PMID- 37063491
OWN - NLM
STAT- PubMed-not-MEDLINE
LR  - 20230419
IS  - 0277-786X (Print)
IS  - 1996-756X (Electronic)
IS  - 0277-786X (Linking)
VI  - 12463
DP  - 2023 Feb
TI  - Multi-energy CT material decomposition using Bayesian deep convolutional neural 
      network with explicit penalty of uncertainty and bias.
LID - 124633M [pii]
LID - 10.1117/12.2654317 [doi]
AB  - Convolutional neural network (CNN)-based material decomposition has the potential 
      to improve image quality (visual appearance) and quantitative accuracy of 
      material maps. Most methods use deterministic CNNs with mean-square-error loss to 
      provide point-estimates of mass densities. Point estimates can be over-confident 
      as the reliability of CNNs is frequently compromised by bias and two major 
      uncertainties - data and model uncertainties originating from noise in inputs and 
      train-test data dissimilarity, respectively. Also, mean-square-error lacks 
      explicit control of uncertainty and bias. To tackle these problems, a Bayesian 
      dual-task CNN (BDT-CNN) with explicit penalization of uncertainty and bias was 
      developed. It is a probabilistic CNN that concurrently conducts material 
      classification and quantification and allows for pixel-wise modeling of bias, 
      data uncertainty, and model uncertainty. CNN was trained with images of physical 
      and simulated tissue-mimicking inserts at varying mass densities. Hydroxyapatite 
      (nominal density 400mg/cc) and blood (nominal density 1095mg/cc) inserts were 
      placed in different-sized body phantoms (30 - 45cm) and used to evaluate 
      mean-absolute-bias (MAB) in predicted mass densities across different images at 
      routine- and half-routine-dose. Patient CT exams were collected to assess 
      generalizability of BDT-CNN in the presence of anatomical background. Noise 
      insertion was used to simulate patient exams at half- and quarter-routine-dose. 
      The deterministic dual-task CNN was used as baseline. In phantoms, BDT-CNN 
      improved consistency of insert delineation, especially edges, and reduced overall 
      bias (average MAB for hydroxyapatite: BDT-CNN 5.4mgHA/cc, baseline 11.0mgHA/cc 
      and blood: BDT-CNN 8.9mgBlood/cc, baseline 14.0mgBlood/cc). In patient images, 
      BDT-CNN improved detail preservation, lesion conspicuity, and structural 
      consistency across different dose levels.
FAU - Gong, Hao
AU  - Gong H
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
FAU - Leng, Shuai
AU  - Leng S
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
FAU - Baffour, Francis
AU  - Baffour F
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
FAU - Yu, Lifeng
AU  - Yu L
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
FAU - Fletcher, Joel G
AU  - Fletcher JG
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
FAU - McCollough, Cynthia H
AU  - McCollough CH
AD  - Department of Radiology, Mayo Clinic, Rochester, MN, USA, 55901.
LA  - eng
GR  - R01 EB028590/EB/NIBIB NIH HHS/United States
PT  - Journal Article
DEP - 20230407
PL  - United States
TA  - Proc SPIE Int Soc Opt Eng
JT  - Proceedings of SPIE--the International Society for Optical Engineering
JID - 101524122
PMC - PMC10099768
MID - NIHMS1880765
OTO - NOTNLM
OT  - Bayesian neural network
OT  - Multi-energy CT
OT  - bias
OT  - deep learning
OT  - material decomposition
OT  - uncertainty
EDAT- 2023/04/18 06:00
MHDA- 2023/04/18 06:01
PMCR- 2023/04/13
CRDT- 2023/04/17 03:30
PHST- 2023/04/18 06:01 [medline]
PHST- 2023/04/17 03:30 [entrez]
PHST- 2023/04/18 06:00 [pubmed]
PHST- 2023/04/13 00:00 [pmc-release]
AID - 124633M [pii]
AID - 10.1117/12.2654317 [doi]
PST - ppublish
SO  - Proc SPIE Int Soc Opt Eng. 2023 Feb;12463:124633M. doi: 10.1117/12.2654317. Epub 
      2023 Apr 7.