Stanford AI Tool Predicts Cancer Gene Activity from Biopsy Images, Potentially Revolutionizing Cancer Diagnosis

Stanford Researchers Develop AI Tool to Predict Cancer Gene Activity

Stanford Medicine researchers have developed an innovative artificial intelligence (AI) tool that could revolutionize cancer diagnosis and treatment planning. The computational program, named SEQUOIA (slide-based expression quantification using linearized attention), can predict the activity of thousands of genes within tumor cells based solely on standard microscopy images of biopsy samples 1

1

How SEQUOIA Works

SEQUOIA was developed using data from 7,584 cancer biopsies across 16 different cancer types. The AI model analyzes thin sections of tumor biopsies prepared with hematoxylin and eosin staining, a standard method for visualizing cancer cells 2

2

By integrating this data with transcriptomic information and images from thousands of healthy cells, SEQUOIA can predict the expression patterns of more than 15,000 different genes from the stained images. For some cancer types, the AI-predicted gene activity showed over 80% correlation with actual gene activity data 3

3

Potential Impact on Cancer Diagnosis and Treatment

Traditional methods of determining genomic changes driving tumor growth require genetic sequencing of RNA isolated from the tumor, a process that can take weeks and cost thousands of dollars. SEQUOIA could potentially bypass this need, offering several advantages:

1. Speed: The AI tool can quickly identify gene signatures in patients' tumors, accelerating clinical decision-making.
2. Cost-effectiveness: By reducing the need for expensive genetic sequencing, SEQUOIA could save the healthcare system significant amounts of money.
3. Comprehensive analysis: The tool can predict the activation of large genomic programs, which is often more clinically relevant than individual gene expression.

Related Stories

AI Model Predicts Gene Activity in Human Cells, Transforming Biological Research

AI Tool Uncovers Five Distinct Cancer Cell Groups, Revolutionizing Tumor Characterization and Treatment

ChatGPT Like AI Model Shows Promise in Cancer Treatment Decision-Making

AI Revolutionizes Breast Cancer Diagnosis and Staging

Promising Results in Breast Cancer

To test SEQUOIA's clinical utility, the researchers focused on breast cancer, a well-studied cancer type with established gene signatures correlated to treatment responses and patient outcomes 1

1

The team demonstrated that SEQUOIA could provide genomic risk scores comparable to the FDA-approved MammaPrint test, which analyzes 70 breast-cancer-related genes. Patients identified as high-risk by SEQUOIA showed worse outcomes, including higher rates of cancer recurrence and shorter time before recurrence 2

2

Future Prospects and Limitations

While SEQUOIA shows great promise, it is not yet ready for clinical use. The tool needs to undergo clinical trials and receive FDA approval before it can be used to guide treatment decisions. However, the research team, led by Olivier Gevaert, PhD, a professor of biomedical data science at Stanford, is continually improving the algorithm and exploring its potential applications 3

3

The development of SEQUOIA represents a significant step forward in the integration of AI and genomics in cancer research and treatment. As the tool continues to evolve, it could potentially be applied to all cancer types, offering a new source of valuable data for oncologists and researchers alike.