{"id":"CU26243","slug":"deep-learning-based-contrast--CU26243","source":{"id":"CU26243","dataset":"techtransfer","title":"Deep learning-based contrast-free MRI enhancement for safer imaging","description_":"<p>This technology is a deep learning framework that generates contrast-enhanced MRI information from pre-contrast scans, enabling safer lesion imaging with reduced reliance on gadolinium-based contrast agents. </p>\r\r<h2>Unmet Need: Noninvasive alternatives to gadolinium-enhanced MRI</h2>\r\r<p>The current standard for enhancing MRI lesion visualization is the use of gadolinium-based contrast agents, but these agents require injection and raise concerns about gadolinium retention, potential toxicity, and added procedural complexity. Despite these limitations, contrast enhancement remains important for accurate lesion detection and assessment. As a result, there is a need for noninvasive approaches that preserve diagnostic contrast without the need to administer contrast agents.</p>\r\r<h2>The Technology: Deep learning-based MRI contrast prediction without injection</h2>\r\r<p>This technology, called DeepContrast, is a deep learning-based imaging platform that predicts contrast-enhanced MRI information from a single pre-contrast T1-weighted MRI scan. The model is trained on paired non-contrast and contrast-enhanced MRI images to learn how enhancement patterns relate to baseline image features. It then generates voxel-level enhancement maps that preserve structural lesion information and support lesion visualization and assessment. By predicting enhancement without administered gadolinium in certain use cases, the technology may enable less invasive MRI workflows.</p>\r\r<h2>Applications:</h2>\r\r<ul>\r<li>AI-predicted contrast-enhanced brain tumor imaging</li>\r<li>AI-predicted contrast-enhanced lesion assessment and visualization</li>\r<li>AI-based MRI lesion analysis</li>\r<li>Reduced-dose or gadolinium-sparing MRI workflows</li>\r</ul>\r\r<h2>Advantages:</h2>\r\r<ul>\r<li>Reduced reliance on gadolinium-based contrast agents</li>\r<li>Preserved diagnostically relevant lesion enhancement information</li>\r<li>Potentially reduced procedural complexity and patient burden</li>\r</ul>\r\r<h2>Lead Inventor:</h2>\r\r<p><a href=\"https://mr.research.columbia.edu/content/jia-guo\">Jia Guo, Ph.D</a></p>\r\r<h2>Related Publications:</h2>\r\r<ul>\r<li><p><a href=\"https://arxiv.org/abs/2001.05551\">Sun H, Liu X, Feng X, Liu C, Zhu N, Gjerswold-Selleck SJ, et al. Substituting gadolinium in brain MRI using DeepContrast. In: 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI); 2020. p. 908-912.</a></p></li>\r<li><p><a href=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC9407020/\">Liu C, Zhu N, Sun H, Zhang J, Feng X, Gjerswold-Selleck S, et al. Deep learning of MRI contrast enhancement for mapping cerebral blood volume from single-modal non-contrast scans of aging and Alzheimer’s disease brains. Front Aging Neurosci. 2022 Aug 11;14:923673.</a></p></li>\r</ul>\r\r<h2>Tech Ventures Reference:</h2>\r\r<ul>\r<li><p>IR CU26243</p></li>\r<li><p>Licensing Contact: <a href=\"mailto:techtransfer@columbia.edu\">Jerry Kokoshka</a> </p></li>\r</ul>\r","tags":["Blood volume","Brain tumor","Deep learning","Lesion","Magnetic resonance imaging"],"file_number":"CU26243","collections":[],"meta_description":"DeepLearning predicts MRI contrast from a single pre-contrast scan, enabling gadolinium-sparing, safer lesion imaging.","apriori_judge_output":"{\"scores\":{\"novelty\":3.0,\"potential_impact\":4.0,\"readiness\":3.0,\"scalability\":3.0,\"timeliness\":4.0},\"weighted_score\":3.6,\"risks\":[\"Tech readiness: research-to-prototype with no published real-world validation yet\",\"Regulatory pathway for MRI software/AI in clinical use\",\"Potential generalization across scanners and institutions\",\"Clinical validation needed for diagnostic equivalence to contrast-enhanced imaging\",\"Regulatory/standards hurdles for safety and interoperability\"],\"one_sentence_take\":\"Promising novelty with meaningful clinical impact potential, but requires real-world validation and regulatory clearance to de-risk adoption.\"}","inventors":["Jia Guo"],"manager":"Jerry Kokoshka","depts":["Psychiatry"],"divs":["Columbia University Medical Center (CUMC)"],"date_released":"2026-04-24"},"highlight":{},"matched_queries":null,"score":0.0}