Fig 1.
MoMic miniature microscope prototype.
The glass slide is placed in the sample holder which is then inserted on the mechanical stage, and used for manual adjustment of the sample position.
Fig 2.
US Air Force 1951 three-bar resolution test chart.
Image of standard 1951 USAF resolution target, captured with the point-of-care microscope. Due to the reversed camera lens the original image appears reversed horizontally.
Fig 3.
Pixel intensity profiles, calculated from four different sets of bars in the USAF three-bar resolution test chart.
Intensity values (RGB) calculated from horizontal lines across regions indicated with red bounding boxes in Fig 2. From left to right: Group 1, element 1; group 8, element 2; group 8 element 5 (smallest resolvable); group 8, element 6.
Fig 4.
A breast cancer core biopsy, immunostained for estrogen receptor-alpha digitized with both imaging devices.
The virtual sample has been digitized with the reference slide-scanner (left) and the point-of-care microscope (right). Regions marked with red bounding boxes represent tiles extracted for analysis. These can be seen at higher magnification in the lower part of the image (A. Reference slide-scanner B. Point-of-care microscope).
Table 1.
Distribution of the classification results of the estrogen receptor-status in the breast cancer samples.
Fig 5.
Estrogen receptor status (percentage of cell nuclei staining positively for ER expression), as assessed by the different methods.
From left to right: 1) Detected ER positivity by image-analysis of images from point-of-care microscope, compared to images from the reference slide-scanner. 2) Detected ER positivity by image-analysis of images from point-of-care microscope, compared to manual scoring. 3) Detected ER positivity by image-analysis of images from reference slide-scanner, compared to images from the point-of-care microscope.
Fig 6.
Results from digital image-analysis of sample expressing high levels of estrogen receptors.
Left column showing results of ImmunoRatio2-analysis of sample digitized with point-of-care microscope. Middle column showing analysis of reference slide-scanner image. Right column showing magnifications of images, as indicated by red bounding boxes. Positive nuclei marked by blue dots, negative nuclei marked by yellow dots. This particular sample was visually scored as strongly ER positive (82%).
Table 2.
Fig 7.
Agreement between the different methods of estrogen receptor-assessment, depicted with Bland-Altman diagrams.
From left to right: 1) Mean value of detected estrogen receptor expression by manual scoring and analysis of slide-scanner images, plotted against difference in detected positivity. 2) Mean value of detected estrogen receptor expression by image-analysis of point-of-care microscope images and of slide-scanner images, plotted against difference in detected positivity. The highest concordance was detected between the two device-based scorings. 3) Mean value of detected estrogen receptor expression by image-analysis of point-of-care microscope images and manual scoring, plotted against difference in detected positivity.
Fig 8.
Results from digital image-analysis of borderline sample.
Left column showing results of ImmunoRatio2-analysis of sample digitized with point-of-care microscope. Middle column showing analysis of reference slide-scanner image. Right column showing magnifications of images, as indicated by red bounding boxes. Positive nuclei marked by blue dots, negative nuclei marked by yellow dots. Sample scored as ER negative (0%) by visual scoring.