approach for using dopamine-borate (DABO) as a suitable boron-containing candidate for potential BNCT is presented here. DABO can spontaneously oxidize and autopolymerize in vitro, giving a soluble, eumelanin-like brown-black poly-DABO product. Melanotic melanoma cell cultures treated with 1 mM DABO for 24 and 48 h were viable and showed no signs of damage or cell death. The stability and possible trans-esterification of DABO is shortly discussed. Chemical calculations and quantitative structure-activity relationships (QSAR) analysis of DABO and the BNCT agent BPA indicated that they should be cell permeant and accumulate within lysosomes and melanosomes. Molecular modeling allows visualization of both the DABO precursor and the structure of a borate derivative of the proposed catechol-porphycene model for eumelanin, showing interesting features from molecular orbital calculations. The main difference between DABO and other agents, such as BPA, is that it is not a boronic acid nor a boron cluster. This simple catechol-borate ester (protected from oxidation and blackening) could be administrated to living cells and organisms, in which biosynthesis of boron-melanin in melanoma melanocytes can lead to improved BNCT.