Voxel-based methods for assessing subjects using magnetic resonance imaging

Patent metadata

Field	Value
Publication number	US-2016239968-A1
Application number	US-201514869862-A
Country	US
Kind code	A1
Filing date	Sep 29, 2015
Priority date	May 28, 2008
Publication date	Aug 18, 2016
Grant date	—

Abstract

The subject matter disclosed herein relates to methods for diagnosing a neurological disorder in a subject. In certain aspects, the methods described herein involve determining one or more critical areas in the brain from PET data where two groups differ and measuring PET signal within determined critical areas in a new subject in order to assign risk or diagnosis.

First claim

1 . A computer-implemented method for diagnosing or determining risk of a neurological disorder in a subject, the method comprising:
(a) determining the presence of a radioactive signal from an administered radioligand in at least one individual in a control group and at least one individual in a reference group to generate primary brain scan image voxel data of radioligand distribution in a brain of at least one individual in a control group and at least one individual in a reference group, (b) generating secondary brain scan image data for the individuals in the control and reference groups, wherein the secondary scan brain image data is generated using a different type of brain scan than the primary brain scan image data, (c) generating a probability-corrected time-activity curve data for each voxel in the primary brain scan image of the at least one individual in the control and reference group, (d) processing the probability-corrected time-activity curve data of the at least one individual in the control and reference group to generate a voxel binding outcome map data of the at least one individual in the control and reference group, (e) transforming the voxel binding outcome map data of the at least one individual in the control and reference group into a normalized space to generate a normalized voxel binding outcome map data of the at least one individual in the control and reference group (f) processing the normalized voxel binding outcome map data of the at least one individual in the control and reference group using statistical analysis to identify one or more voxels of interest (VOI) in the normalized voxel binding outcome map data to generate a VOI map data for differentiating of the at least one individual in the control and reference group, and (g) applying the VOI map data to a voxel binding outcome map data of a test subject to generate a mean cortical binding value to diagnose or determine risk of a neurological disorder in the subject wherein the voxel binding outcome map data of said test subject is generated from Magnetic Resonance Imaging (MRI) using a MRI scanner.

2 . The method of claim 1 , wherein the transforming of step (e) comprises transforming the voxel binding outcome map data of the at least one individual in the control and reference group into a secondary scan space to generate a secondary space voxel binding outcome map data of the at least one individual in the control and reference group.

3 . The method of claim 1 , wherein step (g) comprises:
(i) inverse transforming the VOI map data identified in step (f), into a secondary scan space of the test subject to generate a voxel of interest (VOI) mask for the test subject, (ii) multiplying the VOI mask for the subject by probabilistic brain region (BRP) map data for the subject to generate a brain region VOI mask for the subject, (iii) multiplying a secondary space voxel binding outcome map data of the subject by the brain region VOI mask for the subject to generate cortical binding map data for the subject, and (iv) summing the cortical binding map data of the subject and then dividing it by the sum of the brain region VOI mask to generate a mean cortical binding value.

4 . The method of claim 3 , wherein the method further comprises a step of processing the secondary space voxel binding outcome map data of the subject by partial volume correction analysis.

5 . The method of claim 1 , wherein the generating a probability-corrected time-activity curve data for each voxel in the primary brain scan image of the at least one individual in the control and reference group in step (c) comprises:
(i) processing the secondary brain scan image data of the at least one individual in the control and reference group to generate a binary brain region mask and probabilistic brain region (BRP) map data for each individual, and (ii) processing the probabilistic brain region (BRP) map data and the primary brain scan image data of the at least one individual in the control and reference group onto the binary brain region mask of the individual to generate the a probability-corrected time-activity curve data for each voxel in the primary brain scan image of the at least one individual in the control and reference group.

6 . The method of claim 2 , wherein the normalized voxel binding outcome map data of the at least one individual in the control and reference group is generated by transformation of secondary space voxel binding outcome map data of the individual into a standard brain atlas.

7 . The method of claim 6 , wherein the standard brain atlas is a Talairach brain atlas or a Montreal Neurological Institute (MNI) brain atlas.

8 . The method of claim 6 , wherein the standard brain atlas is a specific brain atlas created for a particular neurological disorder.

9 . The method of claim 6 , wherein the standard brain atlas is a custom brain atlas.

10 . The method of claim 3 , wherein the inverse transforming is performed using parameters from an MRI to standard brain atlas registration.

11 . The method of claim 1 , wherein the processing in step (f) comprises:
(i) generating a binary voxel image mask of the at least one individual in the control and reference group by statistical parametric mapping analysis, (ii) inverse transforming the binary voxel image mask of the at least one individual in the control and reference group into a secondary space voxel binding outcome map data of the individual to generate a voxel of interest (VOI) mask, (iii) multiplying the VOI mask of the at least one individual in the control and reference group by the individual's probabilistic brain region (BRP) map data and secondary space voxel binding outcome map data to generate cortical binding map data for the individual, (iv) dividing the sum of the cortical binding map of the at least one individual in the control and reference group and the reference group by the mean of the probabilistic brain region (BRP) map data of the individual to generate a mean cortical binding outcome value for the statistical parametric mapping analysis applied in step (i), (v) performing statistical analysis between the mean cortical binding outcome values of the at least one individual in the control and reference group to generate a map assigning a probability value to each voxel, and (vi) identifying a scoring threshold providing maximal separation of mean cortical binding outcome values between the at least one individual in the control and reference group, wherein the VOI corresponding to the scoring threshold providing maximal separation of mean cortical binding outcome between individuals from the control group and individuals from the reference group is a VOI map data suitable for differentiating individuals in the reference group from individuals in the control group.

12 . The method of claim 3 or 11 , wherein the secondary space voxel binding outcome map data is generated using a different type of brain scan than the primary brain scan image data used in step (a) of claim 1 .

13 . The method of claim 11 , wherein the statistical analysis in step (v) is a Student's t test.

14 . The method of claim 11 , wherein the generating of the binary voxel image mask in step (i) comprises applying one or more threshold values are to the normalized voxel binding outcome map data such that, for each threshold, data in the voxel binding outcome map data equal to or exceeding the threshold value are retained in the binary voxel image mask and data in the voxel binding outcome map data less than the threshold value are not retained in the binary voxel image mask.

15 . Th

Assignees

• Univ Columbia

Inventors

• Parsey Ramin
• Mikhno Arthur

Classifications

• A61B5/0042

  for the brain · CPC title

• G01R33/0023

  Electronic aspects, e.g. circuits for stimulation, evaluation, control; Treating the measured signals; calibration (G01R33/0017 takes precedence) · CPC title

• G06T2207/20128

  Atlas-based segmentation · CPC title

• G01R33/10

  Plotting field distribution {; Measuring field distribution} · CPC title

• G06T2200/04

  involving 3D image data · CPC title