Variable Importance in Nonlinear Kernels (VINK): Classification of Digitized Histopathology

نویسندگان

  • Shoshana Ginsburg
  • Sahirzeeshan Ali
  • George Lee
  • Ajay Basavanhally
  • Anant Madabhushi
چکیده

Quantitative histomorphometry is the process of modeling appearance of disease morphology on digitized histopathology images via image-based features (e.g., texture, graphs). Due to the curse of dimensionality, building classifiers with large numbers of features requires feature selection (which may require a large training set) or dimensionality reduction (DR). DR methods map the original high-dimensional features in terms of eigenvectors and eigenvalues, which limits the potential for feature transparency or interpretability. Although methods exist for variable selection and ranking on embeddings obtained via linear DR schemes (e.g., principal components analysis (PCA)), similar methods do not yet exist for nonlinear DR (NLDR) methods. In this work we present a simple yet elegant method for approximating the mapping between the data in the original feature space and the transformed data in the kernel PCA (KPCA) embedding space; this mapping provides the basis for quantification of variable importance in nonlinear kernels (VINK). We show how VINK can be implemented in conjunction with the popular Isomap and Laplacian eigenmap algorithms. VINK is evaluated in the contexts of three different problems in digital pathology: (1) predicting five year PSA failure following radical prostatectomy, (2) predicting Oncotype DX recurrence risk scores for ER+ breast cancers, and (3) distinguishing good and poor outcome p16+ oropharyngeal tumors. We demonstrate that subsets of features identified by VINK provide similar or better classification or regression performance compared to the original high dimensional feature sets.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Analysis of convergence of solution of general fuzzy integral equation with nonlinear fuzzy kernels

Fuzzy integral equations have a major role in the mathematics and applications.In this paper, general fuzzy integral equations with nonlinear fuzzykernels are introduced. The existence and uniqueness of their solutions areapproved and an upper bound for them are determined. Finally an algorithmis drawn to show theorems better.

متن کامل

Integrated diagnostics: a conceptual framework with examples.

With the advent of digital pathology, imaging scientists have begun to develop computerized image analysis algorithms for making diagnostic (disease presence), prognostic (outcome prediction), and theragnostic (choice of therapy) predictions from high resolution images of digitized histopathology. One of the caveats to developing image analysis algorithms for digitized histopathology is the abi...

متن کامل

The use of radial basis functions by variable shape parameter for solving partial differential equations

In this paper, some meshless methods based on the local Newton basis functions are used to solve some time dependent partial differential equations. For stability reasons, used variably scaled radial kernels for constructing Newton basis functions. In continuation, with considering presented basis functions as trial functions, approximated solution functions in the event of spatial variable wit...

متن کامل

مقایسه دقت تصاویر رادیوگرافی

Statement of Problem: Computer Sciences, in radiology, like other fields, is of high importance. It should also be noted that the accuracy of the technique and work conditions affects the radiographs information considerably. There for, in order to get more accurate diagnostic information, it seems necessary to investigate different digitized radiographic techniques and to compare them with the...

متن کامل

Accounting for secondary variable for the classification of mineral resources using co-kriging technique; a Case study of Sarcheshmeh porphyry copper deposit

Due to substantial effect of classification of resource models on future mine planning, one should come with an accurate method of estimation to guarantee that the minimum error is acquired in the estimation process. The known world class Cu-Mo deposit, Sarcheshmeh Porphyry deposit (central Iran) selected as the study area. The Hypogene zone of the deposit was chosen as the space in which estim...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:
  • Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention

دوره 16 Pt 2  شماره 

صفحات  -

تاریخ انتشار 2013