Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems

Giuseppe Raso, Donato Cascio, Francesco Fauci, Marius Iacomi, Maria Simone Vasile, Debora Castrogiovanni, Guido Filosto, Raffaele Ienzi, Marius Mihail Iacomi

Risultato della ricerca: Article

12 Citazioni (Scopus)

Abstract

Aim: Within this work, a comparative analysis of two commercial computer-aideddetection or diagnosis (CAD) systems, CyclopusCAD® mammo (v. 6.0) produced byCyclopusCAD Ltd (Palermo, Italy) and SecondLook® (v. 6.1C) produced by iCAD Inc.(OH, USA) is performed by evaluating the results of both systems application on anunique set of mammographic digital images routinely acquired in a hospital structure.Materials &amp; methods: The two CAD systems have been separately applied on a sampleset of 126 mammographic digital cases, having been independently diagnosed by twosenior radiologists. According to the human diagnosis, the cases in the sample referenceset are divided into 61 negatives and 65 pathological cases (21 cases displaying bothmass lesions and microcalcifications and 44 cases characterized only by mass lesions).The images in the pathological subset contain 123 human diagnosed mass lesionsand 37 human diagnosed microcalcifications clusters. In the case of CyclopusCAD,the system offered the possibility to evaluate sensitivity at several threshold levels(working points); five different setting levels (high sensitivity, normal sensitivity,standard, normal specificity and high specificity) have been used. Results: At thestandard threshold level, CyclopusCAD exhibits an overall sensitivity of 83.1 versus66.2% for iCAD (p = 0.04) and an average number of false positives per image (FP/im)of 1.38 against 0.47 for iCAD (p < 0.01). Specifically, for the mass lesions, CyclopusCADexhibits a sensitivity of 76.9% at a rate of 0.73 FP/im, while iCAD displays a sensitivityof 61.5% at 0.28 FP/im. For the microcalcifications, CyclopusCAD exhibits a sensitivityof 76.2% at a rate of 0.64 FP/image, while iCAD displays a sensitivity of 61.9% at0.19 FP/im. The reported results have also been expressed in terms of free-responsereceiver operating characteristic curves, corresponding to five different thresholds inthe case of CyclopusCAD and to one single threshold value for iCAD. Conclusion: Theoverall accuracies of the two systems are fairly comparable up to the uncertainty levelof this analysis. CyclopusCAD may reach a higher sensitivity level for both masses andmicrocalcifications owing to the flexibility in the working point choice, with the priceof a major number of FP/im.
Lingua originaleEnglish
pagine (da-a)13-20
Numero di pagine8
RivistaImaging in Medicine
Volume6
Stato di pubblicazionePublished - 2014

Fingerprint

Calcinosis
Italy
Uncertainty

All Science Journal Classification (ASJC) codes

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging

Cita questo

Raso, G., Cascio, D., Fauci, F., Iacomi, M., Vasile, M. S., Castrogiovanni, D., ... Iacomi, M. M. (2014). Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems. Imaging in Medicine, 6, 13-20.

Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems. / Raso, Giuseppe; Cascio, Donato; Fauci, Francesco; Iacomi, Marius; Vasile, Maria Simone; Castrogiovanni, Debora; Filosto, Guido; Ienzi, Raffaele; Iacomi, Marius Mihail.

In: Imaging in Medicine, Vol. 6, 2014, pag. 13-20.

Risultato della ricerca: Article

Raso, G, Cascio, D, Fauci, F, Iacomi, M, Vasile, MS, Castrogiovanni, D, Filosto, G, Ienzi, R & Iacomi, MM 2014, 'Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems', Imaging in Medicine, vol. 6, pagg. 13-20.
Raso, Giuseppe ; Cascio, Donato ; Fauci, Francesco ; Iacomi, Marius ; Vasile, Maria Simone ; Castrogiovanni, Debora ; Filosto, Guido ; Ienzi, Raffaele ; Iacomi, Marius Mihail. / Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems. In: Imaging in Medicine. 2014 ; Vol. 6. pagg. 13-20.
@article{a9fd5d9942c84a6da0894e6a25b437a2,
title = "Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems",
abstract = "Aim: Within this work, a comparative analysis of two commercial computer-aideddetection or diagnosis (CAD) systems, CyclopusCAD{\circledR} mammo (v. 6.0) produced byCyclopusCAD Ltd (Palermo, Italy) and SecondLook{\circledR} (v. 6.1C) produced by iCAD Inc.(OH, USA) is performed by evaluating the results of both systems application on anunique set of mammographic digital images routinely acquired in a hospital structure.Materials & methods: The two CAD systems have been separately applied on a sampleset of 126 mammographic digital cases, having been independently diagnosed by twosenior radiologists. According to the human diagnosis, the cases in the sample referenceset are divided into 61 negatives and 65 pathological cases (21 cases displaying bothmass lesions and microcalcifications and 44 cases characterized only by mass lesions).The images in the pathological subset contain 123 human diagnosed mass lesionsand 37 human diagnosed microcalcifications clusters. In the case of CyclopusCAD,the system offered the possibility to evaluate sensitivity at several threshold levels(working points); five different setting levels (high sensitivity, normal sensitivity,standard, normal specificity and high specificity) have been used. Results: At thestandard threshold level, CyclopusCAD exhibits an overall sensitivity of 83.1 versus66.2{\%} for iCAD (p = 0.04) and an average number of false positives per image (FP/im)of 1.38 against 0.47 for iCAD (p < 0.01). Specifically, for the mass lesions, CyclopusCADexhibits a sensitivity of 76.9{\%} at a rate of 0.73 FP/im, while iCAD displays a sensitivityof 61.5{\%} at 0.28 FP/im. For the microcalcifications, CyclopusCAD exhibits a sensitivityof 76.2{\%} at a rate of 0.64 FP/image, while iCAD displays a sensitivity of 61.9{\%} at0.19 FP/im. The reported results have also been expressed in terms of free-responsereceiver operating characteristic curves, corresponding to five different thresholds inthe case of CyclopusCAD and to one single threshold value for iCAD. Conclusion: Theoverall accuracies of the two systems are fairly comparable up to the uncertainty levelof this analysis. CyclopusCAD may reach a higher sensitivity level for both masses andmicrocalcifications owing to the flexibility in the working point choice, with the priceof a major number of FP/im.",
author = "Giuseppe Raso and Donato Cascio and Francesco Fauci and Marius Iacomi and Vasile, {Maria Simone} and Debora Castrogiovanni and Guido Filosto and Raffaele Ienzi and Iacomi, {Marius Mihail}",
year = "2014",
language = "English",
volume = "6",
pages = "13--20",
journal = "Imaging in Medicine",
issn = "1755-5191",
publisher = "Future Medicine Ltd.",

}

TY - JOUR

T1 - Computer-aided diagnosis in digitalmammography: comparison of twocommercial systems

AU - Raso, Giuseppe

AU - Cascio, Donato

AU - Fauci, Francesco

AU - Iacomi, Marius

AU - Vasile, Maria Simone

AU - Castrogiovanni, Debora

AU - Filosto, Guido

AU - Ienzi, Raffaele

AU - Iacomi, Marius Mihail

PY - 2014

Y1 - 2014

N2 - Aim: Within this work, a comparative analysis of two commercial computer-aideddetection or diagnosis (CAD) systems, CyclopusCAD® mammo (v. 6.0) produced byCyclopusCAD Ltd (Palermo, Italy) and SecondLook® (v. 6.1C) produced by iCAD Inc.(OH, USA) is performed by evaluating the results of both systems application on anunique set of mammographic digital images routinely acquired in a hospital structure.Materials & methods: The two CAD systems have been separately applied on a sampleset of 126 mammographic digital cases, having been independently diagnosed by twosenior radiologists. According to the human diagnosis, the cases in the sample referenceset are divided into 61 negatives and 65 pathological cases (21 cases displaying bothmass lesions and microcalcifications and 44 cases characterized only by mass lesions).The images in the pathological subset contain 123 human diagnosed mass lesionsand 37 human diagnosed microcalcifications clusters. In the case of CyclopusCAD,the system offered the possibility to evaluate sensitivity at several threshold levels(working points); five different setting levels (high sensitivity, normal sensitivity,standard, normal specificity and high specificity) have been used. Results: At thestandard threshold level, CyclopusCAD exhibits an overall sensitivity of 83.1 versus66.2% for iCAD (p = 0.04) and an average number of false positives per image (FP/im)of 1.38 against 0.47 for iCAD (p < 0.01). Specifically, for the mass lesions, CyclopusCADexhibits a sensitivity of 76.9% at a rate of 0.73 FP/im, while iCAD displays a sensitivityof 61.5% at 0.28 FP/im. For the microcalcifications, CyclopusCAD exhibits a sensitivityof 76.2% at a rate of 0.64 FP/image, while iCAD displays a sensitivity of 61.9% at0.19 FP/im. The reported results have also been expressed in terms of free-responsereceiver operating characteristic curves, corresponding to five different thresholds inthe case of CyclopusCAD and to one single threshold value for iCAD. Conclusion: Theoverall accuracies of the two systems are fairly comparable up to the uncertainty levelof this analysis. CyclopusCAD may reach a higher sensitivity level for both masses andmicrocalcifications owing to the flexibility in the working point choice, with the priceof a major number of FP/im.

AB - Aim: Within this work, a comparative analysis of two commercial computer-aideddetection or diagnosis (CAD) systems, CyclopusCAD® mammo (v. 6.0) produced byCyclopusCAD Ltd (Palermo, Italy) and SecondLook® (v. 6.1C) produced by iCAD Inc.(OH, USA) is performed by evaluating the results of both systems application on anunique set of mammographic digital images routinely acquired in a hospital structure.Materials & methods: The two CAD systems have been separately applied on a sampleset of 126 mammographic digital cases, having been independently diagnosed by twosenior radiologists. According to the human diagnosis, the cases in the sample referenceset are divided into 61 negatives and 65 pathological cases (21 cases displaying bothmass lesions and microcalcifications and 44 cases characterized only by mass lesions).The images in the pathological subset contain 123 human diagnosed mass lesionsand 37 human diagnosed microcalcifications clusters. In the case of CyclopusCAD,the system offered the possibility to evaluate sensitivity at several threshold levels(working points); five different setting levels (high sensitivity, normal sensitivity,standard, normal specificity and high specificity) have been used. Results: At thestandard threshold level, CyclopusCAD exhibits an overall sensitivity of 83.1 versus66.2% for iCAD (p = 0.04) and an average number of false positives per image (FP/im)of 1.38 against 0.47 for iCAD (p < 0.01). Specifically, for the mass lesions, CyclopusCADexhibits a sensitivity of 76.9% at a rate of 0.73 FP/im, while iCAD displays a sensitivityof 61.5% at 0.28 FP/im. For the microcalcifications, CyclopusCAD exhibits a sensitivityof 76.2% at a rate of 0.64 FP/image, while iCAD displays a sensitivity of 61.9% at0.19 FP/im. The reported results have also been expressed in terms of free-responsereceiver operating characteristic curves, corresponding to five different thresholds inthe case of CyclopusCAD and to one single threshold value for iCAD. Conclusion: Theoverall accuracies of the two systems are fairly comparable up to the uncertainty levelof this analysis. CyclopusCAD may reach a higher sensitivity level for both masses andmicrocalcifications owing to the flexibility in the working point choice, with the priceof a major number of FP/im.

UR - http://hdl.handle.net/10447/90383

UR - http://www.futuremedicine.com/doi/abs/10.2217/iim.13.68

M3 - Article

VL - 6

SP - 13

EP - 20

JO - Imaging in Medicine

JF - Imaging in Medicine

SN - 1755-5191

ER -