Systems and methods utilizing artificial intelligence for placental assessment and examination
US11244450B2 · kind B2 · utility
Assignees
Inventors
Key dates
| Filing date | Aug 18, 2020 |
| Grant date | Feb 8, 2022 |
| Priority date | — |
| Expiry date | Aug 18, 2040 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06V2201/03
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Systems and methods for completing a morphological characterization of an image of a placenta and providing suggested pathological diagnoses are disclosed. A system includes programming instructions that, when executed, cause processing devices to execute commands according to the following logic modules: an Encoder module that receives the digital image of the placenta and outputs a pyramid of feature maps, a SegDecoder module that segments the pyramid of feature maps on a fetal side image and on a maternal side image, a Classification Subnet module that classifies the fetal side image and the maternal side image, and a convolutional IPDecoder module that localizes an umbilical cord insertion point of the placenta from the classified fetal side image and the classified maternal side image. The localized umbilical cord insertion point, segmentation maps for the classified fetal side and maternal side images are provided to an external device for determining the morphological characterization.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.