The Computational Image Analysis Laboratory provides services including (but not limited to) image de-identification and transfer, tumor segmentation and characterization, and image-based therapeutic response assessment using both conventional criteria and advanced methods. Image modalities include CT, MRI, PET, and ultrasound. Our software tools work for both 3D and 2D images.
- Standard and customized
- Thru sftp, DVD/CD, diverse software systems, etc.
- All solid tumors such as tumors in lung, liver, lymph node, kidney, pancreas, prostate
- Organs including lung, liver, liver right/left lobes, spleen, kidney, pancreas, prostate
- Breast: lesion, parenchyma enhancement component
- Brain: core infarct, necrosis, edema, brain and its anatomic structures
- Unidimensional, bidimensional and volume measurements
- 1160 radiomics features
- Visceral fat and subcutaneous fat at any specified anatomical level and over a pre-defined body range
- RECIST 1.0/1.1, WHO, Choi, tumor volume, necrosis
- SUVmax, SUVave, total lesion glycolysis (TLG), functional lesion volume
- RANO criteria
- Radiomic features (1160 quantitative image features) and prediction models using multiple ML methods