This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision | |||
|
release_notes [2026/04/30 19:56] rudy |
release_notes [2026/05/04 09:30] (current) rudy |
||
|---|---|---|---|
| Line 10: | Line 10: | ||
| == Release note addendum AI annotation behaviour changes clarification April 30th, 2026 == | == Release note addendum AI annotation behaviour changes clarification April 30th, 2026 == | ||
| - | The new released PMA.core viewer and AI result rendering that are essential components that provide functionality to My Pathomation were updated to PMA.core v3.0.8 and AI 1.1on April 3rd, 2026. These software versions introduced improved handling of AI generated annotations to allow better image reactivity. This has altered the behaviour of viewing the AI generated annotations. When AI annotations are present at high density, meaning most individual cells in a tissue area are annotated the viewer will only show all individual annotations present in the image at higher magnification fields of view. You can notice an incremental annotation apparition behaviour matching the capacity of the observer to discern individual elements in the image, the software will take in account the available field of view of a magnification and saturate it with annotations. As you move from lower magnifications 1x-10x to 20x-40x and the field of view starts to reveal individual cells the software will show every annotation present in densely packed annotation areas. Please be aware that high magnification views are required to see annotations classes that are present only in low minority numbers in between a large majority number of another class. | + | The new released PMA.core viewer and AI result rendering that are essential components providing functionality to My Pathomation were updated to PMA.core v3.0.8 and AI 1.1 on April 3rd, 2026. These software versions introduced improved handling of AI generated annotations to allow better image reactivity. This has altered the behaviour of viewing the AI generated annotations. \\ |
| + | When AI annotations are present at high density, meaning most individual cells in a tissue area are annotated the viewer will only show all individual annotations present in the image at higher magnification fields of view. You can notice an incremental annotation apparition behaviour matching the capacity of the observer to discern individual elements in the image, the software will take in account the available field of view of a magnification and saturate it with annotations. \\ | ||
| + | As you move from lower magnifications 1x-10x to 20x-40x and the field of view starts to reveal individual cells the software will show every annotation present in densely packed annotation areas. \\ | ||
| + | |||
| + | **Please be aware that high magnification views are required to see annotations classes that are present only in low minority numbers in between a large majority number of another class.** | ||
| {{ :annotation_saturation_behaviour_1.png?600 |}} | {{ :annotation_saturation_behaviour_1.png?600 |}} | ||
| - | It is good practice to always check the AI result panel for the presence of minority class annotations. It can be helpful to mouse over the 'visor' icon (2) that highlights the selected class annotations for quick location in the slide. | + | **It is good practice to always check the AI result panel for the presence of minority class annotations.**\\ |
| + | With the PMA.core v3.0.8 update we inadvertently stopped the possibility to hide the annotations with the selection feature (1). A new release PMA.core 3.0.9 will remedy this bug and the release is planned in the beginning of May.\\ | ||
| + | It can be helpful as an alternative to mouse over the 'visor' icon (2) that highlights the selected class annotations for quick location in the slide. | ||
| {{ :annotation_saturation_behaviour.png?600 |}} | {{ :annotation_saturation_behaviour.png?600 |}} | ||
| - | With the PMA.core v3.0.8 update we inadvertently stopped the possibility to hide the annotations with the selection feature (1). A new release PMA.core 3.0.9 will remedy this bug and the release is planned in the beginning of May. | + | |
| | | ||