Utilizing machine learning to accomplish medical picture segmentation is a novel approach.
China’s Beihang University has created a new training approach that automates a significant portion of the manual annotation labor performed by people. Machine learning can do blood cell counts for illness diagnosis instead of costly and often less precise cell analyzer devices. On April 9, an article describing their new training program has published in the journal Cyborg and Bionic Systems.
Since 2015, researchers from China’s Beihang University have created a new training strategy for the CNN, or convolutional network segmentation model, frequently used in medical picture segmentation. They discovered that their training approach for segmenting multiple-cell-type photos reached the same level as training with manually annotated multiple-tone white images: 94.85 percent. Source: https://techxplore.com/news/2022-05-machine-radically-workload-cell-disease.html
