Fine-needle aspiration of follicular adenoma versus parathyroid adenoma: The utility of multispectral imaging in differentiating lesions with subtle cytomorphologic differences

Ibrahim Mansoor, Carola Zalles, Farhan Zahid, Kirk Gossage, Richard M Levenson, David L. Rimm

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

BACKGROUND. Multispectral image analysis is an emerging tool that utilizes both spatial and spectral image information to classify images that can be used for the differentiation between benign versus malignant cells. The aim of the current study was to analyze the ability of this tool in differentiating subtle cytologic differences that cannot be appreciated by the human eye. Herein, the authors used fine-needle aspirations (FNAs) of follicular adenoma (FA) and parathyroid adenoma (PA) as a test case. METHODS. The Nuance platform was used to collect image stacks that were subsequently analyzed with CRI-MLS software, a neural network-based artificial intelligence system that can classify images using automatically "learned" spatial-spectral features. CRI-MLS was trained on random, well-preserved FA cells and PA cells from the training set (n = 45 cells each). An algorithmic solution was developed and then validated on an independent series comprised of 1904 FA cells from 5 FA cases and 690 PA cells from 5 PA cases. RESULTS. The solution from the CRI-MLS classifier showed 1876 FA cells (98.5%) as true FA and 28 FA cells (1.5%) as false PA, whereas 663 PA cells (96%) were true PA and 27 PA cells (4%) were false FA. The summary result of this solution was a sensitivity of 98.5%, a specificity of 96.1%, and a positive predictive value of 98.6%. CONCLUSIONS. The best spatial-spectral imaging solution was able to correctly classify 2534 of 2594 cells (98%) and misclassifled only 55 of 2594 cells (2%). These data suggest that this technology may be valuable in a clinical setting to help differentiate and classify morphologically similar lesions.

Original languageEnglish (US)
Pages (from-to)22-26
Number of pages5
JournalCancer
Volume114
Issue number1
DOIs
StatePublished - Feb 25 2008
Externally publishedYes

Fingerprint

Parathyroid Neoplasms
Fine Needle Biopsy
Adenoma
Artificial Intelligence
Software

Keywords

  • Artificial intelligence
  • CRI-MLS software
  • Cytology
  • Cytopathology
  • Neural network
  • Spectral-spatial imaging

ASJC Scopus subject areas

  • Cancer Research
  • Oncology

Cite this

Fine-needle aspiration of follicular adenoma versus parathyroid adenoma : The utility of multispectral imaging in differentiating lesions with subtle cytomorphologic differences. / Mansoor, Ibrahim; Zalles, Carola; Zahid, Farhan; Gossage, Kirk; Levenson, Richard M; Rimm, David L.

In: Cancer, Vol. 114, No. 1, 25.02.2008, p. 22-26.

Research output: Contribution to journalArticle

@article{69ce5f17afec47da8e6e2a810c96ef25,
title = "Fine-needle aspiration of follicular adenoma versus parathyroid adenoma: The utility of multispectral imaging in differentiating lesions with subtle cytomorphologic differences",
abstract = "BACKGROUND. Multispectral image analysis is an emerging tool that utilizes both spatial and spectral image information to classify images that can be used for the differentiation between benign versus malignant cells. The aim of the current study was to analyze the ability of this tool in differentiating subtle cytologic differences that cannot be appreciated by the human eye. Herein, the authors used fine-needle aspirations (FNAs) of follicular adenoma (FA) and parathyroid adenoma (PA) as a test case. METHODS. The Nuance platform was used to collect image stacks that were subsequently analyzed with CRI-MLS software, a neural network-based artificial intelligence system that can classify images using automatically {"}learned{"} spatial-spectral features. CRI-MLS was trained on random, well-preserved FA cells and PA cells from the training set (n = 45 cells each). An algorithmic solution was developed and then validated on an independent series comprised of 1904 FA cells from 5 FA cases and 690 PA cells from 5 PA cases. RESULTS. The solution from the CRI-MLS classifier showed 1876 FA cells (98.5{\%}) as true FA and 28 FA cells (1.5{\%}) as false PA, whereas 663 PA cells (96{\%}) were true PA and 27 PA cells (4{\%}) were false FA. The summary result of this solution was a sensitivity of 98.5{\%}, a specificity of 96.1{\%}, and a positive predictive value of 98.6{\%}. CONCLUSIONS. The best spatial-spectral imaging solution was able to correctly classify 2534 of 2594 cells (98{\%}) and misclassifled only 55 of 2594 cells (2{\%}). These data suggest that this technology may be valuable in a clinical setting to help differentiate and classify morphologically similar lesions.",
keywords = "Artificial intelligence, CRI-MLS software, Cytology, Cytopathology, Neural network, Spectral-spatial imaging",
author = "Ibrahim Mansoor and Carola Zalles and Farhan Zahid and Kirk Gossage and Levenson, {Richard M} and Rimm, {David L.}",
year = "2008",
month = "2",
day = "25",
doi = "10.1002/cncr.23252",
language = "English (US)",
volume = "114",
pages = "22--26",
journal = "Cancer",
issn = "0008-543X",
publisher = "John Wiley and Sons Inc.",
number = "1",

}

TY - JOUR

T1 - Fine-needle aspiration of follicular adenoma versus parathyroid adenoma

T2 - The utility of multispectral imaging in differentiating lesions with subtle cytomorphologic differences

AU - Mansoor, Ibrahim

AU - Zalles, Carola

AU - Zahid, Farhan

AU - Gossage, Kirk

AU - Levenson, Richard M

AU - Rimm, David L.

PY - 2008/2/25

Y1 - 2008/2/25

N2 - BACKGROUND. Multispectral image analysis is an emerging tool that utilizes both spatial and spectral image information to classify images that can be used for the differentiation between benign versus malignant cells. The aim of the current study was to analyze the ability of this tool in differentiating subtle cytologic differences that cannot be appreciated by the human eye. Herein, the authors used fine-needle aspirations (FNAs) of follicular adenoma (FA) and parathyroid adenoma (PA) as a test case. METHODS. The Nuance platform was used to collect image stacks that were subsequently analyzed with CRI-MLS software, a neural network-based artificial intelligence system that can classify images using automatically "learned" spatial-spectral features. CRI-MLS was trained on random, well-preserved FA cells and PA cells from the training set (n = 45 cells each). An algorithmic solution was developed and then validated on an independent series comprised of 1904 FA cells from 5 FA cases and 690 PA cells from 5 PA cases. RESULTS. The solution from the CRI-MLS classifier showed 1876 FA cells (98.5%) as true FA and 28 FA cells (1.5%) as false PA, whereas 663 PA cells (96%) were true PA and 27 PA cells (4%) were false FA. The summary result of this solution was a sensitivity of 98.5%, a specificity of 96.1%, and a positive predictive value of 98.6%. CONCLUSIONS. The best spatial-spectral imaging solution was able to correctly classify 2534 of 2594 cells (98%) and misclassifled only 55 of 2594 cells (2%). These data suggest that this technology may be valuable in a clinical setting to help differentiate and classify morphologically similar lesions.

AB - BACKGROUND. Multispectral image analysis is an emerging tool that utilizes both spatial and spectral image information to classify images that can be used for the differentiation between benign versus malignant cells. The aim of the current study was to analyze the ability of this tool in differentiating subtle cytologic differences that cannot be appreciated by the human eye. Herein, the authors used fine-needle aspirations (FNAs) of follicular adenoma (FA) and parathyroid adenoma (PA) as a test case. METHODS. The Nuance platform was used to collect image stacks that were subsequently analyzed with CRI-MLS software, a neural network-based artificial intelligence system that can classify images using automatically "learned" spatial-spectral features. CRI-MLS was trained on random, well-preserved FA cells and PA cells from the training set (n = 45 cells each). An algorithmic solution was developed and then validated on an independent series comprised of 1904 FA cells from 5 FA cases and 690 PA cells from 5 PA cases. RESULTS. The solution from the CRI-MLS classifier showed 1876 FA cells (98.5%) as true FA and 28 FA cells (1.5%) as false PA, whereas 663 PA cells (96%) were true PA and 27 PA cells (4%) were false FA. The summary result of this solution was a sensitivity of 98.5%, a specificity of 96.1%, and a positive predictive value of 98.6%. CONCLUSIONS. The best spatial-spectral imaging solution was able to correctly classify 2534 of 2594 cells (98%) and misclassifled only 55 of 2594 cells (2%). These data suggest that this technology may be valuable in a clinical setting to help differentiate and classify morphologically similar lesions.

KW - Artificial intelligence

KW - CRI-MLS software

KW - Cytology

KW - Cytopathology

KW - Neural network

KW - Spectral-spatial imaging

UR - http://www.scopus.com/inward/record.url?scp=39749113150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=39749113150&partnerID=8YFLogxK

U2 - 10.1002/cncr.23252

DO - 10.1002/cncr.23252

M3 - Article

C2 - 18085636

AN - SCOPUS:39749113150

VL - 114

SP - 22

EP - 26

JO - Cancer

JF - Cancer

SN - 0008-543X

IS - 1

ER -