Fibresolve

Fibresolve is an AI-powered software utilizing convolutional neural networks to analyze chest CT scans for suspected interstitial lung disease by detecting key fibrotic patterns like honeycombing and traction bronchiectasis. It aims to improve noninvasive diagnosis and risk stratification for conditions such as idiopathic pulmonary fibrosis through a 'digital biomarker' score, potentially reducing the need for biopsies and aiding multidisciplinary evaluation.

Fibresolve is an advanced artificial intelligence (AI) software designed for the detailed analysis of chest CT scans in patients suspected of having interstitial lung disease. It processes 3D CT images to classify subtypes of fibrotic lung disease, with a particular emphasis on idiopathic pulmonary fibrosis, by identifying specific imaging patterns such as honeycombing, traction bronchiectasis, and ground glass opacities. The software provides a ‘digital biomarker’ score, intended to assist pulmonologists with risk stratification and in deciding whether a multidisciplinary evaluation is warranted for the patient. The core technology of Fibresolve is built upon convolutional neural networks (CNNs) trained using extensive datasets of interstitial lung disease cases. These networks perform segmentation of lung anatomy and quantify the extent of fibrotic features identified in the CT scans. Clinical studies have supported its reliability in distinguishing patterns characteristic of idiopathic pulmonary fibrosis. Notably, Fibresolve is recognized as the first Breakthrough-designated AI tool for lung fibrosis, offering a significant contribution towards noninvasive diagnosis and potentially lessening the necessity for invasive procedures like lung biopsies.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does Fibresolve's performance vary when analyzing CT scans from different manufacturers or with varying acquisition parameters?; What is the specific impact of Fibresolve's 'digital biomarker' on long-term patient outcomes or treatment selection for interstitial lung diseases?; How does the software differentiate between idiopathic pulmonary fibrosis and other fibrotic ILDs that may present with overlapping imaging features like honeycombing?; What are the sensitivity and specificity of Fibresolve in detecting early or subtle signs of fibrotic lung disease compared to expert radiologist interpretation?; How does Fibresolve account for or mitigate the influence of co-existing pulmonary conditions (e.g., emphysema, infection) on the quantification of fibrotic patterns?