Google is fighting the good fight against cancer. Google researchers are using deep structured learning algorithms to train computers to inspect slide images of tissue samples and diagnose cancer correctly.
Google reported that an ongoing investigation into the use of deep structured learning algorithms in the field of digital pathology is showing early promise.
Google's machine learning group researchers are developing algorithms to detect metastatic breast cancer cells from whole slide images of pathological tissue samples.
Radboud University Medical Center in Netherlands supplied images to Google researchers which are then used to train algorithms to detect and recognize that a particular person's breast cancer is likely to have metastasized or spread to the axillary lymph nodes.
A feat that can be difficult or even impossible for a trained human pathologist to achieve. Two different pathologist often arrived at different diagnoses on the same set of images. No surprises here Giving the sheer number of slides they have to go through, Peng Stumped and said.
 |
| Screen shot from the research paper |
Google reported that an ongoing investigation into the use of deep structured learning algorithms in the field of digital pathology is showing early promise.
Google's machine learning group researchers are developing algorithms to detect metastatic breast cancer cells from whole slide images of pathological tissue samples.
Radboud University Medical Center in Netherlands supplied images to Google researchers which are then used to train algorithms to detect and recognize that a particular person's breast cancer is likely to have metastasized or spread to the axillary lymph nodes.
A feat that can be difficult or even impossible for a trained human pathologist to achieve. Two different pathologist often arrived at different diagnoses on the same set of images. No surprises here Giving the sheer number of slides they have to go through, Peng Stumped and said.
No comments:
Post a Comment