Our web-based computer-aided detection is under construction. However we would like to introduce our technology to you.

The Web Second Reading^{sm} services will allow the user to ret second opinion on the matter without visiting radiologist or seating in a waiting line. The user will login onto this page and then will be able to submit his/her digital image to our server. The application interface will be deployed with "feel and touch" of our standalone image analysis and computer-aided detection system known as Image Companion^{™}. Users will have a choice to quantify the loaded image by himself or submit it for a professional staff radiologist available through the site.

The results of computer-aided detection:

Will be displayed to the user;

Will be downloadable to the user who submitted the image;

Will not be stored or saved anywhere on our server unless such storage will be requested by the user;

Will be exportable in numeric form;

Will be available for further statistical analysis;

Will be available to be compared with databases of specific diseases located on this web site.

An ability to compare and calculate similarities between the results of Web Second Reading^{sm} and existing database of diseases using implemented Web Sleuth^{™} (a web version of Med-Sleuth^{™}) and Web Detective^{™} (a web version of Med-Detective^{™}) technologies will revolutionarize the whole world of biomedical data and knowledge retrieval and comparison. We are the first to put Web Radiologist^{sm} in the hands of our users.

You can also use free on-line application of Data Companion^{®} - the statistical analysis system - ,which is currently available freely at the Free On-line Application section.

Statistical Free On-line application Short tutorial for On-line statistical analysis

General Overview
The statistical analysis on-line allows user to process it's data in real time. Input data usually represents field observations. Very often the data is derived from various kinds of processing with numerical results output form. Data Companion Lite^{™} is a powerful tool producing fast and accurate statistical results such as correlation matrix, rank correlation, multiple linear regression, credibility values, etc. The results are produced for an original input matrix or for preprocessed matrices of the differences or logarithms. In the case of absolute difference matrices, each next row value is subtracted from the current one. The number of rows in that case is reduced by one. In the case of a logarithms matrix, the output matrix contains natural logarithms of correspondent initial values of the input matrix. Such additional and non-traditional matrices of data are common in cases where the numeric values of data deviate substantially, or the user is more interested in trends of the acceleration changes instead of acceleration itself.

Parameters Overview
The data should be typed-in row by row. User can use a "Tab" button to move from editable window to window. The number of observations represent a total number of rows in the user's input data. Number of rows must be greater than the doubled number of factors (columns) { number of rows >= 2 x number of columns + 1 }. Each column should have an equal number of observations inserted. User may be interested to average the number of observations (samples, measurements) of the input data. This is a common case in biology, meteorology, climatology, medicine etc. For that case Data Companion Lite^{™} enables user to manipulate its full original data by using averaging by test capability. If user sets parameter "Number of observations in each test" to any value larger than one { > 1 } the system will substitute the original matrix with a new one. Each row value of such new matrix will represent a mean value calculated from a few rows. The rows for averaging are selected starting from the first row and are equal to the number set by user through parameter "Number of observations in each test". When this option is not used the "Number of observations in each test" is equal to one { 1 }, "Number of observations" is equal to "Number of tests" and both parameters are practically identical. For the purpose of regression calculations it is assumed that the first column represents the response function and the rest of the columns assumed to be factors.