Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism
Detector-based spectral CT offers the possibility of obtaining spectral information from which discrete acquisitions at different energy levels can be derived, yielding so-called virtual monoenergetic images (VMI). In this study, we aimed to develop a jointly optimized deep-learning framework based...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-05-01
|
Series: | Diagnostics |
Subjects: | |
Online Access: | https://www.mdpi.com/2075-4418/12/5/1224 |
_version_ | 1797500473621610496 |
---|---|
author | Matthias A. Fink Constantin Seibold Hans-Ulrich Kauczor Rainer Stiefelhagen Jens Kleesiek |
author_facet | Matthias A. Fink Constantin Seibold Hans-Ulrich Kauczor Rainer Stiefelhagen Jens Kleesiek |
author_sort | Matthias A. Fink |
collection | DOAJ |
description | Detector-based spectral CT offers the possibility of obtaining spectral information from which discrete acquisitions at different energy levels can be derived, yielding so-called virtual monoenergetic images (VMI). In this study, we aimed to develop a jointly optimized deep-learning framework based on dual-energy CT pulmonary angiography (DE-CTPA) data to generate synthetic monoenergetic images (SMI) for improving automatic pulmonary embolism (PE) detection in single-energy CTPA scans. For this purpose, we used two datasets: our institutional DE-CTPA dataset <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>1</mn></msub></semantics></math></inline-formula>, comprising polyenergetic arterial series and the corresponding VMI at low-energy levels (40 keV) with 7892 image pairs, and a 10% subset of the 2020 RSNA Pulmonary Embolism CT Dataset <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>2</mn></msub></semantics></math></inline-formula>, which consisted of 161,253 polyenergetic images with dichotomous slice-wise annotations (PE/no PE). We trained a fully convolutional encoder-decoder on <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>1</mn></msub></semantics></math></inline-formula> to generate SMI from single-energy CTPA scans of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>2</mn></msub></semantics></math></inline-formula>, which were then fed into a ResNet50 network for training of the downstream PE classification task. The quantitative results on the reconstruction ability of our framework revealed high-quality visual SMI predictions with reconstruction results of 0.984 ± 0.002 (structural similarity) and 41.706 ± 0.547 dB (peak signal-to-noise ratio). PE classification resulted in an AUC of 0.84 for our model, which achieved improved performance compared to other naïve approaches with AUCs up to 0.81. Our study stresses the role of using joint optimization strategies for deep-learning algorithms to improve automatic PE detection. The proposed pipeline may prove to be beneficial for computer-aided detection systems and could help rescue CTPA studies with suboptimal opacification of the pulmonary arteries from single-energy CT scanners. |
first_indexed | 2024-03-10T03:02:24Z |
format | Article |
id | doaj.art-eb6d6b4e6b67401e8ed29f1464f611b2 |
institution | Directory Open Access Journal |
issn | 2075-4418 |
language | English |
last_indexed | 2024-03-10T03:02:24Z |
publishDate | 2022-05-01 |
publisher | MDPI AG |
record_format | Article |
series | Diagnostics |
spelling | doaj.art-eb6d6b4e6b67401e8ed29f1464f611b22023-11-23T10:41:08ZengMDPI AGDiagnostics2075-44182022-05-01125122410.3390/diagnostics12051224Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary EmbolismMatthias A. Fink0Constantin Seibold1Hans-Ulrich Kauczor2Rainer Stiefelhagen3Jens Kleesiek4Clinic for Diagnostic and Interventional Radiology, Heidelberg University Hospital, Im Neuenheimer Feld 420, 69120 Heidelberg, GermanyDepartment of Computer Science, Karlsruhe Institute of Technology (KIT), 76131 Karlsruhe, GermanyClinic for Diagnostic and Interventional Radiology, Heidelberg University Hospital, Im Neuenheimer Feld 420, 69120 Heidelberg, GermanyDepartment of Computer Science, Karlsruhe Institute of Technology (KIT), 76131 Karlsruhe, GermanyInstitute for AI in Medicine (IKIM), University Hospital Essen, 45131 Essen, GermanyDetector-based spectral CT offers the possibility of obtaining spectral information from which discrete acquisitions at different energy levels can be derived, yielding so-called virtual monoenergetic images (VMI). In this study, we aimed to develop a jointly optimized deep-learning framework based on dual-energy CT pulmonary angiography (DE-CTPA) data to generate synthetic monoenergetic images (SMI) for improving automatic pulmonary embolism (PE) detection in single-energy CTPA scans. For this purpose, we used two datasets: our institutional DE-CTPA dataset <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>1</mn></msub></semantics></math></inline-formula>, comprising polyenergetic arterial series and the corresponding VMI at low-energy levels (40 keV) with 7892 image pairs, and a 10% subset of the 2020 RSNA Pulmonary Embolism CT Dataset <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>2</mn></msub></semantics></math></inline-formula>, which consisted of 161,253 polyenergetic images with dichotomous slice-wise annotations (PE/no PE). We trained a fully convolutional encoder-decoder on <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>1</mn></msub></semantics></math></inline-formula> to generate SMI from single-energy CTPA scans of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>D</mi><mn>2</mn></msub></semantics></math></inline-formula>, which were then fed into a ResNet50 network for training of the downstream PE classification task. The quantitative results on the reconstruction ability of our framework revealed high-quality visual SMI predictions with reconstruction results of 0.984 ± 0.002 (structural similarity) and 41.706 ± 0.547 dB (peak signal-to-noise ratio). PE classification resulted in an AUC of 0.84 for our model, which achieved improved performance compared to other naïve approaches with AUCs up to 0.81. Our study stresses the role of using joint optimization strategies for deep-learning algorithms to improve automatic PE detection. The proposed pipeline may prove to be beneficial for computer-aided detection systems and could help rescue CTPA studies with suboptimal opacification of the pulmonary arteries from single-energy CT scanners.https://www.mdpi.com/2075-4418/12/5/1224artificial intelligencedeep learningimage-to-image translationdual-energy computed tomographypulmonary embolismemergency radiology |
spellingShingle | Matthias A. Fink Constantin Seibold Hans-Ulrich Kauczor Rainer Stiefelhagen Jens Kleesiek Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism Diagnostics artificial intelligence deep learning image-to-image translation dual-energy computed tomography pulmonary embolism emergency radiology |
title | Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism |
title_full | Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism |
title_fullStr | Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism |
title_full_unstemmed | Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism |
title_short | Jointly Optimized Deep Neural Networks to Synthesize Monoenergetic Images from Single-Energy CT Angiography for Improving Classification of Pulmonary Embolism |
title_sort | jointly optimized deep neural networks to synthesize monoenergetic images from single energy ct angiography for improving classification of pulmonary embolism |
topic | artificial intelligence deep learning image-to-image translation dual-energy computed tomography pulmonary embolism emergency radiology |
url | https://www.mdpi.com/2075-4418/12/5/1224 |
work_keys_str_mv | AT matthiasafink jointlyoptimizeddeepneuralnetworkstosynthesizemonoenergeticimagesfromsingleenergyctangiographyforimprovingclassificationofpulmonaryembolism AT constantinseibold jointlyoptimizeddeepneuralnetworkstosynthesizemonoenergeticimagesfromsingleenergyctangiographyforimprovingclassificationofpulmonaryembolism AT hansulrichkauczor jointlyoptimizeddeepneuralnetworkstosynthesizemonoenergeticimagesfromsingleenergyctangiographyforimprovingclassificationofpulmonaryembolism AT rainerstiefelhagen jointlyoptimizeddeepneuralnetworkstosynthesizemonoenergeticimagesfromsingleenergyctangiographyforimprovingclassificationofpulmonaryembolism AT jenskleesiek jointlyoptimizeddeepneuralnetworkstosynthesizemonoenergeticimagesfromsingleenergyctangiographyforimprovingclassificationofpulmonaryembolism |