Instrument Development: Depth-Resolved Chemical Imaging for Highly Absorbing, Scattering, and Curved Specimens

This project is developing a volumetric chemical imaging platform to noninvasively quantify the internal chemical composition of thick, highly scattering specimens. By integrating shortwave infrared (SWIR) optical technology with nonlinear inverse reconstruction, the research should enable nondestructive analysis of millimeter-scale complex specimens, including 3D-printed pharmaceutical tablets, engineered human tissues, and environmental microparticles embedded within filter matrices. For pharmaceutical applications, the platform is designed to provide quantitative 3D maps of active ingredients within layered tablets, improving quality control and counterfeit detection. For biotechnology applications, this noninvasive imaging capability will enable longitudinal, quantitative monitoring of molecular and metabolic changes in human organoids, supporting advances in tissue engineering and disease modeling. Broader impacts include training undergraduate and graduate students in optical instrumentation, LabVIEW programming, computational imaging, and deep learning. The project also engages K–12 students and educators through hands-on workshops and public demonstrations while establishing a validated imaging platform for shared research use. The intellectual merit of this project lies in advancing diffuse optical tomography (DOT) toward quantitative chemical imaging in strongly scattering and absorbing media across the 1000–2500 nm spectral range. The system will integrate spatial frequency domain imaging with Fourier-transform spectroscopy to simultaneously recover depth-resolved optical scattering, absorption, and chemical signatures. To achieve quantitative molecular mapping, the project is developing a Monte Carlo–based nonlinear reconstruction framework incorporating a three-dimensional radiative transfer equation solver that accounts for realistic boundary conditions, sample curvature, and refractive index mismatch. Deep learning approaches will be developed to analyze raw interferometric measurements, resolve overlapping spectral features, and improve robustness against measurement noise. Together, these innovations should extend DOT beyond conventional qualitative or bulk measurements toward a quantitative volumetric chemical imaging platform capable of four-dimensional spatial–spectral characterization of complex specimens. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.