Quantitative Image Analysis and Shape Analysis
We're interested in developing statistical methodology for evaluating effectiveness of quantitative image biomarkers (such as volumetrics, 2D area and 1D largest diameter--RECIST) in inter-laboratory benchmark studies with partial or full ground-truth information. We're also interested in statistically sound image analysis algorithms for segmentation and volume or shape measurements. Our work in functional data analysis and nonparametric regression such as curvature estimation proved to be useful for image segmentation and shape analysis.

Useful Resources:

• Imaging Response Criteria in Clinical Trials.
• NIH Cancer Image Clinical Reference Data.
• NIST Biochange 2008: Benchmarking Pilot Study of Lung Cancer CT Change Analysis.
• NIH's Cancer Image Program.
• ITK Medical Image Segmentation and Registration.
• RSNA. QIBA: Quantitative Imaging Biomarker Standard Activity. Nodule Volume Measurement in Lung CTs. Biomedical engineering as a discipline.
• Medical Physics. Academic Radiology. Radiology.
• Digital imaging processing. Principles of Computerized Tomographic Imaging .
• Conferences: MICCAI. IPAM. SPIE.

Related Publications:

• Z.Q. J. Lu, N. Lowhorn, W. Wong-Ng, W. Zhang et al (2009). Statistical analysis of a round-robin measurement survey of two candidate materials for a Seebeck coefficient Standard Reference Material. NIST Journal of Research, Vol.114, No.1, 37-55.

• N. Lowhorn, W. Wong-Ng, W. Zhang, Z.Q. Lu et al (2009). Round-robin measurements of two candidate materials for a Seebeck coefficient Standard Reference Material. Applied Physics A, Volume 94, Number 2, February 2009 , pp. 231-234(4).

• Z. Q. Lu (2007) Pattern Theory: From Representation to Inference (by Ulf Grenander and Michael Miller), Book Review. Journal of Applied Statistics, Vol.34, No.6, 763-763. August 2007.

• Z.Q. John Lu (2007). Toward 3-D Topography Using Curvature-based Geometry Measuring Machine. The First Workshop on 3D and 2D Content Representation, Analysis and Retrieval in the context of Interoperability Week (April 23-25 2007).

• Z.Q. John Lu (2007). Stereology for statisticians (Adrian Baddeley, Jensen EBV), Book Review, Statistical Methods in Medical Research, Vol. 16, No. 4, 377-378 (2007).

• N. Machkour-Deshayes, J. Stoup, Z. Q. Lu, J. Soons, U. Griesmann, R. Polvani (2006) Form-profiling of Optics Using the Geometry Measuring Machine and the M-48 CMM at NIST. Journal of Research of NIST, September-October 2006, Vol.111, No.5, pp.373-384.

• Z.Q. Lu (2002). Local Polynomial Prediction and Volatility Estimation in Financial Time Series. Chapter 5 of Modeling and Forecasting Financial Data (Kluwer, 2002), pp.115-135. In which I gave a survey of work on variance estimation using local polynomial regression and a general regression framework.

• Z.Q. John Lu, D.B. Clarkson (1999). Monotone spline and multidimensional scaling. Proceedings of American Statistical Association, Section of Statistical Computing, 185-190.

• Nonparametric regression with singular design. Journal of Multivariate Analysis, 1999, Vol.70, 177-201.
  In which I propose and develop a singular design framework for nonparametric estimation. This approach extends to regression models with covariates in high-dimensional space, or with nonlinear confounding, or manifold or curved design. Note that the same theory applies to nonparametric functional data analysis, i.e. regression models in which the covariates take on discretely-sampled measurements of curves in some functional space.

• Multivariate local polynomial fitting for martingale nonlinear regression models. Annals of Institute of Statistical Mathematics, Vol.51, No.4, 691-706. December, 1999.
  I propose a general model for nonlinearly dependent time series data which may be more suitable for some real data applications such as economic or financial data.

• Multivariate locally weighted polynomial fitting and partial derivative estimation. Journal of Multivariate Analysis, 1996, 59, 187-205.
  Interestingly, Canny's edge detection method is based on partial derivative estimation and the statistical theory developed in this paper should be very relevant.

---

Back to Research
Go to Lu' main page.