Method and apparatus for early warning of cerebral vasospasm after subarachnoid hemorrhage
The invention is directed to a method and apparatus for early warning of cerebral vasospasm after subarachnoid hemorrhage using brain imaging equipment to identify blood leakage within the subarachnoid space to detect brain zones with increased risk of cerebral vasospasm.
1 . An apparatus for predicting a cerebral vasospasm comprising;
a CT scanner configured to take CT images of a patient;
a computer system with a processor and a database of reference tomography images of segmented CT, CTA and MRI images;
the processor is configured for conducting computational fluid dynamics calculations, for conducting automatic tomography image segmentation, to receive uploaded brain tomography images, and to extract 3D information of leaked blood, cerebral arteries and subarachnoid space;
the processor is further configured to create a database of automated segmented images, process extracted 3D information on leaked blood, cerebral arteries and subarachnoid space (SAS) using computational fluid dynamics calculations to identify time-dependent wash out and spreading of leaked blood breakdown products (BBPs) within subarachnoid space, and to display to a user the results of the automatic tomography image segmentation with calculated 3D distribution of BBPs concentration on cerebral arteries walls and subarachnoid space locations.
2 . The apparatus of claim 1 wherein the automatic tomography image segmentation includes a machine learning algorithm for the automatic segmentation of CT and CTA images.
3 . The apparatus of claim 1 wherein the processor is configured to permit a user to manually alter the automatic tomography image segmentation.
4 . The apparatus of claim 3 wherein the processor is configured to add manually altered automatic tomography image segmentations into the database of automated segmented images.
5 . The apparatus of claim 1 wherein the processor is further configured to calculate a time-dependent spatial distribution of BBPs concentrations from the calculation of cerebrospinal fluid (CSF) flow and BBPs transport.
6 . The apparatus of claim 5 wherein the processor is configured to use the calculated BBPs concentrations to locate zones with higher risk of cerebral vasospasm (CV) and delayed cerebral ischemia (DCI).
7 . The apparatus of claim 6 wherein the processor is configured to use the distribution of BBPs concentration to provide a 3D map of CV and DCI probability.
8 . The apparatus of claim 7 wherein the processor is configured to provide a 3D map of CV and DCI probability using the following equation:
P
risk
(
r
a
,
t
)
=
F
(
c
i
(
r
-
r
a
,
t
)
,
D
a
(
r
a
)
,
C
a
(
r
a
)
)
where
t represents time;
r represents a spatial position vector;
r a represents a spatial position vector pointing to a location on a cerebral arterial wall;
P risk (r a , t) represents a time-dependent three-dimensional map of the probability of cerebral vasospasm and delayed cerebral ischemia;
c j (r-r a , t) represents a time-dependent concentration of a blood breakdown product species at a spatial location offset from the cerebral arterial wall position r a ;
D a (r a ) represents a diameter of the cerebral artery at position r a ;
C a (r a ) represents a calcification level of the cerebral artery at position r a ;
F represents a function defining an integrative relationship between the probability of delayed cerebral ischemia and cerebral vasospasm at position r a and the parameters c i , D a , and C a .
9 . The apparatus of claim 1 wherein the processor is configured to issue a warning of cerebral vasospasm and delayed cerebral ischemia by determining risk zones with a high concentration of BBPs close to specific cerebral arteries at specific times.
10 . The apparatus of claim 1 wherein the processor is configured to generate a segmentation mask which is an array that is an output of the neural network and the algorithm to indicate the likelihood that the location corresponding to a pixel belongs to the hemorrhage.
11 . The apparatus of claim 1 wherein the process is configured to output an original CT scan of a patient and a pixel array of the same dimensions of the CT scan, each pixel in the pixel array has only values 0 or 1, where the value 1 of a pixel indicates the corresponding pixel of the CT scan is predicted to be hemorrhage location.
12 . An apparatus for predicting a cerebral vasospasm comprising;
a CT scanner configured to take CT images of a patient;
a computer system with a processor and a database of reference tomography images of segmented CT, CTA and MRI images;
the processor is configured for conducting computational fluid dynamics calculations, for conducting automatic tomography image segmentation, to receive uploaded brain tomography images, and to extract 3D information of leaked blood, cerebral arteries and subarachnoid space (SAS);
the processor is further configured to process extracted 3D information on leaked blood, cerebral arteries and subarachnoid space using computational fluid dynamics calculations to identify time-dependent wash out and spreading of leaked blood breakdown products within subarachnoid space, and to display to a user the results of the automatic tomography image segmentation with calculated 3D distribution of blood breakdown product concentration on cerebral arteries walls and subarachnoid space locations.
13 . A method for predicting a cerebral vasospasm comprising; acquiring CT, CTA and MRI images of leaked blood within the subarachnoid space (SAS) of a patient,
extracting a 3D spatial distribution of leaked blood from within the SAS from CT and MRI images,
calculating the volume of leaked blood using a manually segmented CT image,
extracting from the CTA and MRI images a 3D distribution of main cerebral arteries,
extracting a 3D image of the SAS from the MRI images,
using a collection of segmented CT, CTA and MRI images to create a database used to train an automatic segmentation algorithm,
using the automated segmentation algorithm to automatically segment CT and CTA images,
calculating a generation of blood breakdown products (BBPs) from the spatial distribution of the leaked blood,
calculating wash out and spreading of BBPs within the SAS by calculating the generation of BBPs in a leaked blood clot, BBPs diffusion and advection by cerebrospinal fluid (CSF) flow within the SAS,
identifying critical zones of high concentrations of BBPs near arteries at specific times that are potential locations for cerebral vasospasm (CV) and delayed cerebral ischemia (DCI), and
issuing a warning of a CV occurrence and DCI location and time in specific areas of the patient's cerebral vasculature where blood flow is maintained from identified vasospasmic artery.
14 . The method of claim 13 further including calculating BBPs concentration within the SAS to identify ischemic zone locations.
15 . The method of claim 14 further including using the distribution of BBPs concentration to provide a 3D map of CV and DCI probability.
16 . The method of claim 15 further providing a 3D map of CV and DCI probability generated using the following equation:
P
risk
(
r
a
,
t
)
=
F
(
c
i
(
r
-
r
a
,
t
)
,
D
a
(
r
a
)
,
C
a
(
r
a
)
)
where
t represents time;
r represents a spatial position vector;
r a represents a spatial position vector pointing to a location on a cerebral arterial wall;
P risk (r a , t) represents a time-dependent three-dimensional map of the probability of cerebral vasospasm and delayed cerebral ischemia;
c i (r-r a , t) represents a time-dependent concentration of a blood breakdown product species at a spatial location offset from the cerebral arterial wall position r a ;
D a (r a ) represents a diameter of the cerebral artery at position r a ;
C a (r a ) represents a calcification level of the cerebral artery at position r a ;
F represents a function defining an integrative relationship between the probability of delayed cerebral ischemia and cerebral vasospasm at position r a and the parameters c i , D a , and C a .