Broad Goals
- Devising methods to quantify image features
- Assessing methods to quantify lesions and tissues in morphological, textual, and material biomarkers
- Devising task-based radiomics
- Relating image-based biomarkers to clinical outcome
Medical imaging technologies improve patient care by providing both qualitative visualization of internal human anatomy and quantitative data. Quantitative data, such as mathematical descriptions of object size, shape, internal heterogeneity, and radiomics features, are being explored by the medical imaging community as biomarkers for patient outcomes. In our lab, the purpose of quantitative imaging research is to ensure that the metrics extracted from images, including radiomics, are valid biomarkers which are measurable, relevant, and objective.
Quantitative biomarkers quantitatively describe patient phenotypical attributes, which are relevant to patient outcomes and are objective measurements relating to the truth of the patient. Current projects include development of truth-based lesion models, development of truth-based reference datasets, and metrology for lesion quantification. Truth-based lesion models and reference data are used to assess the objectivity, precision, and accuracy of quantification across different measurement tools, medical imaging scanners, segmentation tools, and observers. Lesion quantification metrology is currently being studied to determine which quantifications are objective and relevant to patient outcomes.
Related Publications
- Volumetric quantification of lung nodules in CT with iterative reconstruction (ASiR and MBIR)
- Quantitative features of liver lesions, lung nodules, and renal stones at multi-detector row CT examinations: dependency on radiation dose and reconstruction algorithm
- Quantitative breast tomosynthesis: from detectability to estimability
- Techniques for virtual lung nodule insertion: volumetric and morphometric comparison of projection-based and image-based methods for quantitative CT
