Prompt gamma ray activation analysis for determining chemical composition of 3D printing and casting materials used in biomedical applications.

Three-dimensional printing and casting materials were analyzed by prompt gamma-ray activation analysis (PGAA) to determine their suitability as human tissue surrogates for the fabrication of phantoms for medical imaging and radiation dosimetry applications. Measured elemental compositions and densities of five surrogate materials simulating soft tissue and bone were used to determine radiological properties (x-ray mass attenuation coefficient and electron stopping power). When compared with radiological properties of International Commission on Radiation Units and Measurements (ICRU) materials, it was determined that urethane rubber and PLA plastic yielded the best match for soft tissue, while silicone rubber and urethane resin best simulated the properties of bone.

Application of machine learning classifiers to X-ray diffraction imaging with medically relevant phantoms.

PURPOSE: Recent studies have demonstrated the ability to rapidly produce large field of view X-ray diffraction (XRD) images, which provide rich new data relevant to the understanding and analysis of disease. However, work has only just begun on developing algorithms that maximize the performance toward decision-making and diagnostic tasks. In this study, we present the implementation of and comparison between rules-based and machine learning (ML) classifiers on XRD images of medically relevant phantoms to explore the potential for increased classification performance. METHODS: Medically relevant phantoms were utilized to provide well-characterized ground-truths for comparing classifier performance. Water and polylactic acid (PLA) plastic were used as surrogates for cancerous and healthy tissue, respectively, and phantoms were created with varying levels of spatial complexity and biologically relevant features for quantitative testing of classifier performance. Our previously developed X-ray scanner was used to acquire co-registered X-ray transmission and diffraction images of the phantoms. For classification algorithms, we explored and compared two rules-based classifiers (cross-correlation, or matched-filter, and linear least-squares unmixing) and two ML classifiers (support vector machines and shallow neural networks). Reference XRD spectra (measured by a commercial diffractometer) were provided to the rules-based algorithms, while 60% of the measured XRD pixels were used for training of the ML algorithms. The area under the receiver operating characteristic curve (AUC) was used as a comparative metric between the classification algorithms, along with the accuracy performance at the midpoint threshold for each classifier. RESULTS: The AUC values for material classification were 0.994 (cross-correlation [CC]), 0.994 (least-squares [LS]), 0.995 (support vector machine [SVM]), and 0.999 (shallow neural network [SNN]). Setting the classification threshold to the midpoint for each classifier resulted in accuracy values of CC = 96.48%, LS = 96.48%, SVM = 97.36%, and SNN = 98.94%. If only considering pixels ±3 mm from water-PLA boundaries (where partial volume effects could occur due to imaging resolution limits), the classification accuracies were CC = 89.32%, LS = 89.32%, SVM = 92.03%, and SNN = 96.79%, demonstrating an even larger improvement produced by the machine-learned algorithms in spatial regions critical for imaging tasks. Classification by transmission data alone produced an AUC of 0.773 and accuracy of 85.45%, well below the performance levels of any of the classifiers applied to XRD image data. CONCLUSIONS: We demonstrated that ML-based classifiers outperformed rules-based approaches in terms of overall classification accuracy and improved the spatially resolved classification performance on XRD images of medical phantoms. In particular, the ML algorithms demonstrated considerably improved performance whenever multiple materials existed in a single voxel. The quantitative performance gains demonstrate an avenue to extract and harness XRD imaging data to improve material analysis for research, industrial, and clinical applications.

PVA matches human liver in needle-tissue interaction.

Medical phantoms can be used to study needle-tissue interaction and to train medical residents. The purpose of this research is to study the suitability of polyvinyl alcohol (PVA) as a liver tissue mimicking material in terms of needle-tissue interaction. Insertions into ex-vivo human livers were used for reference. Six PVA samples were created by varying the mass percentage of PVA to water (4m% and 7m%) and the number of freeze-thaw cycles (1, 2 and 3 cycles, 16hours of freezing at -19°C, 8hours of thawing). The inner needle of an 18 Gauge trocar needle with triangular tip was inserted 13 times into each of the samples, using an insertion velocity of 5 mm/s. In addition, 39 insertions were performed in two ex-vivo human livers. Axial forces on the needle were captured during insertion and retraction and characterized by friction along the needle shaft, peak forces, and number of peak forces per unit length. The concentration of PVA and the number of freeze-thaw cycles both influenced the mechanical interaction between needle and specimen. Insertions into 4m% PVA phantoms with 2 freeze-thaw cycles were comparable to human liver in terms of estimated friction along the needle shaft and the number of peak forces. Therefore, these phantoms are considered to be suitable liver mimicking materials for image-guided needle interventions. The mechanical properties of PVA hydrogels can be influenced in a controlled manner by varying the concentration of PVA and the number of freeze-thaw cycles, to mimic liver tissue characteristics.