Chemical imaging, especially mid-infrared spectroscopic microscopy, enables label-free biomedical analyses while achieving expansive molecular sensitivity. However, its slow speed and poor image quality impede widespread adoption. We present a microscope that provides high-throughput recording, low noise, and high spatial resolution where the bottom-up design of its optical train facilitates dual-axis galvo laser scanning of a diffraction-limited focal point over large areas using custom infinity-corrected objectives. The data quality enables applications of modern machine learning and capabilities not previously feasible. Distinct from conventional approaches that focus on morphological investigations or immunostaining, this development makes label-free imaging of minimally processed tissue practical.
Infrared spectroscopic imaging combines the ability to record molecular content with the ability to visualize chemistry in its spatial diversity. Given the need to record a significantly larger quantity of data than a typical microscopy image (MB vs. GB) and the extensive bandwidth of the spectra (~10 m), trade-offs often have to be made between the closely related considerations of signal to noise ratio, spatial-spectral coverage, resolution and optical arrangements. Here, we present a path from rigorous theory to modeling and design to realizing the advantages offered by new ideas on fundamentally changing these trade-offs. We first describe a new microscope design for increased speed and rapid coverage that is useful for biomedical and clinical tissue imaging. Next, we describe a configuration to measure chirality in samples that promises higher spectral information that present methods. Finally, we present a new approach to nanoscale IR imaging that provides greater fidelity and speed at unprecedented levels of signal to noise ratio. Finally, we show how emerging machine learning approaches can further augment these advances. For each instrumentation advance, examples of use cases will be presented.
Applications of machine learning in pathology is an active research area in modern medicine. Here, we presented a classifier for label-free renal histopathology. Three frequently encountered categories of monoclonal gammopathy-associated kidney disease were studied, which included light chain amyloidosis, monoclonal light chain disease deposition (MIDD) and myeloma cast nephropathy. Biopsies with diabetic nephropathy and normal baseline transplant biopsies were used as control. The samples are imaged using a FTIR hyperspectral microscope. More than three million infrared spectra are adopted for the training and evaluation of the computational model. The model recognizes the pixels associated with the glomerulus, and diagnoses the disease based on infrared absorption features.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.