3D-Arterial analysis software and CEUS in the assessment of severity and vulnerability of carotid atherosclerotic plaque: a comparison with CTA and histopathology

Abstract

Purpose Our purpose is to assess Multiparametric Ultrasound (MPUS) efficacy for evaluation of carotid plaque vulnerability and carotid stenosis degree in comparison with Computed Tomography angiography (CTA) and histology. Material and methods 3D-Arterial Analysis is a 3D ultrasound software that automatically provides the degree of carotid stenosis and a colorimetric map of carotid plaque vulnerability. We enrolled 106 patients who were candidates for carotid endarterectomy. Prior to undergoing surgery, all carotid artery plaques were evaluated with Color-Doppler-US (CDUS), Contrast-Enhanced Ultrasound (CEUS), and 3D Arterial analysis (3DAA) US along with Computerized Tomographic Angiography (CTA) to assess the carotid artery stenosis degree. Post-surgery, the carotid specimens were fixed with 10% neutral buffered formalin solution, embedded in paraffin and used for light microscopic examination to assess plaque vulnerability morphological features. Results The results of the CTA examinations revealed 91 patients with severe carotid stenoses with a resultant diagnostic accuracy of 82.3% for CDUS, 94.5% for CEUS, 98.4% for 3DAA, respectively. The histopathological examination showed 71 vulnerable plaques with diagnostic accuracy values of 85.8% for CDUS, 93.4% for CEUS, 90.3% for 3DAA, 92% for CTA, respectively. Conclusions The combination of CEUS and 3D Arterial Analysis may provide a powerful new clinical tool to identify and stratify “at-risk” patients with atherosclerotic carotid artery disease, identifying vulnerable plaques. These applications may also help in the postoperative assessment of treatment options to manage cardiovascular risks.

Radiation Recall Pneumonitis: The Open Challenge in Differential Diagnosis of Pneumonia Induced by Oncological Treatments

“…Some promising tools in the diagnosis and follow-up of cancer patients who develop adverse reactions to treatments are artificial intelligence (AI) and radiomics [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. Computers are able to accumulate and evaluate higher volumes of data compared to the human brain, so AI can resolve unsolved complexities in cancer patient management [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. Machine learning (ML) is a sub-area of AI which uses mathematical algorithms and can learn specific tasks [ 162 , 163 , 164 ,…”

“…Computers are able to accumulate and evaluate higher volumes of data compared to the human brain, so AI can resolve unsolved complexities in cancer patient management [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. Machine learning (ML) is a sub-area of AI which uses mathematical algorithms and can learn specific tasks [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. These models are supervised and unsupervised, depending on the knowledge of the desired outcome of interest [ 162 , 163 , 164 , 165 ,…”

“…Machine learning (ML) is a sub-area of AI which uses mathematical algorithms and can learn specific tasks [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. These models are supervised and unsupervised, depending on the knowledge of the desired outcome of interest [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. The ML models analyze the input producing the necessary adjustments to obtain the desired output, or the models assess uncurated data and classify thanks to defined features within the dataset that can be grouped and analyzed further to reach a specific outcome, respectively, [ 162 , 163 , …”

“…These models are supervised and unsupervised, depending on the knowledge of the desired outcome of interest [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. The ML models analyze the input producing the necessary adjustments to obtain the desired output, or the models assess uncurated data and classify thanks to defined features within the dataset that can be grouped and analyzed further to reach a specific outcome, respectively, [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. These models have made possible the development of a new approach named radiomics [ 162 , 163 , 164 ...…”

“…The ML models analyze the input producing the necessary adjustments to obtain the desired output, or the models assess uncurated data and classify thanks to defined features within the dataset that can be grouped and analyzed further to reach a specific outcome, respectively, [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ]. These models have made possible the development of a new approach named radiomics [ 162 , 163 , 164 , 165 , 166 , 167 , 168 , 169 , 170 , 171 , 172 , 173 , 174 , 175 , 176 , 177 , 178 , 179 , 180 , 181 , 182 , 183 , 184 , 185 ].…”