Despite recent advances of deep Convolutional Neural Networks (CNNs) in various computer vision tasks, their potential for classification of multispectral remote sensing images has not been thoroughly explored. In particular, the applications of deep CNNs using optical remote sensing data have focused on the classification of very high-resolution aerial and satellite data, owing to the similarity of these data to the large datasets in computer vision. Accordingly, this study presents a detailed investigation of state-of-the-art deep learning tools for classification of complex wetland classes using multispectral RapidEye optical imagery. Specifically, we examine the capacity of seven well-known deep convnets, namely DenseNet121, InceptionV3, VGG16, VGG19, Xception, ResNet50, and InceptionResNetV2, for wetland mapping in Canada. In addition, the classification results obtained from deep CNNs are compared with those based on conventional machine learning tools, including Random Forest and Support Vector Machine, to further evaluate the efficiency of the former to classify wetlands. The results illustrate that the full-training of convnets using five spectral bands outperforms the other strategies for all convnets. InceptionResNetV2, ResNet50, and Xception are distinguished as the top three convnets, providing state-of-the-art classification accuracies of 96.17%, 94.81%, and 93.57%, respectively. The classification accuracies obtained using Support Vector Machine (SVM) and Random Forest (RF) are 74.89% and 76.08%, respectively, considerably inferior relative to CNNs. Importantly, InceptionResNetV2 is consistently found to be superior compared to all other convnets, suggesting the integration of Inception and ResNet modules is an efficient architecture for classifying complex remote sensing scenes such as wetlands.

No abstract

Cross sections (CSs) for the interaction of low-energy electrons (LEE) with condensed macromolecules are essential parameters for accurate modeling of radiation-induced molecular decomposition and chemical synthesis. Electron irradiation of dry nanometer-scale macromolecular solid films has often been employed to measure CSs and other quantitative parameters for LEE interactions. Since such films have thicknesses comparable with electron thermalization distances, energy deposition varies throughout the film. Moreover, charge accumulation occurring inside the films shields a proportion of the macromolecules from electron irradiation. Such effects complicate the quantitative comparison of the CSs obtained in films of different thicknesses and limit the applicability of such measurements. Here, we develop a simple mathematical model, termed the molecular survival model, that employs a CS for a particular damage process together with an attenuation length related to the total CS, to investigate how a measured CS might be expected to vary with experimental conditions. As a case study, we measure the absolute CS for the formation of DNA strand breaks (SBs) by electron irradiation at 10 and 100 eV of lyophilized plasmid DNA films with thicknesses between 10 and 30 nm. The measurements are shown to depend strongly on the thickness and charging condition of the nanometer-scale films. Such behaviors are in accord with the model and support its validity. Via this analysis, the CS obtained for SB damage is nearly independent of film thickness and charging effects. In principle, this model can be adapted to provide absolute CSs for electron-induced damage or reactions occurring in other molecular solids across a wider range of experimental conditions.

KeywordsCancer; Cisplatin; DNA damage; Low-energy electrons; OxaliplatinThe sensitization of malignant cells to ionizing radiation is the clinical rationale for the use of platinum-based concurrent chemoradiotherapy (CCRT) for cancer treatment; however, the specific mechanisms of radiosensitization and their respective contributions still remain unknown [1] . Biological mechanisms such as inhibition of DNA repair may contribute to the efficacy of CCRT [2] ; nevertheless, there is a dearth of information on the possible contribution of nanoscopic mechanisms to the generation of lethal DNA lesions such as double-strand breaks (DSB). The present study demonstrates that the abundant near zero-eV (0.5-eV) electrons, created by ionizing radiation during radiotherapy, induce DSB in supercoiled plasmid DNA modified by platinum anticancer drugs (Pt-drugs), but not in unmodified DNA. They do so more efficiently than other types of radiation, including soft X-rays and 10-eV electrons. The formation of DSB by 0.5-eV electrons is found to be a single-hit process. These findings reveal insights into the radiosensitization mechanism of Pt-drugs that can have implications for the development of optimal clinical protocols for platinum-based CCRT and the deployment of in-situ sources of subexcitation-energy electrons (e.g., Auger-electron emitting radionuclides) to efficiently enhance DSB formation in DNA modified by Pt-drugs in malignant cells.Pt-drug molecules react with cellular components, including DNA, RNA and proteins, via ligand exchange at the platinum atom [3] . Binding of Pt-drugs to DNA causes a local distortion of the DNA conformation affecting the physical and chemical properties of the molecule, particularly at the site of platination [4] . These modifications are believed to enhance the effects of radiation on the modified DNA by perturbing processes occurring during the physical and chemical stages of the interaction of radiation with DNA [5] . In an aqueous solution of DNA, cisplatin enhances the formation of DNA strand breaks by the reaction of both hydroxyl radicals and hydrated electrons at the site of platination through CIHR Author ManuscriptCIHR Author Manuscript CIHR Author Manuscript the increased accessibility of these species to sugar moieties and dissociative electron transfer [6] . Additionally, in the presence of Pt-DNA adducts, low energy electrons (LEEs, i.e., electrons of 0-30 eV) have a higher cross section for DNA damage than other types of radiations [7,8] . Previously, we hypothesized that at such energies, an incident electron is captured by a DNA constituent at the platination site to form a local transient negative ion (TNI) via a resonance process, which may subsequently rupture chemical bonds via either dissociative electron attachment (DEA) or dissociative electronic excitation (DEE) [8,9] . To determine the primary mechanism responsible for the enhanced damage to DNA modified by Pt-drugs, in the present study, we investigated the contribution and efficiency of these decay channels in ...

Platinum chemotherapeutic drugs preferentially enhance the formation of cluster damage to DNA induced by the direct effect of ionizing radiation, and LEEs are the main species responsible for such an enhancement via the formation of electron resonances.

The synergistic interaction of cisplatin with ionizing radiation is the clinical rationale for the treatment of several cancers including head and neck, cervical and lung cancer. The underlying molecular mechanism of the synergy has not yet been identified, although both DNA damage and repair processes are likely involved. Here, we investigate the indirect effect of γ rays on strand break formation in a supercoiled plasmid DNA (pGEM-3Zf-) covalently modified by cisplatin. The yields of single- and double-strand breaks were determined by irradiation of DNA and cisplatin/DNA samples with (60)Co γ rays under four different scavenging conditions to examine the involvement of hydrated electrons and hydroxyl radicals in inducing the DNA damage. At 5 mM tris in an N2 atmosphere, the presence of an average of two cisplatins per plasmid increased the yields of single- and double-strand breaks by factors of 1.9 and 2.2, respectively, relative to the irradiated unmodified DNA samples. Given that each plasmid of 3,200 base pairs contained an average of two cisplatins, this represents an increase in radiosensitivity of 3,200-fold on a per base pair basis. When hydrated electrons were scavenged by saturating the samples with N2O, these enhancement factors decreased to 1.5 and 1.2, respectively, for single- and double-strand breaks. When hydroxyl radicals were scavenged using 200 mM tris, the respective enhancement factors were 1.2 and 1.6 for single- and double-strand breaks, respectively. Furthermore, no enhancement in DNA damage by cisplatin was observed after scavenging both hydroxyl radicals and hydrated electrons. These findings show that hydrated electrons can induce both single- and double-strand breaks in the platinated DNA, but not in unmodified DNA. In addition, cisplatin modification is clearly an extremely efficient means of increasing the formation of both single- and double-strand breaks by the hydrated electrons and hydroxyl radicals created by ionizing radiation.

No abstract

Given the distinct characteristics of low-energy electrons (LEEs), particularly at energies less than 30 eV, they can be applied to a wide range of therapeutic modalities to improve cancer treatment. LEEs have been shown to efficiently produce complex molecular damage resulting in substantial cellular toxicities. Since LEEs are produced in copious amounts from high-energy radiation beam, including photons, protons and ions; the control of LEE distribution can potentially enhance the therapeutic radio of such beams. LEEs can play a substantial role in the synergistic effect between radiation and chemotherapy, particularly halogenated and platinum-based anticancer drugs. Radiosensitizing entities containing atoms of high atomic number such as gold nanoparticles can be a source of LEE production if high-energy radiation interacts with them. This can provide a high local density of LEEs in a cell and produce cellular toxicity. Auger-electron-emitting radionuclides also create a high number of LEEs in each decay, which can induce lethal damage in a cell. Exploitation of LEEs in cancer treatment, however, faces a few challenges, such as dosimetry of LEEs and selective delivery of radiosensitizing and chemotherapeutic molecules close to cellular targets. This review first discusses the rationale for utilizing LEEs in cancer treatment by explaining their mechanism of action, describes theoretical and experimental studies at the molecular and cellular levels, then discusses strategies for achieving modification of the distribution and effectiveness of LEEs in cancerous tissue and their associated clinical benefit.