Performance of CAD

Several academic institutions have conducted clinical trials to demonstrate the performance of their CAD systems (Kiss et al. 2002; Nappi et al. 2004b; Nappi and Yoshida 2003; Paik et al. 2004; Summers et al. 2001; Yoshida et al. 2002a, b; Yoshida and Nappi 2001) that implement the full CAD scheme in the previous section or a part of it. In these studies, optical colonoscopy was used as the gold standard, i.e., the locations of the polyps detected by CAD were compared with the "true" locations of polyps that were determined visually in CTC data sets based on colonoscopy reports. In most of theses studies, the performance of CAD was evaluated on CTC cases that were collected retrospectively at a single institution, and that were acquired with a protocol that is currently widely used for CTC, i.e., standard pre-colonoscopy cathartic bowel cleansing, insufflation of the colon with room air or carbon dioxide, and standard-dose CT scanning with CT parameter settings such as the following: a collimation of 2.55.0 mm, pitch of 1-2, a tube current of 50-200 mA, and a reconstruction interval of 1.25-3.0 mm.

Among the studies published in peer-reviewed journals that describe a full CAD scheme, the CAD scheme developed at the University of Chicago yielded a 95% by-polyp sensitivity, with an average of 1.5 false positives per patient (0.7 false positives per b a c

Fig. 11.3a-d. Example of polyps detected by CAD. (Reprint, with permission, from Yqshida and Dachman 2005)

Fig. 11.3a-d. Example of polyps detected by CAD. (Reprint, with permission, from Yqshida and Dachman 2005)

data set), based on 72 patients (144 data sets), including a total of 21 polyps >5 mm in 14 patients. In a by-patient analysis, the sensitivity was 100%, with 1.3 false positives per patient (Nappi and Yoshida

2003). The same group reported, in a follow-up study that was published in a conference proceedings paper, a 93% by-polyp sensitivity with 4.0 false positives per patient (2.0 false positives per data set) based on 121 patients (242 data sets), including a total of 42 polyps >5 mm in 28 patients (Nappi et al. 2004b). Figure 11.4 shows a free-response receiver-operating characteristic (FROC) curve that shows the sensitivity of this CAD scheme as a function of the average number of false positives per data set. Generally, sensitivity of CAD increases as the number of false positives increases.

The CAD system at the University Hospital Gast-huisberg achieved an 80% by-polyp sensitivity, with 8.2 false positives per patient (4.1 false positives per data set), based on 18 patients, with 15 polyps >5 mm in 9 patients (Kiss et al. 2002). In this study, fecal tagging was used for most of the cases. A group at Stanford reported a 100% sensitivity with 7.0 false positives per data set (only the supine data set of each patient was used) based on 8 patients that included a total of 7 polyps >10 mm in 4 patients (Paik et al.

2004). The sensitivity was less than 50% at the same false positive rate for 11 polyps between 5 and 9 mm that were found in 3 of the above 8 patients. A group at the NIH reported a 90% sensitivity with 15.7 false positives per data set, based on 40 patients (80 data sets) that included a total of 39 polyps >3 mm in 20 patients (Jerebko et al. 2003b). In a separate study, they reported that multiple artificial neural networks could potentially be employed to increase the sensitivity by an average of 6.9% and to reduce the false-positive rate by 30-36% (Jerebko et al. 2003a, Yao et al. 2004).

These studies indicate that CAD is promising in detecting polyps with high sensitivity and a low false-positive rate. It appears that the detection performance can reach up to 100% by-patient sensitivity with 1.3 false positives per patient for polyps >5 mm (Nappi and Yoshida 2003). Generally, however, the performance of CAD systems appears to range between 70 and 100% by-patient sensitivity for

Fig. 11.4. Free-response receiver-operating characteristic curve showing the performance of CAD in the detection of polyps

b a d c polyps >6 mm, with 2-8 false positives per patient. A meta-analysis of the reported performance of CTC showed that, for human readers, the pooled by-patient sensitivity for polyps >10 mm and for those 6-9 mm was 85 and 70%, respectively (Mulhall et al. 2005). Comparing this performance with that of CAD, it appears that the performance, especially the sensitivity, of CAD is approaching that of an average human reader.

Was this article helpful?

0 0
Managing Diverticular Disease

Managing Diverticular Disease

Stop The Pain. Manage Your Diverticular Disease And Live A Pain Free Life. No Pain, No Fear, Full Control Normal Life Again. Diverticular Disease can stop you from doing all the things you love. Seeing friends, playing with the kids... even trying to watch your favorite television shows.

Get My Free Ebook


Post a comment