By Brittany Wade
May 31, 2022 | Crosscope, Silicon Valley, and Farcast Bioscience, Florida, created OrionAI–an artificially intelligent (AI) diagnostic assistant for tumor histological examination. Crosscope developed the AI-enabled digital pathology platform and proprietary algorithms, while Farscape Bioscience created the histoculture platform that preserves entire human tumor microenvironments.
Both companies are particularly interested in examining head and neck squamous cell carcinoma (HNSCC). According to the Indian Journal of Medical and Paediatric Oncology, it accounts for nearly one-third of all cancers in India (DOI: 10.4103/ijmpo.ijmpo_252_17). This technology–a combination of digital pathology, AI, and machine learning–alleviates the Indian cancer burden by increasing diagnostic speed and significantly lowering misdiagnosis rates.
“This collaboration will improve the power of digital transformation in accelerating drug discovery and driving new breakthroughs… in better cancer diagnosis,” said Dr. Jayendra Shinde, CEO of Crosscope in a press release.
Squamous cell carcinoma of the head and neck is one of the most prevalent cancers in India due to the widespread use of tobacco. Worldwide, HNSCC makes up 90% of head and neck cancers, making it the sixth leading cancer by incidence. Squamous cell DNA mutations cause abnormal replication, which leads to tumor formation. However, when detected early, HNSCC is often treated successfully. Innovations like OrionAI bridge the gap created by a lack of available pathologists, making early detection feasible and a regular occurrence.
Digital Pathology and Deep Learning
Digital pathology extracts data from digitized specimen slides to diagnose and treat disease. OrionAI uses digital hematoxylin and eosin slides to detect and quantify significant tumor pathology hallmarks. Predefined algorithms perform histopathological analyses that pinpoint predictive biomarkers for several cancer types. “Crosscope’s deep learning-based OrionAI is dedicated to accurately detecting and evaluating clinically-relevant findings in digitized slides from cancer biopsies,” says Dr. Shinde.
When digital pathology combines with deep learning–a type of machine learning that creates artificial networks to extract high-level information from data–pathologists can predict disease progression, patient prognosis, and effective therapy options.
Human Tumor Microenvironments
Perhaps the most impressive component of OrionAI is its ability to preserve human tumors histologically. Farcast’s technology recreates the tumor immune microenvironment (TiME) ex vivo to maintain morphology and native architecture as tumor drug response is measured.
“TiME is a… platform that has demonstrated the preservation of all elements of the human tumor microenvironment for up to 72 hours. Digitization and AI-based histopathology assessment will go a long way in improving the speed, efficiency, and robustness of our pathology workflows,” said Farcast Biosciences CSO Satish Sankaran.
With an accurate depiction of tumor cells, immune cells, blood vessels, and the extracellular matrix coupled with AI-facilitated histopathological analysis, pathologists will be well-suited to detect and diagnose HNSCC and other cancers across India and beyond.