Efficacy evaluation study comparing physician alon
e and FS-AI688 combination for diagnostic ability f
or pulmonary nodules.
fficacy evaluation test of FS-AI688
Machida Tadashi
Nippon Medical School Hospital
1-1-5 Sendagi, Bunkyo-Ku
+81-3-3822-2131
zeke@nms.ac.jp
Machida Tadashi
Nippon Medical School Hospital
1-1-5 Sendagi, Bunkyo-Ku
+81-3-3822-2131
zeke@nms.ac.jp
Pending
Aug. 31, 2024
390
Observational
1.Patients who had a chest CT scan conducted bet
ween January 1, 2020 and March 31, 2024 at Nipp
on Medical School Hospital or Amakusa Regional
Medical Center, Amakusa County Medical Associat
ion.
2.Patients who have undergone chest CT imaging
who meet the following imaging conditions
CT image size:512 x 512
Slice thickness:5 mm or less
Slice interval:5 mm or less (gapless)
Imaging conditions:Image quality is suitable for inte
rpretation. This is a chest image that allows you to
see the entire lung field.
Reconstruction conditions: There are images for bo
th lung field and mediastinal conditions (the contra
st conditions match, and the slice thickness and int
erval are the same).
1. Minor patients (under 18 years old)
2.Patients who have indicated their intention to refu
se to participate in the study by the research subje
ct or their legal representative
3.Other patients who are judged to be inappropriat
e for participation in this study by the principal inve
stigator or co-investigator.
18age old over
No limit
Both
Lung Cancer
FOM (Figure of merit)
Reading test
1.The detectability (sensitivity, average number of false positives, and positive predictive value)of the nodule unit is calculated.
2.Subgroup analysis of the detectability (sensitivity, mean number of false positives, and positive predictive value)of the nodule unit according to the following factors.
Additionally, FOM will be calculated for the interpreter group.
Nodule type
Nodule location
Relationship to the surrounding structure of the nodule
Maximum diameter of the nodule
Slice thickness
Contrast/non-contrast
Imaging research institute
Imaging equipment
Group of interpreters (specialists/late residents).
3.The detectability (sensitivity, specificity, positive predictive value, and negative predictive value) of each case is calculated.
4.Subgroup analysis of the detectability (sensitivity, specificity, positive predictive value, and negative predictive value) of the case unit is performed according to the following factors:
Slice thickness
Contrast/non-contrast
Imaging research institute
Imaging equipment
Reading physicians
5.To evaluate the effect of false positives in the nodal units
Single performance test
1.The detectability (sensitivity, average number of false positives, and positive predictive value) of the nodule unit is calculated.
2.Subgroup analysis of the detectability (sensitivity, mean number of false positives, and positive predi
ctive value)of the nodule unit according to the following factors.
Nodule type
Nodule location
Relationship to the surrounding structure of the nodule
Maximum diameter of the nodule
Slice thickness
Contrast/non-contrast
Imaging research institute
Imaging equipment
Lung field condition/mediastinum condition
3.The detectability (sensitivity, specificity, positive predictive value, and negative predictive value) of e
ach case is calculated.
4.Subgroup analysis of the detectability (sensitivity,
specificity, positive predictive value, and negative p redictive value) of the case unit is performed accor
ding to the following factors:
Slice thickness
Contrast/non-contrast
Imaging research institute
Imaging equipment
Lung field condition/mediastinum condition
common to each study
1.To evaluate the concordance rate of GS.
For GS agreement, the following is calculated:
Case adoption rate, mean number of false positives, and false negative rate of every GS author for GS
Percentage of GS by Combination of GS Creators
Positive agreement of GS author
Summarized statistic of Bounding Box by combination of GS and individual GS producers
Summarized statistic of differences in Bounding Box and relative differences by combination between
GS and individual GS producers
Summarized statistic of Dice factors for Bounding
Box between GS and individual GS producers
