WO2004077333A1

WO2004077333A1 - Method and apparatus for analyzing image data

Info

Publication number: WO2004077333A1
Application number: PCT/IB2004/050138
Authority: WO
Inventors: Frans W. Zonneveld
Original assignee: Koninklijke Philips Electronics NV
Current assignee: Koninklijke Philips NV
Priority date: 2003-02-28
Filing date: 2004-02-23
Publication date: 2004-09-10
Anticipated expiration: 2005-08-28

Abstract

Method for analyzing image data representing a person's body party comprising a first step of collecting a first image data set and a second step of collecting a second image data set, wherein both the first image data set and the second image data set represent the same body part, whereafter the first image data set and the second image data set are compared to identify differences with respect to a region of interest in said body part, wherein the region of interest is selected to include at least one nodule present in the first image data set, and that the comparison between the first image data set and the second image data set is carried out to identify a growth behavior of the at least one nodule present in the first image data set.

Description

Method and apparatus for analyzing image data

The invention relates to a method and apparatus for analyzing image data.

From WO 98/36683 a method for analyzing image data representing a person's body part is known, comprising a first step of collecting a first image data set and a second step of collecting a second image data set, wherein both the first image data set and the second image data set represent the same body part, whereafter the First image data set and the second image data set are compared to identify differences with respect to a region of interest in said body part.

From this citation also an apparatus is known to carry out the known method. Analysis of the image data with the known method and apparatus results into the identification of abnormalities or deviations in the scanned body structure.

A problem with the prior art method and apparatus is that once an abnormality or a deviation is identified, the issue arises to determine whether the concerning abnormality or deviation, in particular the nodules that are found, are malignant or not.

The invention is intended to provide a method and apparatus, notably a medical workstation by which this determination of malignancy or non-malignancy can be carried out.

The method according to the invention is to this end characterized in that the region of interest is selected to include at least one nodule present in the first image data set, and that the comparison between the first image data set and the second image data set is carried out to identify a growth-behavior of the at least one nodule present in the first image data set. The growth behavior is derived from examing differences between the first and second images. The growth behavior represents the changes, for example in size, over time of the nodule at issue. The change and hence the growth behavior may be postive, negative or neutral. The workstation according to the invention receives the first and image datasets by way of an input module. The image datasets represent the same part of an object to be examined, such as a patient to be examined. The workstation is provided with a selection module to select a region of interest in which at least one nodule is shown, from the image datasets. The workstation is further provided with a comparator which functions to compare the first and second image datasets and identify the growth behavior of the nodule at issue. It is possible that the nodule being examined has disappeared in the second image data set or has remained essentially the same in regard of its characterizing features as compared to the nodule in the first image data set. If in such case the time elapsed between the moment of collecting the first image data set and the moment of collecting the second image data set is large enough, for instance typically three to six months, the conclusion is warranted that the nodule under examination is non-malignant. If, on the other hand, the nodule under examination has doubled in volume in the said timeframe of typically three to six months (the exact timeframe being dependent on the age of the person under examination), the conclusion is warranted that the nodule under examination is malignant. The workstation of the invention is for example provided with a qualifier module which attributes the relevant qualification such as malignant or non-malignant to the nodule at issue. To support the ease of comparison the first image data set with the second image data set, it is desirable that prior to comparing these data sets, the data sets are matched to each other.

Depending on the circumstances it is desirable that matching of the first and the second image data sets includes translating and/or rotating and/or deforming of either one of the first and the second image data sets. In this connection it can be remarked that translating and/or rotating of the image data sets may be required in order to correct for different positioning of the person on the examining table at the time the first image data set and the second image data set are collected. This may particularly be the case when much time has elapsed between the two moments that the image data sets are collected. The deformation of the first image data set or second image data set, may be desirable to take account for differences in breath hold situation; such differences account for possible differences in the location of the region of interest under examination. The workstation of the invention is at option provided with a matching module to perform the matching of the image data sets to one another. The function of matching includes translating and/or rotating and/or deforming of either one of the first and the second image data sets When the method is used to examine a person's lungs, it is desirable that prior to matching the first and second image data sets a dividing of the data sets into sub- volumes is carried out, whereby the said sub- volumes of one image data set jointly represent the person's lungs. By applying this division into sub- volumes, a more accurate matching and subsequent comparison of the regions of interest under examination can be carried out. To perform the selection of sub- volumes the workstation of the invention is provided with a selection module.

Relatively high repeatability of the method according to the invention can be secured when the division into sub-volumes comprises the definition of separation-planes, at least one of which includes the carina of the person's lungs. In this connection it is remarked that the carina identifies with the splitting point of the person's two main stem bronchi.

A second separation plane may be located in the middle of the person's diaphragm or midriff.

In a further aspect of the invention it is desirable that matching of the first and second image data sets covering a region of interest near the person's heart is executed with image data sets representing the person's lungs during end-diastolic phases of the heart. In this way, disturbing factors due to movement of the heart, are effectively obviated.

The invention is further embodied in an apparatus for analyzing image data representing a person's body party, comprising a processor for receiving and processing a first image data set and a second image data set representing said person's body part, said processor including a comparator for comparing the first and the second image data sets and an output organ for reporting a result from said comparison, and which is characterized in that the processor is adapted to operate the method according to the invention. It is noted that the various functions of for example the selection module, comparator, qualifier module, matching module are implemented in software, but they may be implemented in specially designed circuitry, such as in the form of integrated circuits.

The invention also relates to a computer program as defined in Claim 11. Loading the computer program of the invention in the working memory of a general workstation enable the workstation to perform the method of the invention. The computer program of the invention can be made available on a data carrier such as a CD-ROM, or may be downloaded from a data network such as the world- wide web. Finally it is remarked that the invention is also embodied in a software carrier and in software which is designed to control the operation of a processor, so as to implement the method as discussed above.

These and other aspects of the invention will be elucidated with reference to the embodiments described hereinafter and with reference to the accompanying drawing wherein

Figure 1 shows a diagrammatic representation of a workstation in which the invention is employed. The workstation 1 is provided with the input module 11 which receives the first and second datasets. Upon receipt of the first and second datasets they are appliedto the verifying module 111. The verifying module assesses tha the images if the first and second data sets relate to the same body part. Meaningful comparison to assess the growth behavior of a nodule requires the images to concern the same body part. Moreover, the verifying mdoule assesses from the datasets the timesapn that elapsed between the acquisition of the images. Particularly meaningful results are acieved when the elapsed timespan is in the range of three to six months. The first and second data sets are applied to the selection module 12 which selects 121 the sub- volumes form the first and second datasets. For example the selection module is arranged to select sub-volumes that relate to the lung fields of the patient to be examined. The selected sub- volumes of the first and second data sets are fed to the selection module to select 13 the region of interest on the basis of the occurrence of at least one nodule in the region of interest in the first and second datasets, or more precisely in the selected sub- volumes. The respective regions of interest of both first and second datasets are mutually matched by the matching module 14. The matching module 14 notably carries out geometric transformations, such as translation and rotation in order to remove differences that are irrelevant for comparing growth of nodules, but which are causes by differences in the imaging circumstances that applied when the first and second datasets are generated. The matching module 14 derives matched versions of the first and second datasets and applies these to the qualifier module 16. The qualifier module attributes a qualification, such as benign or malignant, to the at least one nodule in the first and second datasets. This qualification is further assigned to the first and/or second datasets by an assignment module 17 which supports a graphical representation of the attributed qualification and outputs the attributed dataset to a monitor 18. On the monitor 18 the attributed dataset is displayed which shows both the anatomy represented by the first or second datasets with the attributed qualification.

Claims

CLAIMS:

1. A workstation for analyzing image data comprising an input module (11) to receive a first and a second image dataset, a verifier module (111) to assess if the first and second image datasets represent the same part of an object, - a selection module (12) to select a region of interest so as to include a representation of at least one nodule in the first image dataset and a comparator (15) to compare the first and second datasets so as to identify differences between the first and second datasets with respect to the region of interest in the object and derive a growth behavior of the at least one nodule from the identified differences.

2. A workstation as claimed in Claim 1, including a qualifier module (16) to attribute a qualification selected from the group malignant, non-malignant to the at least one nodule and to derive the qualification from the growth behavior of the at least one nodule.

3. A workstation according to claim 1 or 2, wherein, the verifier module assess if the first image data set and the second image data set are spaced apart by an amount in the range of three to six months and that the nodule present in the first image data set is qualified malignant if said nodule in the second image data set shows approximately twice the size of the nodule in the first image data set.

4. A workstation as claimed in any one of Claims 1 to 3, including a matching module (14) to - match the received first and second datasets to each other so as to produce a matched pair of first and second datasets and apply the matched pair of first and second datasets to the comparator.

5. A workstation according to claim 4, wherein, matching of the first and the second image data sets includes translating and/or rotating and/or deforming of either one of the first and the second image data sets.

6. A workstation as claimed in any one of Claims 1 to 5, which when in operation receives the first and second image datasets that include the lungs of a patient to be examined and the workstation includes a selection module to select sub- volumes from the received first and second image datasets, which sub- volumes represent the lungs of the patient to be examined.

7. A workstation according to claim 6, wherein, the selection of sub- volumes comprises the definition of separation-planes, at least one of which includes the carina of the person's lungs.

8. A workstation according to claim 7, wherein, one of the separation-planes is located in the middle of the person's diaphragm.

9. A workstation according to any one of claims 6-8, characterized in that, matching of the first and second image data sets covering a region of interest near the person's heart is executed with image data sets representing the person's lungs during end- diastolic phases of the heart.

10. Method for analyzing image data including a first and a second image dataset, representing the same part of an object, the method involving the steps of - selecting a region of interest so as to include a representation of at least one nodule in the first image dataset and comparing the first and second datasets so as to identify differences between the first and second datasets with respect to a region of interest in the object and derive a growth behavior of the at least one nodule from the identified differences.

11. Computer program for analyzing image data including a first and a second image dataset, representing the same part of an object, the method involving the instructions to select a region of interest so as to include a representation of at least one nodule in the first image dataset and compare the first and second datasets so as to identify differences between the first and second datasets with respect to a region of interest in the object and derive a growth behavior of the at least one nodule from the identified differences.