Angular range, sampling and noise considerations for inverse light scattering analysis of nuclear morphology

In recent years, significant work has been devoted to the use of angle‐resolved elastic scattering for the extraction of nuclear morphology in tissue. By treating the nucleus as a Mie scattering object, techniques such as angle‐resolved low‐coherence interferometry (a/LCI) have demonstrated substantial success in identifying nuclear alterations associated with dysplasia. Because optical biopsies are inherently noninvasive, only a small, discretized portion of the 4π scattering field can be collected from tissue, limiting the amount of information available for diagnostic purposes. In this work, we comprehensively characterize the diagnostic impact of variations in angular sampling, range and noise for inverse light scattering analysis of nuclear morphology, using a previously reported dataset from 40 patients undergoing a/LCI optical biopsy for cervical dysplasia. The results from this analysis are applied to a benchtop scanning a/LCI system which compromises angular range for wide‐area scanning capability. This work will inform the design of next‐generation optical biopsy probes by directing optical design towards parameters which offer the most diagnostic utility. Angle‐resolved low‐coherence interferometry measures nuclear morphology in tissue using a narrow, discretized portion of the scattering function from cell nuclei. In this study, we examine the diagnostic impact of compression of the scattering function, including reductions in angular range and sampling frequency, and addition of additive noise. Diagnostic parameters are investigated using a clinical dataset of 40 patients undergoing optical biopsy of the cervix. A benchtop probe for detecting dysplasia which sacrifices scattering data for scanning capability is discussed.