REWIRE NEUROSCIENCE LLC — Department of Health and Human Services SBIR Phase I: 400

Description

ABSTRACT Manual analysis of biomedical images by researchers and pathologists has the potential to introduce bias and error that compromise the reliability of research and clinical findingsThese problems are significant barriers to delivering the most beneficial evidence based medicinedeveloping effective medical treatmentsand promoting confidence in scientific inquiryIdentification of biomarkers and cellular targets following microscopy requires manual analysis of biomedical imageswhich is time intensivedifficultand prone to bias and errorsUnintentional bias and attentional limitations during analysis of biomarkers can underlie poor reproducibility of findings in biomedical research and potentially introduce error in clinical diagnosticsWe recently developed abetasoftware package designed to improve automation and standardization of image analysiscalledPIPSQUEAKPerineuronal net Intensity Program for the Standardization and Quantification of Extracellular matrix Analysis KitSince its publication inPIPSQUEAK beta has amassed approximatelyusers worldwide who use it to quantify the intensity and number of perineuronal nets and other neural markers in the brainThis technology significantly increases data reliability between image raters and decreases the time required for analysis by more thanfoldHoweverPIPSQUEAK beta currently uses target detection algorithms that require high contrast images to automatically identify neurons as clusters of bright pixels on dark backgroundsA significant current limitation to PIPSQUEAK betaand other available imaging programsis that detection of biomarkers can be difficult unless image conditions are idealSuboptimal conditionslike high background stainingoff target structuresoverlapping or clustered biomarkersand atypical morphologiescan lead to artifacts and consequently to inaccurate results and erroneous conclusionsHerewe propose to develop a user friendly artificial intelligenceAIplatform for the automated detection of targeted biomarkers in digital microscopy that reduces this error by learning to distinguish between true cellular biomarkers and artifactsWe propose to integrate AI capabilities into our PIPSQUEAK technology to produce an adaptivehigh throughputbiomedical image analysis platform that quickly and accurately identifies biomarker targets from bench to bedsideA key advantage is that this AI program will be user friendly and available onlinemaking it highly accessible to basic researchers and to technicians and clinicians identifying human pathologiesThussuccessful development of our AI program has a high translational potentialThe goal of this proposal isto develop and validate a machine learning model that is capable of detecting common histological marker morphologies in digital microscopyandto test the feasibility of adapting our AI platform to new biomarker datasets with minimal additional supervised trainingOur end goal is to advance the reliability and speed of research findings and clinical diagnoses by making this technology widely available to researchers and clinicians PROJECT NARRATIVE Manual analysis of biomedical images by researchers and pathologists has the potential to introduces bias and error that compromise the reliability of research and clinical findingsproblems which are significant barriers to delivering the most beneficial evidence based medicine and developing effective medical treatmentsApplication of artificial intelligence for the detection of disease or cellular targets has the potential to improve the reliability of research findings and clinical diagnoseswhile reducing wastetimeand expenseWe propose a method to improve the quality of biomedical research reproducibility and clinical diagnoses by developing a high throughputadaptive artificial intelligence platform for automated analysis of cellular and disease targets in digital microscopy imageswhich will be made available to scientists and clinicians as a user friendly analysis platform