The ARPA-H Precision Surgical Interventions (PSI) program aims to develop methods and techniques to improve cancer detection and increase the visibility of critical anatomical structures during tumor removal surgery.

Funding for awardees varies in amount and is contingent upon the recipient meeting aggressive milestones specific to their project.

The PSI performers are:

• Tulane University will build an imaging system that uses a large aperture camera and structured illumination microscopy, an imaging technique that uses patterned light to achieve high resolution in three dimensions. It takes advantage of light wave interference patterns to image entire excised tumors. The team will also develop an AI algorithm to automatically identify cancerous cells for fast data classification.
• Rice University will build a novel microscope that images tumor slices with ultraviolet epifluorescence. They will use advanced methods to create fluorescent stains that label cells and cellular components and will develop automated AI algorithms to transform their images into ones that look similar to conventional pathology. They will also develop an automated pathology algorithm to classify the imaged cells. 
• University of Washington will develop a microscopy system to allow surgeons to image the entire surface of the tumor by placing it on a lightsheet scanner. The team is also developing algorithms to pseudo-stain the resulting images, so that the sample doesn’t need to be dyed in the operating room; instead, AI methods will take a greyscale image and render it similar to conventional pathology images in order to better classify it.  
• University of California, San Francisco is inventing a microscope that uses an optical array that is pressed into the cavity’s surface. Each pixel is its own multicolor microscope. The investigators are also developing a multi-cancer dyeing agent that activates based on enzyme activity in tumors. 
• University of Illinois Urbana-Champaign will develop optical coherence tomography techniques to find suspicious tissue structures in the surgical cavity, then image those regions with nonlinear optics, which will give a multilayered view of the cells’ metabolism and structural properties. 
• Johns Hopkins University, which is performing on both technical area 1-B and technical area two, will develop a novel non-contact, photoacoustic endoscope to provide a more colorful view of the surgical field without altering the surgeons’ workflow. They will also develop a multi-cancer fluorescent contrast agent.  
• Dartmouth College is creating a laparoscope-integrating imaging solution that will be especially helpful in prostate cancer surgeries. They will use nerve-dyeing and ureter-dyeing contrast agents, in addition to vascular dyes, to cause these critical anatomical structures to fluoresce. They will then map and visualize the 3D shape and depth of the structures.  
• Johns Hopkins University will use existing fluorescent dyes in combination with their novel photoacoustic endoscope to visualize anatomical structures for surgeons. The endoscope will ‘see’ deep into human tissue to reveal hidden blood vessels and nerves, such as they are not accidentally cut. (See above)
• Cision Vision will use shortwave infrared and hyperspectral images to help surgeons visualize blood vessels, nerves, and especially lymphatic structures. Going well beyond red, green, and blue, hyperspectral imaging is enhanced by AI algorithms. This would allow the team to distinguish between tissue types without administering dyes.