We developed numerous advanced image analysis algorithms for efficient, accurate and robust segmentation and quantification of organs (liver, spleen, kidney, etc.), lesions (lung lesion, liver metastasis, lymph nodes, brain tumor, breast cancer, etc.) and other diseases / tissues (emphysema, lung cyst, body fat, etc.) on multiple imaging modalities such as CT, MRI and/or PET/ CT. We are always working to improve our algorithms and develop novel ones for new clinical applications.

We have two patents and three filed patent applications for our in-house developed organ and lesion segmentation algorithms, lung nodule and lymph node detection algorithms and quality assurance software.

We are Cycle 1 Awardees of Columbia-Coulter Translational Partnership.
Project Title: Evaluating Tumor Response to Cancer Therapy
Drs. Binsheng Zhao and Lawrence Schwartz are developing CT and MRI software to accurately measure subtle changes in tumor volume and density. Ultimately, their technology will enable doctors to better gauge tumor responses to specific cancer treatments. The algorithms are being optimized and converted into suitable formats for rapid integration with commercial imaging systems.

We are Cycle 3 Awardees of Columbia-Coulter Translational Partnership.
Project Title: Novel software for quantification of core infarct volume in stroke
Dr. Binsheng Zhao's group and Dr. Christopher Filippi are developing brain MR neuroimaging software to reliably quantify core infarct volume on diffusion weighted MRI in stroke. The envisioned product will be efficient ready-to-be-licensed software for segmentation / quantification of core infarct volume in stroke to better inform diagnosis, treatment option and response, and prognostication. It will be easy-to-use, reliable, and readily integrated into the routine clinical workflow of all hospital PACS workstations.

Our available technologies can be found at:


News Reports:

  • October 23, 2015
  • Hybrid lung CAD scheme improves accuracy by AuntMinnie.com
    A hybrid computer-aided detection (CAD) algorithm for lung nodules developed by Columbia University showed high sensitivity with very few false positives, according to new research in Medical Physics.

  • October 9, 2014
  • "Tumor Segmentation Software Receives 510k Clearance From FDA"

    Press Releases:

  • February 6, 2015
  • "Imbio Signs License Agreement for Brain Tumor Segmentation Algorithm with Columbia University"

  • April 4, 2014
  • "Hinacom Licenses 3-D Lesion Segmentation Software from Columbia"

  • May 16, 2013
  • "Columbia Licenses Novel 3-D Organ and Tumor Segmentation Software to Varian Medical Systems"

  • November 12, 2012
  • "AG Mednet Licenses Software from Columbia University to Reduce Clinical Trial Protocol Amendments and Unnecessary Costs"