Date of Award
Master of Science (MS)
Electrical and Computer Engineering
Hye Ye, Dong
Povinelli, Richard J.
Mass Spectrometry Imaging (MSI) extracts molecular mass data to form visualizations of molecular spatial distributions. The involved scanning procedure is conducted by moving a probe across and around a rectilinear grid, as in the case of nanoscale Desorption Electro-Spray Ionization (nano-DESI) MSI, where singular measurements can take up to ~5 seconds to acquire high-resolution (better than 10 μm) results. This temporal expense creates a high inefficiency in sample processing and throughput. For example, in a high-resolution nano-DESI study, a single mouse uterine tissue section (2.5 mm by 1.7 mm) had an acquisition time of ~4 hours to acquire 104,400 pixels. Anywhere from ~25-30% of those pixels were outside the actual tissue, and a further portion of those locations lacked relevant information. An existing method, a Supervised Learning Approach for Dynamic Sampling (SLADS), utilizes information obtained during an active scan to infer, using a least-squares regression, regions of interest that most likely contain meaningful information, and a computationally inexpensive weighted mean interpolation to perform sparse sample reconstruction. This approach could potentially be used to significantly improve throughput in this and other biological tissue scanning applications. However, existing SLADS implementations were neither designed nor optimized for leveraging or handling the 3rd dimension in MSI of molecular spectra. Further, integrating more recent advances in machine learning since the last SLADS publication issuance, such as Convolutional Neural Network (CNN) architectures, offers additional performance gains. The objective of this research is the updating, re-design, and optimization of the SLADS methodology, to form a Deep Learning Approach for Dynamic Sampling (DLADS) for high-resolution biological tissues and integration with nano-DESI MSI instrumentation.