Recent advances in tissue microarray technology have allowed immunohistochemistry to become a powerful medium-to-high throughput analysis tool, particularly for the validation of diagnostic and prognostic biomarkers. However, as study size grows, the manual evaluation of these assays becomes a prohibitive limitation; it vastly reduces throughput and greatly increases variability and expense. We propose an algorithm—Tissue Array Co-Occurrence Matrix Analysis (TACOMA)—for quantifying cellular phenotypes based on textural regularity summarized by local inter-pixel relationships. The algorithm can be easily trained for any staining pattern, is absent of sensitive tuning parameters and has the ability to report salient pixels in an image that contribute to its score. Pathologists’ input via informative training patches is an important aspect of the algorithm that allows the training for any specific marker or cell type. With co-training, the error rate of TACOMA can be reduced substantially for a very small training sample (e.g., with size 30). We give theoretical insights into the success of co-training via thinning of the feature set in a high dimensional setting when there is “sufficient” redundancy among the features. TACOMA is flexible, transparent and provides a scoring process that can be evaluated with clarity and confidence. In a study based on an estrogen receptor (ER) marker, we show that TACOMA is comparable to, or outperforms, pathologists’ performance in terms of accuracy and repeatability.

Due to the enormous dynamic range of human photoreceptors in response to light, studying their visual function in the intact retina challenges the stimulation hardware, specifically with regard to the displayable luminance contrast. The adaptive optics scanning laser ophthalmoscope (AOSLO) is an optical platform that focuses light to extremely small retinal extents, approaching the size of single photoreceptor cells. However, the current light modulation techniques produce spurious visible backgrounds which fundamentally limit experimental options. To remove unwanted background light and to improve contrast for high dynamic range visual stimulation in an AOSLO, we cascaded two commercial fiber-coupled acousto-optic modulators (AOMs) and measured their combined optical contrast. By compensating for zero-point differences in the individual AOMs, we demonstrate a multiplicative extinction ratio in the cascade that was in accordance with the extinction ratios of both single AOMs. When latency differences in the AOM response functions were individually corrected, single switch events as short as 50 ns with radiant power contrasts up to 1:10 10 were achieved. This is the highest visual contrast reported for any display system so far. We show psychophysically that this contrast ratio is sufficient to stimulate single foveal photoreceptor cells with small and bright enough visible targets that do not contain a detectable background. Background-free stimulation will enable photoreceptor testing with custom adaptation lights. Furthermore, a larger dynamic range in displayable light levels can drive photoreceptor responses in cones as well as in rods.

Multi-wavelength ophthalmic imaging and stimulation of photoreceptor cells require consideration of chromatic dispersion of the eye, manifesting in longitudinal and transverse chromatic aberrations. Contemporary image-based techniques to measure and correct transverse chromatic aberration (TCA) and the resulting transverse chromatic offset (TCO) in an adaptive optics retinal imaging system are precise but lack compensation of small but significant shifts in eye position occurring during in vivo testing. Here, we present a method that requires only a single measurement of TCO during controlled movements of the eye to map retinal chromatic image shifts to the image space of a pupil camera. After such calibration, TCO can be compensated by continuously monitoring eye position during experimentation and by interpolating correction vectors from a linear fit to the calibration data. The average change rate of TCO per head shift and the correlation between Kappa and the individual foveal TCA are close to the expectations based on a chromatic eye model. Our solution enables continuous compensation of TCO with high spatial precision and avoids high light intensities required for re-measuring TCO after eye position changes, which is necessary for foveal cone-targeted psychophysical experimentation.

8 Multi-wavelength ophthalmic imaging and stimulation of photoreceptor cells 9 requires consideration of chromatic dispersion of the eye, manifesting in 10 longitudinal and transverse chromatic aberrations. Current image-based techniques 11 to measure and correct transverse chromatic aberration (TCA) and the resulting 12 transverse chromatic offset (TCO) in an adaptive optics retinal imaging system are 13 precise, but lack compensation of small but significant shifts in eye position 14 occurring during in vivo testing. Here we present a method that requires only a 15 single measurement of TCO during controlled movements of the eye to map retinal 16 chromatic image shifts to the image space of a pupil camera. After such calibration, 17 TCO can be compensated by continuously monitoring eye position during 18 experimentation and by interpolating correction vectors from a linear fit to the 19 calibration data. The average change rate of TCO per head shift and the correlation 20 between Kappa and the individual foveal TCA are close to the expectations based 21 on a chromatic eye model. Our solution enables continuous correction of TCO with 22 high spatial precision and avoids high light intensities required for re-measuring 23 TCO after eye position changes, which is necessary for foveal cone-targeted 24 psychophysical experimentation. 25 26 27 28 29