Analysis of the static time-of-flight secondary ion mass spectrometry (ToF-SIMS) spectra of adsorbed protein films is reported using principal component analysis (PCA) and a novel artificial neural network (ANN) approach, NeuroSpectraNet, to classify chemically the spectra of the protein films. The ease of application and the efficiency with which each approach classified positive ion spectra from adsorbed films of 13 different proteins is reported and assessed. The ToF-SIMS spectra of adsorbed protein films are especially difficult to analyze owing to the absence of unique peaks in the spectra of different proteins. Although PCA was able to differentiate successfully ToF-SIMS spectra of adsorbed protein films using the ions generated from the fragmentation of the amino acids, differentiation of the spectra using the entire spectrum was unsuccessful. Outliers in several of the protein groups make classification of unknown spectra difficult, despite the use of only amino-acid-specific ions. However, NeuroSpectraNet successfully classified the spectra from 11 of the protein films using the whole positive ion spectra after a vector analysis enhancement had been incorporated into the neural network. Full classification of all 13 proteins was achieved by using the combined positive and negative ion spectra. However, as with PCA, ANN classification was enhanced when the input patterns only contained amino-acid-specific ions. The complex and multivariate nature of static SIMS spectra is a domain well suited to the application of neural networks for pattern recognition and classification.

No abstract

Breast cancer is the most common type of cancer worldwide. A robotic system performing autonomous breast palpation can make a significant impact on the related health sector worldwide. However, robot programming for breast palpating with different geometries is very complex and unsolved. Robot learning from demonstrations (LfD) reduces the programming time and cost. However, the available LfD are lacking the modelling of the manipulation path/trajectory as an explicit function of the visual sensory information. This paper presents a novel approach to manipulation path/trajectory planning called deep Movement Primitives that successfully generates the movements of a manipulator to reach a breast phantom and perform the palpation. We show the effectiveness of our approach by a series of real-robot experiments of reaching and palpating a breast phantom. The experimental results indicate our approach outperforms the state-of-the-art method.

Breast cancer is the most common type of cancer worldwide. A robotic system performing autonomous breast palpation can make a significant impact on the related health sector worldwide. However, robot programming for breast palpating with different geometries is very complex and unsolved. Robot learning from demonstrations (LfD) reduces the programming time and cost. However, the available LfD are lacking the modelling of the manipulation path/trajectory as an explicit function of the visual sensory information. This paper presents a novel approach to manipulation path/trajectory planning called deep Movement Primitives that successfully generates the movements of a manipulator to reach a breast phantom and perform the palpation. We show the effectiveness of our approach by a series of realrobot experiments of reaching and palpating a breast phantom. The experimental results indicate our approach outperforms the state-of-the-art method.