Quantitative method for unbiased categorization of lung tissue

This technology provides an objective, quantitative, robust method to evaluate lung morphometry and extract unbiased categorization of the lung parenchyma.

Unmet Need: Method to evaluate lung morphometry and categorize the lung parenchyma

Lung morphometry analysis is used to evaluate lung physiology, pathology and development. Techniques and software currently designed to perform morphometry analysis, however, suffer from overly simplistic implementations and biased approaches. The existing parameters, such as mean linear intercept and destructive index, are subjective and limited in their ability to interpret data. Therefore, there is a need for an objective, quantitative, robust method to evaluate lung morphometry and extract and unbiased categorization of the lung parenchyma for better understanding of lung physiology and pathology.

The Technology: Objective, robust method to profile the parenchymal airspace

This technology allows for improved understanding of lung physiology and pathology via an unbiased, reproducible and quantitative approach. To do this, normally time-consuming and subjective tasks are converted to standardized and computerized processes. The combination of machine learning and computer vision provided by this technology make it possible to perform complex analyses, such as accurately defining the actual size and count of alveolus in samples. Therefore, this technology provides an improved method for accurate parenchymal airspace profiling and quantification of parenchymal destruction.

This technology has been validated with images of fixed, inflated, serial sections of a mouse lung and mouse lung with emphysema.

Applications:

• Lung morphometry for clinical and research applications
• Evaluation of biopsy samples
• Image analysis of fixed and sectioned tissue

Advantages:

• Application of machine learning and computer vision technologies lead to more accurate and unbiased quantification of normal vs damaged lung tissue
• Increased sensitivity to changes in images
• Replaces subjective parameters with computational, objective methods for accurate assessment of airspace size distribution and distinction between normal and destructed region

Lead Inventor:

Jeanine D’Armiento, M.D., Ph.D.

Related Publications:

• Xio R, Goldklang MP, D’Armiento JM. “Parenchymal Airspace Profiling: Sensitive Quantification and Characterization of Lung Structure Evaluating Parenchymal Destruction” Am J Respir Cell Mol Biol. 2016 Nov; 55(5): 708-715.

Tech Ventures Reference:

• IR CU16048

• Licensing Contact: Sara Gusik