Leveraging coded ptychography to deliver AI-powered rapid analysis of unstained lung biopsy samples at the point of care

The National Cancer Institute awarded $1,057,841 to Pathware Inc. on September 1, 2026, under the Cancer Detection and Diagnosis Research program (CFDA 93.394) as a Small Business Innovation Research Direct-to-Phase II grant (R44CA302204).

Pathware is developing an automated imaging platform combining coded ptychography with artificial intelligence algorithms to evaluate the diagnostic adequacy of unstained lung biopsy samples at the point of care. The system provides rapid on-site evaluation feedback to proceduralists during biopsy procedures and prioritizes slides for downstream pathology workflow efficiency. The technology aims to reduce repeat biopsies caused by insufficient sample cellularity, which currently delays lung cancer diagnosis and treatment by up to 90 days and increases cancer upstaging risk. By addressing financial and operational barriers that have historically limited rapid on-site evaluation adoption in lung biopsies, the platform seeks to improve early-stage lung cancer detection and treatment outcomes.

Performance occurs in Golden, Colorado, with a period of performance from September 1, 2026, through August 31, 2028.

Abstract

This small business innovation research (SBIR) direct-to-phase II project aims to continue research and development of a novel imaging platform combined with artificial intelligence (AI) algorithms to analyze the adequacy of lung biopsy samples at the point of care (POC). Lung cancer is among the deadliest of all cancers and is the first leading cause of cancer death in the united states. Early diagnosis and treatment of lung cancer is essential, given the survival rate plummets from 63.7% for localized stage cases to 8.9% for distant stage cases. Unfortunately, fewer than half of the cases are identified in early stages.40,41 accurate diagnosis of a suspicious lesion requires the proceduralist to acquire enough sample from the correct location. Unfortunately, many biopsy procedures must be repeated due to a lack of sufficient cellularity for diagnosis-especially for early-stage cases when the target nodule tends to be smaller and therefore harder to access with the biopsy tool. The result of this delay can be devastating on patient outcomes by increasing the time to diagnosis and treatment by up to 90 days, thus increasing the likelihood of the cancer upstaging.2 rapid on-site evaluation (ROSE) of biopsy specimens mitigates the risk of repeat procedures by verifying the diagnostic quality of a sample at the POC during the biopsy procedure. Yet, despite a high rate of non- diagnostic samples, ROSE historically is underutilized in lung biopsies due to financial and operational barriers, including opportunity costs and the lack of on-site pathologist availability. Through NIH grant funding and private resources, pathware is developing technology that autonomously provides rapid biopsy sample feedback to the proceduralist during the biopsy procedure and slide prioritization to pathology for downstream workflow efficiency. This direct-to-phase II project is designed to build upon promising initial proof-of-concept results to continue development and validation of this novel technology with the following aims: 1) optimize a next generation bioptic® system for clinical adoption and AI analysis to non- destructively image unstained lung lesion cytology specimens rapidly and efficiently; 2) collect multi-center clinical data on fresh, unstained lung tissue samples to expand AI model development and training to a large and diversified population; 3) refine autonomous algorithms to determine adequacy of unstained lung qpis to develop a robust AI system for a diverse and representative patient population. Achievement of these aims will enable pathware to take the key next steps of submission for FDA regulatory clearance and subsequent commercialization. This SBIR funding is vital to the development of a product that will improve treatment and outcomes for lung cancer patients by enabling healthcare providers to perform lung biopsies more effectively and by improving the diagnostic pathway for one of the deadliest cancers.