American diagnostic startup Droplet Biosciences has announced a collaboration with technology giant NVIDIA. As part of the partnership, the company will use NVIDIA’s artificial intelligence (AI) infrastructure to accelerate the detection of residual cancer cells following surgery, Reuters reports.
The company is applying the NVIDIA Parabricks software suite, accelerated by graphics processing units (GPUs), to analyze genomic data from DNA sequencing.
The method developed by Droplet Biosciences can detect residual disease within 24 hours by analyzing lymphatic fluid collected immediately after surgery. In comparison, traditional blood tests can take four to six weeks to identify remaining tumor cells. According to Droplet’s Chief Scientific Officer, Wendy Winckler, using NVIDIA Parabricks has reduced the most resource-intensive analysis steps from over a day to just a few hours.
The company noted that while GPU-based computing has a higher hourly cost, this approach offers significant operational advantages. The drastic reduction in processing time ultimately lowers the total cost per sample analyzed.
The rapid turnaround allows patients to receive their results while still in the hospital, avoiding the need for additional clinic visits or the long wait associated with standard blood tests.
The startup’s initial clinical test is designed to detect human papillomavirus (HPV)-negative head and neck cancer. It has been approved under the Clinical Laboratory Improvement Amendments (CLIA), the US federal standards governing laboratory testing on human specimens.
Earlier, it was reported that pharmaceutical giant Eli Lilly is also betting on AI. The company launched a supercomputer in partnership with NVIDIA to accelerate drug development. According to developers, the NVIDIA DGX SuperPOD system, featuring the latest DGX B300 chips, is unmatched in the world.
