Instance segmentation of upper aerodigestive tract cancer : site-specific outcomes
Copyright © 2023 Società Italiana di Otorinolaringoiatria e Chirurgia Cervico-Facciale, Rome, Italy..
Objective: To achieve instance segmentation of upper aerodigestive tract (UADT) neoplasms using a deep learning (DL) algorithm, and to identify differences in its diagnostic performance in three different sites: larynx/hypopharynx, oral cavity and oropharynx.
Methods: A total of 1034 endoscopic images from 323 patients were examined under narrow band imaging (NBI). The Mask R-CNN algorithm was used for the analysis. The dataset split was: 935 training, 48 validation and 51 testing images. Dice Similarity Coefficient (Dsc) was the main outcome measure.
Results: Instance segmentation was effective in 76.5% of images. The mean Dsc was 0.90 ± 0.05. The algorithm correctly predicted 77.8%, 86.7% and 55.5% of lesions in the larynx/hypopharynx, oral cavity, and oropharynx, respectively. The mean Dsc was 0.90 ± 0.05 for the larynx/hypopharynx, 0.60 ± 0.26 for the oral cavity, and 0.81 ± 0.30 for the oropharynx. The analysis showed inferior diagnostic results in the oral cavity compared with the larynx/hypopharynx (p < 0.001).
Conclusions: The study confirms the feasibility of instance segmentation of UADT using DL algorithms and shows inferior diagnostic results in the oral cavity compared with other anatomic areas.
| Medienart: |
E-Artikel |
|---|
| Erscheinungsjahr: |
2023 |
|---|---|
| Erschienen: |
2023 |
| Enthalten in: |
Zur Gesamtaufnahme - volume:43 |
|---|---|
| Enthalten in: |
Acta otorhinolaryngologica Italica : organo ufficiale della Societa italiana di otorinolaringologia e chirurgia cervico-facciale - 43(2023), 4 vom: 24. Aug., Seite 283-290 |
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Paderno, Alberto [VerfasserIn] |
|---|
| Links: |
|---|
| Themen: |
Artificial intelligence |
|---|
| Anmerkungen: |
Date Completed 26.07.2023 Date Revised 02.08.2023 published: Print Citation Status MEDLINE |
|---|
| doi: |
10.14639/0392-100X-N2336 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
NLM359892027 |
|---|
| LEADER | 01000naa a22002652 4500 | ||
|---|---|---|---|
| 001 | NLM359892027 | ||
| 003 | DE-627 | ||
| 005 | 20231226081838.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 231226s2023 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.14639/0392-100X-N2336 |2 doi | |
| 028 | 5 | 2 | |a pubmed24n1199.xml |
| 035 | |a (DE-627)NLM359892027 | ||
| 035 | |a (NLM)37488992 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Paderno, Alberto |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Instance segmentation of upper aerodigestive tract cancer |b site-specific outcomes |
| 264 | 1 | |c 2023 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ƒaComputermedien |b c |2 rdamedia | ||
| 338 | |a ƒa Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Date Completed 26.07.2023 | ||
| 500 | |a Date Revised 02.08.2023 | ||
| 500 | |a published: Print | ||
| 500 | |a Citation Status MEDLINE | ||
| 520 | |a Copyright © 2023 Società Italiana di Otorinolaringoiatria e Chirurgia Cervico-Facciale, Rome, Italy. | ||
| 520 | |a Objective: To achieve instance segmentation of upper aerodigestive tract (UADT) neoplasms using a deep learning (DL) algorithm, and to identify differences in its diagnostic performance in three different sites: larynx/hypopharynx, oral cavity and oropharynx | ||
| 520 | |a Methods: A total of 1034 endoscopic images from 323 patients were examined under narrow band imaging (NBI). The Mask R-CNN algorithm was used for the analysis. The dataset split was: 935 training, 48 validation and 51 testing images. Dice Similarity Coefficient (Dsc) was the main outcome measure | ||
| 520 | |a Results: Instance segmentation was effective in 76.5% of images. The mean Dsc was 0.90 ± 0.05. The algorithm correctly predicted 77.8%, 86.7% and 55.5% of lesions in the larynx/hypopharynx, oral cavity, and oropharynx, respectively. The mean Dsc was 0.90 ± 0.05 for the larynx/hypopharynx, 0.60 ± 0.26 for the oral cavity, and 0.81 ± 0.30 for the oropharynx. The analysis showed inferior diagnostic results in the oral cavity compared with the larynx/hypopharynx (p < 0.001) | ||
| 520 | |a Conclusions: The study confirms the feasibility of instance segmentation of UADT using DL algorithms and shows inferior diagnostic results in the oral cavity compared with other anatomic areas | ||
| 650 | 4 | |a Journal Article | |
| 650 | 4 | |a artificial intelligence | |
| 650 | 4 | |a deep learning | |
| 650 | 4 | |a instance segmentation | |
| 650 | 4 | |a videomics | |
| 700 | 1 | |a Villani, Francesca Pia |e verfasserin |4 aut | |
| 700 | 1 | |a Fior, Milena |e verfasserin |4 aut | |
| 700 | 1 | |a Berretti, Giulia |e verfasserin |4 aut | |
| 700 | 1 | |a Gennarini, Francesca |e verfasserin |4 aut | |
| 700 | 1 | |a Zigliani, Gabriele |e verfasserin |4 aut | |
| 700 | 1 | |a Ulaj, Emanuela |e verfasserin |4 aut | |
| 700 | 1 | |a Montenegro, Claudia |e verfasserin |4 aut | |
| 700 | 1 | |a Sordi, Alessandra |e verfasserin |4 aut | |
| 700 | 1 | |a Sampieri, Claudio |e verfasserin |4 aut | |
| 700 | 1 | |a Peretti, Giorgio |e verfasserin |4 aut | |
| 700 | 1 | |a Moccia, Sara |e verfasserin |4 aut | |
| 700 | 1 | |a Piazza, Cesare |e verfasserin |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Acta otorhinolaryngologica Italica : organo ufficiale della Societa italiana di otorinolaringologia e chirurgia cervico-facciale |d 1985 |g 43(2023), 4 vom: 24. Aug., Seite 283-290 |w (DE-627)NLM01279046X |x 1827-675X |7 nnns |
| 773 | 1 | 8 | |g volume:43 |g year:2023 |g number:4 |g day:24 |g month:08 |g pages:283-290 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.14639/0392-100X-N2336 |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a GBV_NLM | ||
| 951 | |a AR | ||
| 952 | |d 43 |j 2023 |e 4 |b 24 |c 08 |h 283-290 | ||