In-silico investigation into optimal photoacoustic probe design using a digital thyroid phantom.

Significance: Thyroid nodules have a high incidence, but a low malignancy rate, placing demands on imaging specificity. Photoacoustic imaging is being studied as a method that could potentially help increase the unsatisfactory specificity of ultrasound, the current first-line imaging method, which would decrease the number of unnecessary clinical interventions. Aim: We employ a digital thyroid phantom and numerical simulations to evaluate photoacoustic probe designs on their suitability for thyroid nodule risk estimation. Approach: Using the SIMPA toolkit, we examined 1080 probe configurations with varying illumination geometry, detector shape, pitch, and wavelength combinations. The performance of each probe was quantified on its sensitivity and its accuracy of oxygen saturation estimation. Results: Optimal performance was found using wavelengths 757 and 800 nm, and a curved detector (0.5 split, 0.40 mm pitch) for external illumination. However, performance increases significantly with endotracheal illumination, using wavelengths of 700 and 800 nm, and a curved detector (0.25 split, 0.40 mm pitch). Conclusions: These results could provide design guidance for optimizing photoacoustic hardware, assisting with the clinical adoption of PAI, which could improve the diagnosis of thyroid nodules.