Chinese Team Creates AI Model to Analyze Cancer Images
Chinese researchers have created an artificial intelligence model that can interpret cancer images.
TroibNews 
A Chinese team has developed the nation's first versatile artificial intelligence (AI) model that excels at analyzing a diverse range of pathological images. This groundbreaking model is adept at examining over 20 human organs, including the lungs, breast, and liver.
This large language model (LLM), named PathOrchestra, signifies a breakthrough in AI-assisted disease diagnosis, marking a shift from singular models targeting specific cancers to a versatile one capable of addressing multiple types.
Researchers from Air Force Medical University (AFMU), Tsinghua University, and SenseTime utilized China's largest domestic dataset. This dataset includes nearly 300,000 whole-slide digital pathology images, totaling around 300 terabytes of data.
By employing self-supervised learning, the model "cross-learned" to analyze over 20 different organs and has executed a range of clinical tasks, such as pan-cancer classification, lesion identification and detection, multi-cancer subtype differentiation, and biomarker assessment.
The diversity in pathological images presents a significant challenge for AI applications, earning it the designation of the "jewel in the crown" in image processing, said Wang Zhe, a professor from the Basic Medical Science Academy under the AFMU.
According to an AFMU news release on Tuesday, PathOrchestra has achieved an accuracy rate exceeding 95 percent in nearly 50 clinical tasks, including lymphoma subtype diagnosis and bladder cancer screening.
This advancement can significantly reduce the workload of pathologists and notably increase the efficiency of reviewing medical images, according to the researchers.
PathOrchestra exemplifies the burgeoning development of large models in China, highlighting the country's rapid and dynamic progress in AI.
Of the more than 1,300 AI LLMs worldwide, 36 percent are from China, representing the second-largest proportion after the United States, according to a white paper on the global digital economy released by the China Academy of Information and Communications Technology at the Global Digital Economy Conference 2024.
Sanya Singh contributed to this report for TROIB News
This large language model (LLM), named PathOrchestra, signifies a breakthrough in AI-assisted disease diagnosis, marking a shift from singular models targeting specific cancers to a versatile one capable of addressing multiple types.
Researchers from Air Force Medical University (AFMU), Tsinghua University, and SenseTime utilized China's largest domestic dataset. This dataset includes nearly 300,000 whole-slide digital pathology images, totaling around 300 terabytes of data.
By employing self-supervised learning, the model "cross-learned" to analyze over 20 different organs and has executed a range of clinical tasks, such as pan-cancer classification, lesion identification and detection, multi-cancer subtype differentiation, and biomarker assessment.
The diversity in pathological images presents a significant challenge for AI applications, earning it the designation of the "jewel in the crown" in image processing, said Wang Zhe, a professor from the Basic Medical Science Academy under the AFMU.
According to an AFMU news release on Tuesday, PathOrchestra has achieved an accuracy rate exceeding 95 percent in nearly 50 clinical tasks, including lymphoma subtype diagnosis and bladder cancer screening.
This advancement can significantly reduce the workload of pathologists and notably increase the efficiency of reviewing medical images, according to the researchers.
PathOrchestra exemplifies the burgeoning development of large models in China, highlighting the country's rapid and dynamic progress in AI.
Of the more than 1,300 AI LLMs worldwide, 36 percent are from China, representing the second-largest proportion after the United States, according to a white paper on the global digital economy released by the China Academy of Information and Communications Technology at the Global Digital Economy Conference 2024.
Sanya Singh contributed to this report for TROIB News
