FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387249433> ?p ?o ?g. }

• W4387249433 endingPage "A4223" @default.
• W4387249433 startingPage "A4223" @default.
• W4387249433 abstract "SESSION TITLE: Lung Cancer Posters 1 SESSION TYPE: Original Investigation Posters PRESENTED ON: 10/10/2023 12:00 pm - 12:45 pm PURPOSE: Precise evaluation of nodule dimensions plays a critical role in clinical management. This study aims to evaluate the accuracy of a deep learning model and image acquisition process by estimating their respective errors and associated factors. METHODS: The data utilized in the study was acquired from Phantom FDA, which involved obtaining CT scans using anthropomorphic thoracic phantom containing a vasculature insert on which synthetic nodules were inserted or attached. This dataset contained nodules of varying characteristics, such as size, density, shape, and location, along with acquisition protocols such as Hounsfield units (HU) and slice thickness. The dataset comprised of four distinct configurations, encompassing a total of 670 nodules. The nodules come in five distinct sizes, with nominal diameters of 8, 10, 12, 15, and 20 mm, and various shapes such as lobulated, elliptical, spiculated, spherical, and irregular. The volume of these nodules ranges from 64 mm3 to a maximum of 4441 mm3, with an average volume of 1972 mm3 across all nodules. The study aimed to calculate and examine the normalized volume error across different nodules characteristics and acquisition parameters. RESULTS: The overall normalized error was 20.71% (95% confidence interval: 19.02-22.41). 51 (7.6%) of 670 had an error greater than 50%. 49 of these 51 nodules were from two layouts. Irregular nodules had a higher occurrence of errors compared to other nodule shapes, such as spiculated and lobulated nodules. Elliptical nodule shape has best prediction with mean normalized error of 7.41%. Slice thickness did not appear to be a factor. Surprisingly, smaller nodule had better predictions than larger ones. For nodules with diameter 8mm (smallest nodule size), mean normalized error is 14.67% and nodules with diameter 20mm (largest nodule size), the mean normalized error is 22.08%. All other available acquisition parameters did not have a significant effect on errors. CONCLUSIONS: The DL (Deep Learning) model had an error which is comparable to radiologists’ assessing volume in real CTs. Scans taken in different settings reported similar errors. A larger study with real CTs would be required to confirm these findings and to find its clinical relevance. Normalized error was chosen so that this error can be interpreted clinically for cases with follow-up scans. CLINICAL IMPLICATIONS: DL algorithm can aid the clinicians (physicians other than radiologists) for better and timely diagnosis in low resource settings. This is particularly more relevant in the areas where there is a lack of experienced radiologists. Furthermore, AI can also be utilized to highlight the limitations and errors inherent in human interpretation and image acquisition especially in busy clinical settings. DISCLOSURES: No relevant relationships by Rohitashva Agrawal No disclosure on file for Vikash Challa Employee relationship with Qure.ai Please note: Current employee by Souvik Mandal, value=Salary Employee relationship with Qure.ai Please note: Current employee by Ankit Modi, value=Salary Employee relationship with Qure AI Please note: Current Employee by Vanapalli Prakash, value=Salary Employee relationship with Qure.ai Please note: Current employee by Preetham Putha, value=Salary Employee relationship with Qure.ai Please note: Current Employee by Saigopal Sathyamurthy, value=Salary No relevant relationships by Moksh Shukla No disclosure on file for Prashant Warier" @default.
• W4387249433 created "2023-10-03" @default.
• W4387249433 creator A5004810217 @default.
• W4387249433 creator A5007829534 @default.
• W4387249433 creator A5013138148 @default.
• W4387249433 creator A5055547835 @default.
• W4387249433 creator A5064304476 @default.
• W4387249433 creator A5069881941 @default.
• W4387249433 creator A5072323444 @default.
• W4387249433 creator A5003528059 @default.
• W4387249433 creator A5052143253 @default.
• W4387249433 date "2023-10-01" @default.
• W4387249433 modified "2023-10-03" @default.
• W4387249433 title "ERROR ANALYSIS OF A DEEP-LEARNING MODEL FOR NODULE VOLUME ESTIMATION: IMPLICATIONS FOR CLINICAL MANAGEMENT" @default.
• W4387249433 doi "https://doi.org/10.1016/j.chest.2023.07.2748" @default.
• W4387249433 hasPublicationYear "2023" @default.
• W4387249433 type Work @default.
• W4387249433 citedByCount "0" @default.
• W4387249433 crossrefType "journal-article" @default.
• W4387249433 hasAuthorship W4387249433A5003528059 @default.
• W4387249433 hasAuthorship W4387249433A5004810217 @default.
• W4387249433 hasAuthorship W4387249433A5007829534 @default.
• W4387249433 hasAuthorship W4387249433A5013138148 @default.
• W4387249433 hasAuthorship W4387249433A5052143253 @default.
• W4387249433 hasAuthorship W4387249433A5055547835 @default.
• W4387249433 hasAuthorship W4387249433A5064304476 @default.
• W4387249433 hasAuthorship W4387249433A5069881941 @default.
• W4387249433 hasAuthorship W4387249433A5072323444 @default.
• W4387249433 hasBestOaLocation W43872494331 @default.
• W4387249433 hasConcept C104293457 @default.
• W4387249433 hasConcept C121332964 @default.
• W4387249433 hasConcept C126322002 @default.
• W4387249433 hasConcept C126838900 @default.
• W4387249433 hasConcept C151730666 @default.
• W4387249433 hasConcept C187954543 @default.
• W4387249433 hasConcept C20556612 @default.
• W4387249433 hasConcept C2776731575 @default.
• W4387249433 hasConcept C2989005 @default.
• W4387249433 hasConcept C44249647 @default.
• W4387249433 hasConcept C544519230 @default.
• W4387249433 hasConcept C62520636 @default.
• W4387249433 hasConcept C71924100 @default.
• W4387249433 hasConcept C86803240 @default.
• W4387249433 hasConceptScore W4387249433C104293457 @default.
• W4387249433 hasConceptScore W4387249433C121332964 @default.
• W4387249433 hasConceptScore W4387249433C126322002 @default.
• W4387249433 hasConceptScore W4387249433C126838900 @default.
• W4387249433 hasConceptScore W4387249433C151730666 @default.
• W4387249433 hasConceptScore W4387249433C187954543 @default.
• W4387249433 hasConceptScore W4387249433C20556612 @default.
• W4387249433 hasConceptScore W4387249433C2776731575 @default.
• W4387249433 hasConceptScore W4387249433C2989005 @default.
• W4387249433 hasConceptScore W4387249433C44249647 @default.
• W4387249433 hasConceptScore W4387249433C544519230 @default.
• W4387249433 hasConceptScore W4387249433C62520636 @default.
• W4387249433 hasConceptScore W4387249433C71924100 @default.
• W4387249433 hasConceptScore W4387249433C86803240 @default.
• W4387249433 hasIssue "4" @default.
• W4387249433 hasLocation W43872494331 @default.
• W4387249433 hasOpenAccess W4387249433 @default.
• W4387249433 hasPrimaryLocation W43872494331 @default.
• W4387249433 hasRelatedWork W1965389463 @default.
• W4387249433 hasRelatedWork W2059417794 @default.
• W4387249433 hasRelatedWork W2066429881 @default.
• W4387249433 hasRelatedWork W2075565253 @default.
• W4387249433 hasRelatedWork W2152852918 @default.
• W4387249433 hasRelatedWork W2317883499 @default.
• W4387249433 hasRelatedWork W2407861221 @default.
• W4387249433 hasRelatedWork W3028984494 @default.
• W4387249433 hasRelatedWork W3036272248 @default.
• W4387249433 hasRelatedWork W4309305996 @default.
• W4387249433 hasVolume "164" @default.
• W4387249433 isParatext "false" @default.
• W4387249433 isRetracted "false" @default.
• W4387249433 workType "article" @default.