Detection of hidden tumors by imaging and machine learning
December 2020
National Cancer Center, Kashiwa, Japan
The diagnosis of gastrointestinal stromal tumor (GIST) using conventional endoscopy is difficult because submucosal tumor lesions like GIST are covered by a mucosal layer and thus can be overseen. To overcome this issue, near-infrared hyperspectral imaging (NIR-HSI) was combined with machine learning. 12 gastric GIST lesions were surgically resected and imaged ex vivo with a near-infrared (NIR) hyperspectral camera. The site of the GIST was defined by a pathologist using the NIR image to prepare training data for normal and GIST regions. A machine learning algorithm was then used to predict normal and GIST regions. Accuracy was around 86%, therefore NIR-HSI analysis may have the potential to distinguish deep lesions. An endoscope is planned in order to use the method in vivo during standard endoscopy.
Distinction of surgically resected gastrointestinal stromal tumor by near-infrared hyperspectral imaging
Toshihiro Takamatsu
Added on: 02-09-2021
[1] https://www.nature.com/articles/s41598-020-79021-7[2] https://www.bionity.com/en/news/1169742/deep-vision-near-infrared-imaging-and-machine-learning-can-identify-hidden-tumors.html
