scholarly journals Subject-centered free-response ROC (FROC) analysis

2013 ◽  
Vol 40 (5) ◽  
pp. 051706 ◽  
Author(s):  
Andriy I. Bandos ◽  
Howard E. Rockette ◽  
David Gur
Keyword(s):  
2021 ◽  
pp. 159101992110009
Author(s):  
Xinke Liu ◽  
Junqiang Feng ◽  
Zhenzhou Wu ◽  
Zhonghao Neo ◽  
Chengcheng Zhu ◽  
...  

Objective Accurate diagnosis and measurement of intracranial aneurysms are challenging. This study aimed to develop a 3D convolutional neural network (CNN) model to detect and segment intracranial aneurysms (IA) on 3D rotational DSA (3D-RA) images. Methods 3D-RA images were collected and annotated by 5 neuroradiologists. The annotated images were then divided into three datasets: training, validation, and test. A 3D Dense-UNet-like CNN (3D-Dense-UNet) segmentation algorithm was constructed and trained using the training dataset. Diagnostic performance to detect aneurysms and segmentation accuracy was assessed for the final model on the test dataset using the free-response receiver operating characteristic (FROC). Finally, the CNN-inferred maximum diameter was compared against expert measurements by Pearson’s correlation and Bland-Altman limits of agreement (LOA). Results A total of 451 patients with 3D-RA images were split into n = 347/41/63 training/validation/test datasets, respectively. For aneurysm detection, observed FROC analysis showed that the model managed to attain a sensitivity of 0.710 at 0.159 false positives (FP)/case, and 0.986 at 1.49 FP/case. The proposed method had good agreement with reference manual aneurysmal maximum diameter measurements (8.3 ± 4.3 mm vs. 7.8 ± 4.8 mm), with a correlation coefficient r = 0.77, small bias of 0.24 mm, and LOA of -6.2 to 5.71 mm. 37.0% and 77% of diameter measurements were within ±1 mm and ±2.5 mm of expert measurements. Conclusions A 3D-Dense-UNet model can detect and segment aneurysms with relatively high accuracy using 3D-RA images. The automatically measured maximum diameter has potential clinical application value.


2017 ◽  
Vol 0 (0) ◽  
Author(s):  
Ehsan Kozegar ◽  
Mohsen Soryani

AbstractMammography is the most widely used modality for early breast cancer detection. This work proposes a new computer-aided mass detection approach, in which a denoising method called BM3D is first applied to mammograms. Afterwards, using an adaptive segmentation algorithm, images are segmented to suspicious regions of interest (ROIs) and then a classifier is used to understand the features of true positive (TP) and false positive (FP) patterns. In this way, from selected suspicious ROIs, fractal dimension, texture and intensity features are extracted. Subsequently, a discretization approach followed by correlation-based feature selection (CFS) is combined with a genetic algorithm to obtain the most representative features. To neutralize the classifier’s bias in favor of the major class in imbalanced datasets, an oversampling algorithm is used. In the next step, a cost-sensitive ensemble classifier based on a trainable combiner is proposed in order to reduce the number of FP samples. Finally, the presented method is validated on miniMIAS and INBreast datasets. The free-response receiver operating characteristic (FROC) analysis results prove the efficiency of the proposed approach. A sensitivity of 88% and false positive per image (FPpI) of 0.78 for miniMIAS and also a sensitivity of 86% and FPpI of 0.75 for INBreast dataset were obtained.


1991 ◽  
Vol 47 (4) ◽  
pp. 620-626 ◽  
Author(s):  
JUNJI SHIRAISHI ◽  
KAZUHISA KOSAKAI ◽  
MASAKATSU HATAGAWA ◽  
MITSUJI HIGASHIDA ◽  
SHINICHI WATANABE

2020 ◽  
Vol 132 (5) ◽  
pp. 1405-1413 ◽  
Author(s):  
Michael D. Staudt ◽  
Holger Joswig ◽  
Gwynedd E. Pickett ◽  
Keith W. MacDougall ◽  
Andrew G. Parrent

OBJECTIVEThe prevalence of trigeminal neuralgia (TN) in patients with multiple sclerosis (MS-TN) is higher than in the general population (idiopathic TN [ITN]). Glycerol rhizotomy (GR) is a percutaneous lesioning surgery commonly performed for the treatment of medically refractory TN. While treatment for acute pain relief is excellent, long-term pain relief is poorer. The object of this study was to assess the efficacy of percutaneous retrogasserian GR for the treatment of MS-TN versus ITN.METHODSA retrospective chart review was performed, identifying 219 patients who had undergone 401 GR procedures from 1983 to 2018 at a single academic institution. All patients were diagnosed with medically refractory MS-TN (182 procedures) or ITN (219 procedures). The primary outcome measures of interest were immediate pain relief and time to pain recurrence following initial and repeat GR procedures. Secondary outcomes included medication usage and presence of periprocedural hypesthesia.RESULTSThe initial pain-free response rate was similar between groups (p = 0.726): MS-TN initial GR 89.6%; MS-TN repeat GR 91.9%; ITN initial GR 89.6%; ITN repeat GR 87.0%. The median time to recurrence after initial GR was similar between MS-TN (2.7 ± 1.3 years) and ITN (2.1 ± 0.6 years) patients (p = 0.87). However, there was a statistically significant difference in the time to recurrence after repeat GR between MS-TN (2.3 ± 0.5 years) and ITN patients (1.2 ± 0.2 years; p < 0.05). The presence of periprocedural hypesthesia was highly predictive of pain-free survival (p < 0.01).CONCLUSIONSPatients with MS-TN achieve meaningful pain relief following GR, with an efficacy comparable to that following GR in patients with ITN. Initial and subsequent GR procedures are equally efficacious.


2011 ◽  
Vol 60 (10) ◽  
pp. 815-824 ◽  
Author(s):  
Mauro Cimino ◽  
Prabhakar R. Pagilla

1990 ◽  
Vol 1990 (1) ◽  
pp. i-29 ◽  
Author(s):  
Randy Elliot Bennett ◽  
Donald A. Rock ◽  
Minhwei Wang

1963 ◽  
Vol 13 (1) ◽  
pp. 159-169 ◽  
Author(s):  
Peter F. Merenda ◽  
Walter V. Clarke

Two self-rating adjective check lists (ACL) were administered to 44 students in a course on psychological measurement. The first ACL administered was the regular free response list, followed immediately with a forced-choice version in which the adjectives were arranged into tetrad sets. Ipsative scoring was used and profiles compared. The correlations between the profiles ranged from −1.00 to 1.00, more than 40% falling in the negative range. Ss gave their impressions and reactions to both inventories, and evaluated the relative validity of the results. A majority favored the free-response technique and felt that it would yield a more accurate description of their self-concepts and personality structures. General consensus was that the forced-choice instructions led to frustration, increased irritability, and decreased motivation. Ss felt that the free-response instrument presented a more relaxing situation and was even enjoyable to take. Ss' verbal reactions were consistent with the statistical results. The findings suggest that the forced-choice method is likely to be inappropriate for use with adjective check lists in self-concept assessment and analysis, and may lead not only to distortion in the personality profiles, but also to reversals.


Sign in / Sign up

Export Citation Format

Share Document