- Salesforce collaborated with Ellison Institute of Transformative Medicine
- ReceptorNet is an algorithm that scans mammography for signs of cancer
- ReceptorNet detected potential cancer signs with 92% accuracy
Every year more than two million women are affected by breast cancer. The rates of cancer are increasing alarmingly all around the world. But, the risk can be mitigated if signs of cancer are discovered at an early stage. To address this issue, a team at Salesforce Research came up with a machine learning system named ReceptorNet.
Salesforce Research is an American cloud-based software company. Recently, it has collaborated with Ellison Institute which developed the deep-learning algorithm. ReceptorNet can determine the hormone-receptor status, which is a critical biomarker for oncologists when deciding on an appropriate treatment for breast cancer patients. ReceptorNet achieved 92% accuracy thanks to its excellent sensitivity and specificity numbers.
The work has been published in the journal ‘Nature Communications’ with the title ‘Deep learning-enabled breast cancer, hormonal receptor status determination from base-level H&E stains.’ The method used by Salesforce employs artificial intelligence to try to improve outcomes for breast cancer patients. Other similar efforts like Google’s AI breast cancer screening tool have also focused mainly on diagnosing cancer. Google Health’s AI model was trained on over 90,000 mammogram X-rays and produced better results than human radiologists. It could also recognize false negatives – ones that look normal but contain breast cancer.
What sets ReceptorNet apart from others is its focus on improving the way treatment decisions are made for breast cancer patients. Its prediction of hormone-receptor status from an inexpensive and easily available tissue image is exemplary. Typical methods involve the extraction of breast cancer cells during a biopsy or surgery and tested against certain proteins. When estrogen and progesterone hormones are attached to these receptors, cancer starts growing.
About ReceptorNet
The development of the AI-powered tool began when Salesforce researchers and Dr. David Agus started discussing the hidden information within tumor cells. Together they collaborated to develop this technology with the view of creating a positive impact and not just serve the purpose of the company. Salesforce believes that this tech if deployed clinically, could help reduce the cost of care and time required to begin cancer treatment. The AI-powered tool will also improve the overall accuracy and deliver better health outcomes for patients.
