Method and system using magnetic resonance imaging for tissue classification and bulk-density assignment

Abstract

An apparatus includes a magnetic resonance imaging system, a processor for controlling the apparatus, and a memory containing machine executable instructions and a pulse sequence. The machine executable instructions and pulse sequence cause the processor to control the apparatus to: acquire magnetic resonance data from an imaging volume, wherein the magnetic resonance data includes gradient echo data; segment the magnetic resonance data into a plurality of segments, the segments including a fat segment, a water segment, a cortical bone segment, and an air segment; and create a bulk density map of the imaging volume from the segments.

What is claimed is: 
     
         1 . An apparatus, comprising:
 a magnetic resonance imaging system which acquires magnetic resonance data from an imaging volume;   a processor for controlling the apparatus; and   a memory containing machine executable instructions and a pulse sequence, wherein the magnetic resonance data is acquired using the pulse sequence includes gradient echo data, wherein execution of the instructions causes the processor to:
 acquire the magnetic resonance data using the magnetic resonance imaging system and the pulse sequence; and 
 segment the magnetic resonance data into a plurality of segments, including a fat segment, a water segment, a cortical bone segment, and an air segment; and 
 create a bulk density map of the imaging volume from the segments. 
   
     
     
         2 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to create the bulk density map from the segments by:
 determining for each of a plurality of voxels of the imaging volume, which element among fat, water, cortical bone and air segment is primarily represented in the voxel; and   assigning a corresponding bulk density value to each voxel, where the assigned bulk density value depends on which element among fat, water, cortical bone and air is primarily represented in the voxel.   
     
     
         3 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to create the bulk density map from the segments by:
 assigning corresponding bulk density values to fat, water, cortical bone and air;   determining, for each of a plurality of voxels of the imaging volume, a plurality of fractions which pertain to the voxel, including a fat fraction, a water fraction, a cortical bone fraction, and an air fraction; and   for each of the plurality of voxels, weighting each of the plurality of fractions by the corresponding bulk density value.   
     
     
         4 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to generate one or more digitally reconstructed radiographs (DRRs) from the magnetic resonance data. 
     
     
         5 . The apparatus of  claim 4 , wherein execution of the instructions further causes the processor to transfer the one or more DRRs to a radiation treatment planning system. 
     
     
         6 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to generate an artificial computed tomography image based on fractions of fat, water, air, and cortical bone in each voxel of the imaging volume. 
     
     
         7 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to reconstruct an in-phase image, a fat-saturated image, a water-saturated image, and an ultra-short echo time image from the magnetic resonance data, and to produce the cortical bone segment by subtracting a scaled multiple of the in-phase image from the ultra-short echo time image. 
     
     
         8 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to reconstruct an in-phase image, a fat-saturated image, and a water-saturated image from the magnetic resonance data, and to produce the cortical bone segment by automatically thresholding a noise level in the in-phase image and subsequently removing background noise. 
     
     
         9 . The apparatus of  claim 8 , wherein execution of the instructions further causes the processor to produce the cortical bone segment by registering the in-phase image with a bone probability atlas. 
     
     
         10 . The apparatus of  claim 1 , wherein execution of the instructions further causes the processor to transfer the bulk density map to a radiation treatment planning system. 
     
     
         11 . A method, comprising:
 acquiring magnetic resonance data from an imaging volume via a magnetic resonance imaging system and a pulse sequence, wherein the magnetic resonance data includes gradient echo data;   segmenting the magnetic resonance data into a plurality of segments, the segments including a fat segment, a water segment, a cortical bone segment, and an air segment; and   creating a bulk density map of the imaging volume from the segments.   
     
     
         12 . The method of  claim 11 , wherein creating the bulk density map from the segments comprises:
 determining, for each of a plurality of voxels of the imaging volume, which element among fat, water, air and cortical bone is primarily represented in the voxel; and   assigning a corresponding bulk density value to each voxel, where the assigned bulk density value depends on which element among fat, water, cortical bone, and air is primarily represented in the voxel.   
     
     
         13 . The method of  claim 11 , wherein creating the bulk density map from the segments comprises:
 assigning corresponding bulk density values to fat, water, cortical bone and air;   determining, for each of a plurality of voxels of the imaging volume, a plurality of fractions which pertain to the voxel, including an a fat fraction, a water fraction, a cortical bone fraction and an air fraction; and   for each of the plurality of voxels, weighting each of the plurality of fractions by the corresponding bulk density value.   
     
     
         14 . The method of  claim 11 , further comprising generating one or more digitally reconstructed radiographs (DRRs) from the magnetic resonance data. 
     
     
         15 . The method of  claim 14 , further comprising transferring the one or more DRRs to a radiation treatment planning system. 
     
     
         16 . The method of  claim 11 , further comprising generating an artificial computed tomography image based on fractions of fat, water, cortical bone, and air in each voxel of the imaging volume. 
     
     
         17 . The method of  claim 11 , further comprising:
 reconstructing an in-phase image, a fat-saturated image, a water-saturated image, and an ultra-short echo time image from the magnetic resonance data; and   producing the cortical bone segment by subtracting a scaled multiple of the in-phase image from the ultra-short echo time image.   
     
     
         18 . The method of  claim 11 , further comprising:
 reconstructing an in-phase image, a fat-saturated image, and a water-saturated image from the magnetic resonance data; and   producing the cortical bone segment by automatically thresholding a noise level in the in-phase image and subsequently removing background noise.   
     
     
         19 . The method of  claim 18 , wherein producing the cortical bone segment further comprises registering the in-phase image with a bone probability atlas. 
     
     
         20 . A non-transitory computer-readable storage medium having stored therein a pulse sequence and machine readable instructions configured to be executed by a processor to control an apparatus including a magnetic resonance imaging system, the machine readable instructions being configured in conjunction with the pulse sequence to cause the apparatus to execute a process comprising:
 acquiring magnetic resonance data from an imaging volume using the magnetic resonance imaging system and the pulse sequence, wherein the magnetic resonance data includes gradient echo data;   segmenting the magnetic resonance data into a plurality of segments, the segments including a fat segment, a water segment, a cortical bone segment, and an air segment; and   creating a bulk density map of the imaging volume from the segments.