RESUMO
The tissue diagnosis of adenocarcinoma and intraductal carcinoma of the prostate includes Gleason grading of tumor morphology on the hematoxylin and eosin stain and immunohistochemistry markers on the prostatic intraepithelial neoplasia-4 stain (CK5/6, P63, and AMACR). In this work, we create an automated system for producing both virtual hematoxylin and eosin and prostatic intraepithelial neoplasia-4 immunohistochemistry stains from unstained prostate tissue using a high-throughput hyperspectral fluorescence microscope and artificial intelligence and machine learning. We demonstrate that the virtual stainer models can produce high-quality images suitable for diagnosis by genitourinary pathologists. Specifically, we validate our system through extensive human review and computational analysis, using a previously validated Gleason scoring model, and an expert panel, on a large data set of test slides. This study extends our previous work on virtual staining from autofluorescence, demonstrates the clinical utility of this technology for prostate cancer, and exemplifies a rigorous standard of qualitative and quantitative evaluation for digital pathology.
RESUMO
The uterine cervix softens, shortens, and dilates throughout pregnancy in response to progressive disorganization of its layered collagen microstructure. This process is an essential part of normal pregnancy, but premature changes are associated with preterm birth. Clinically, there are no reliable noninvasive methods to objectively measure cervical softening or assess cervical microstructure. The goal of these preliminary studies was to evaluate the feasibility of using an intracavity ultrasound array to generate acoustic radiation force impulse (ARFI) excitations in the uterine cervix through simulation, and to optimize the acoustic radiation force (ARF) excitation for shear wave elasticity imaging (SWEI) of the tissue stiffness. The cervix is a unique soft tissue target for SWEI because it has significantly greater acoustic attenuation (α = 1.3 to 2.0 dB·cm(-1)·MHz(-)1) than other soft tissues, and the pathology being studied tends to lead to an increase in tissue compliance, with healthy cervix being relatively stiff compared with other soft tissues (E ≈ 25 kPa). Additionally, the cervix can only be accessed in vivo using a transvaginal or catheter-based array, which places additional constraints on the excitation focal characteristics that can be used during SWEI. Finite element method (FEM) models of SWEI show that larger-aperture, catheter-based arrays can utilize excitation frequencies up to 7 MHz to generate adequate focal gain up to focal depths 10 to 15 mm deep, with higher frequencies suffering from excessive amounts of near-field acoustic attenuation. Using full-aperture excitations can yield ~40% increases in ARFI-induced displacements, but also restricts the depth of field of the excitation to ~0.5 mm, compared with 2 to 6 mm, which limits the range that can be used for shear wave characterization of the tissue. The center-frequency content of the shear wave particle velocity profiles ranges from 1.5 to 2.5 kHz, depending on the focal configuration and the stiffness of the material being imaged. Overall, SWEI is possible using catheter-based imaging arrays to generate adequate displacements in cervical tissue for shear wave imaging, although specific considerations must be made when optimizing these arrays for this shear wave imaging application.