Dataset Information

Assessment of New Coronary Features on Quantitative Coronary Angiographic Images With Innovative Unsupervised Artificial Adaptive Systems: A Proof-of-Concept Study

ABSTRACT: Background and Purpose: The Active Connection Matrixes (ACMs) are unsupervised artificial adaptive systems able to extract from digital images features of interest (edges, tissue differentiation, etc.) unnoticeable with conventional systems. In this proof-of-concept study, we assessed the potentiality of ACMs to increase measurement precision of morphological structures (e.g., stenosis and lumen diameter) and to grasp morphological features (arterial walls) from quantitative coronary angiography (QCA), unnoticeable on the original images. Methods: Archive images of QCA and intravascular ultrasound (IVUS) of 10 patients (8 men, age 69.1 ± 9.7 years) who underwent both procedures for clinical reasons were retrospectively analyzed. Arterial features derived from “IVUS images,” “conventional QCA images,” and “ACM-reprocessed QCA images” were measured in 21 coronary segments. Portions of 1-mm length (263 for lumen and 526 for arterial walls) were head-to-head compared to assess quali-quantitative between-methods agreement. Results: When stenosis was calculated on “ACM-reprocessed QCA images,” the bias vs. IVUS (gold standard) did not improve, but the correlation coefficient of the QCA–IVUS relationship increased from 0.47 to 0.83. When IVUS-derived lumen diameters were compared with diameters obtained on ACM-reprocessed QCA images, the bias (−0.25 mm) was significantly smaller (p < 0.01) than that observed with original QCA images (0.58 mm). ACMs were also able to extract arterial wall features from QCA. The bias between the measures of arterial walls obtained with IVUS and ACMs, although significant (p < 0.01), was small [0.09 mm, 95% CI (0.03, 0.14)] and the correlation was fairly good (r = 0.63; p < 0.0001). Conclusions: This study provides proof of concept that ACMs increase the measurement precision of coronary lumen diameter and allow extracting from QCA images hidden features that mirror well the arterial walls derived by IVUS.

SUBMITTER: Amato M

PROVIDER: S-EPMC8551448 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Many biophysical problems involve molecular and nanoscale targets moving next to a curvilinear track, e.g., a cytosolic cargo transported by motor proteins moving along a microtubule. For this type of problem, fluorescence imaging is usually the primary tool of choice. There is, however, an ∼20-fold mismatch between target-localization precision and track-imaging resolution such that questions requiring high-fidelity definition of the target's track remain inaccessible. On the other hand, if the contextual image of the tracks can be refined to a level comparable to that of the target, many intuitive yet mechanistically important issues can begin to be addressed. This work demonstrates that it is possible to statistically infer, to subpixel precision, curvilinear features in a low signal/noise image. This is achieved by a framework that consists of three stages: the Hessian-based feature enhancement, the subimage feature sampling and registration, and the statistical learning of the underlying curvilinear structure using a new, to our knowledge, method developed here for inferring the principal curves. In each stage, the descriptive prior information that the features come from curvilinear elements is explicitly taken into account. It is fully automated without user supervision, which is distinctly different from approaches that require user seeding or well-defined training data sets. Computer simulations of realistic images are used to investigate the performance of the framework and its implementation. The characterization results suggest that curvilinear features are refined to the same order of precision as that of the target and that the bootstrap confidence intervals from the analysis allow an estimate for the statistical bounds of the simulated "true" curve. Also shown are analyses of experimental images from three different microscopy modalities: two-photon laser-scanning microscopy, epifluorescence microscopy, and total internal reflection fluorescence microscopy. The practical application of this prior-apprised unsupervised learning framework as well as its potential outlook are discussed.

Project description:PurposeAutomation and computer assistance can support quality assurance tasks in radiotherapy. Retrospective image review requires significant human resources, and automation of image review remains a noteworthy missing element in previous work. Here, we present initial findings from a proof-of-concept clinical implementation of an AI-assisted review of CBCT registrations used for patient setup.MethodsAn automated pipeline was developed and executed nightly, utilizing python scripts to interact with the clinical database through DICOM networking protocol and automate data retrieval and analysis. A previously developed artificial intelligence (AI) algorithm scored CBCT setup registrations based on misalignment likelihood, using a scale from 0 (most unlikely) through 1 (most likely). Over a 45-day period, 1357 pre-treatment CBCT registrations from 197 patients were retrieved and analyzed by the pipeline. Daily summary reports of the previous day's registrations were produced. Initial action levels targeted 10% of cases to highlight for in-depth physics review. A validation subset of 100 cases was scored by three independent observers to characterize AI-model performance.ResultsFollowing an ROC analysis, a global threshold for model predictions of 0.87 was determined, with a sensitivity of 100% and specificity of 82%. Inspecting the observer scores for the stratified validation dataset showed a statistically significant correlation between observer scores and model predictions.ConclusionIn this work, we describe the implementation of an automated AI-analysis pipeline for daily quantitative analysis of CBCT-guided patient setup registrations. The AI-model was validated against independent expert observers, and appropriate action levels were determined to minimize false positives without sacrificing sensitivity. Case studies demonstrate the potential benefits of such a pipeline to bolster quality and safety programs in radiotherapy. To the authors' knowledge, there are no previous works performing AI-assisted assessment of pre-treatment CBCT-based patient alignment.

Project description:BackgroundPainful diabetic peripheral neuropathy (PDPN) affects up to 26% of patients with diabetes mellitus, with major impacts on their general health and well-being. Most available drugs fail to deliver acceptable pain reduction in the majority of patients and are often poorly tolerated. NRD.E1 is a novel product that has shown anti-nociceptive preclinical effects and good tolerability in healthy volunteer studies.MethodsThis phase 2a, randomized, dose-finding, Proof of Concept study enrolled patients with PDPN of ≥3 months duration. After at least one treatment-free week (WO week), 88 patients entered a 1-week single-blind (SB)-placebo run-in period, followed by 3 weeks' double-blind (DB) treatment, during which they received NRD.E1 at 10, 40 or 150 mg/day or placebo.ResultsThe primary endpoint (change from SB-placebo run-in week to week 3 in weekly mean of daily average numerical rating scale [NRS] pain intensity) showed clinically relevant placebo-corrected treatment effect pain reductions at 40 mg and 150 mg/day of 0.82 (95% CI: 0.07, 1.58, p = 0.034) and 0.66 (95% CI: -0.03, 1.35; p = 0.061) NRS points, respectively, though did not meet the pre-specified value of p = 0.016 required due to multiplicity. An additional post hoc endpoint looking at the change from WO baseline to week 3 in weekly mean of daily average NRS showed the placebo-corrected treatment effect was 1.46 (95% CI: 0.26, 2.66), and 1.20 (95% CI: 0.10, 2.29) NRS points, respectively. Secondary and post hoc analyses of NRS pain data (including 30 & 50% responder rate and NNT), sleep interference, Short-form McGill pain questionnaire (especially pain intensity assessed on Visual Analogue Scale), Patient's and Clinician's Global Impression of Change showed effects consistent with the primary findings. NRD.E1 was well tolerated, with only headache reported in more than two patients and more frequently on NRD.E1 than placebo.ConclusionsThe data suggest that NRD.E1 potentially represents a novel non-opioid therapeutic option for patients with PDPN, with at least similar efficacy and better tolerability than available therapies, justifying its further evaluation in larger-scale confirmatory studies.SignificanceNRD.E1 is a novel non-opioid therapeutic which is being developed for the treatment of PDPN. In this randomized, controlled, dose-finding, Proof of Concept study, NRD.E1 induced a clinically relevant pain reduction and it was well tolerated. Available data suggest that NRD.E1 has at least similar efficacy and better tolerability than the currently available therapies, potentially offering a promising new therapeutic option to patients with PDPN and possibly other neuropathic pain indications.

Dataset Information

Assessment of New Coronary Features on Quantitative Coronary Angiographic Images With Innovative Unsupervised Artificial Adaptive Systems: A Proof-of-Concept Study

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets