Identification of Anonymous MRI Research Participants with Face-Recognition Software.
Ontology highlight
ABSTRACT:
SUBMITTER: Schwarz CG
PROVIDER: S-EPMC7091256 | biostudies-literature | 2019 Oct
REPOSITORIES: biostudies-literature
Publications
Identification of Anonymous MRI Research Participants with Face-Recognition Software.
Schwarz Christopher G CG Kremers Walter K WK Therneau Terry M TM Sharp Richard R RR Gunter Jeffrey L JL Vemuri Prashanthi P Arani Arvin A Spychalla Anthony J AJ Kantarci Kejal K Knopman David S DS Petersen Ronald C RC Jack Clifford R CR
The New England journal of medicine 20191001 17
Similar Datasets
Project description:In recent years, wide deployment of automatic face recognition systems has been accompanied by substantial gains in algorithm performance. However, benchmarking tests designed to evaluate these systems do not account for the errors of human operators, who are often an integral part of face recognition solutions in forensic and security settings. This causes a mismatch between evaluation tests and operational accuracy. We address this by measuring user performance in a face recognition system used to screen passport applications for identity fraud. Experiment 1 measured target detection accuracy in algorithm-generated 'candidate lists' selected from a large database of passport images. Accuracy was notably poorer than in previous studies of unfamiliar face matching: participants made over 50% errors for adult target faces, and over 60% when matching images of children. Experiment 2 then compared performance of student participants to trained passport officers-who use the system in their daily work-and found equivalent performance in these groups. Encouragingly, a group of highly trained and experienced "facial examiners" outperformed these groups by 20 percentage points. We conclude that human performance curtails accuracy of face recognition systems-potentially reducing benchmark estimates by 50% in operational settings. Mere practise does not attenuate these limits, but superior performance of trained examiners suggests that recruitment and selection of human operators, in combination with effective training and mentorship, can improve the operational accuracy of face recognition systems.
| S-EPMC4605725 | biostudies-literature
Project description:Face recognition, as one of the major biometrics identification methods, has been applied in different fields involving economics, military, e-commerce, and security. Its touchless identification process and non-compulsory rule to users are irreplaceable by other approaches, such as iris recognition or fingerprint recognition. Among all face recognition techniques, principal component analysis (PCA), proposed in the earliest stage, still attracts researchers because of its property of reducing data dimensionality without losing important information. Nevertheless, establishing a PCA-based face recognition system is still time-consuming, since there are different problems that need to be considered in practical applications, such as illumination, facial expression, or shooting angle. Furthermore, it still costs a lot of effort for software developers to integrate toolkit implementations in applications. This paper provides a software framework for PCA-based face recognition aimed at assisting software developers to customize their applications efficiently. The framework describes the complete process of PCA-based face recognition, and in each step, multiple variations are offered for different requirements. Some of the variations in the same step can work collaboratively and some steps can be omitted in specific situations; thus, the total number of variations exceeds 150. The implementation of all approaches presented in the framework is provided.
| S-EPMC8384131 | biostudies-literature
Project description:Diffusion MRI (dMRI) is the only noninvasive method for mapping white matter connections in the brain. We describe SlicerDMRI, a software suite that enables visualization and analysis of dMRI for neuroscientific studies and patient-specific anatomic assessment. SlicerDMRI has been successfully applied in multiple studies of the human brain in health and disease, and here, we especially focus on its cancer research applications. As an extension module of the 3D Slicer medical image computing platform, the SlicerDMRI suite enables dMRI analysis in a clinically relevant multimodal imaging workflow. Core SlicerDMRI functionality includes diffusion tensor estimation, white matter tractography with single and multi-fiber models, and dMRI quantification. SlicerDMRI supports clinical DICOM and research file formats, is open-source and cross-platform, and can be installed as an extension to 3D Slicer (www.slicer.org). More information, videos, tutorials, and sample data are available at dmri.slicer.org Cancer Res; 77(21); e101-3. ©2017 AACR.
| S-EPMC5679308 | biostudies-literature
Project description:It is well known that de-identified research brain images from MRI and CT can potentially be re-identified using face recognition; however, this has not been examined for PET images. We generated face reconstruction images of 182 volunteers using amyloid, tau, and FDG PET scans, and we measured how accurately commercial face recognition software (Microsoft Azure's Face API) automatically matched them with the individual participants' face photographs. We then compared this accuracy with the same experiments using participants' CT and MRI. Face reconstructions from PET images from PET/CT scanners were correctly matched at rates of 42% (FDG), 35% (tau), and 32% (amyloid), while CT were matched at 78% and MRI at 97-98%. We propose that these recognition rates are high enough that research studies should consider using face de-identification ("de-facing") software on PET images, in addition to CT and structural MRI, before data sharing. We also updated our mri_reface de-identification software with extended functionality to replace face imagery in PET and CT images. Rates of face recognition on de-faced images were reduced to 0-4% for PET, 5% for CT, and 8% for MRI. We measured the effects of de-facing on regional amyloid PET measurements from two different measurement pipelines (PETSurfer/FreeSurfer 6.0, and one in-house method based on SPM12 and ANTs), and these effects were small: ICC values between de-faced and original images were > 0.98, biases were <2%, and median relative errors were < 2%. Effects on global amyloid PET SUVR measurements were even smaller: ICC values were 1.00, biases were <0.5%, and median relative errors were also <0.5%.
| S-EPMC9358410 | biostudies-literature
Project description:It is now widely known that research brain MRI, CT, and PET images may potentially be re-identified using face recognition, and this potential can be reduced by applying face-deidentification ("de-facing") software. However, for research MRI sequences beyond T1-weighted (T1-w) and T2-FLAIR structural images, the potential for re-identification and quantitative effects of de-facing are both unknown, and the effects of de-facing T2-FLAIR are also unknown. In this work we examine these questions (where applicable) for T1-w, T2-w, T2*-w, T2-FLAIR, diffusion MRI (dMRI), functional MRI (fMRI), and arterial spin labelling (ASL) sequences. Among current-generation, vendor-product research-grade sequences, we found that 3D T1-w, T2-w, and T2-FLAIR were highly re-identifiable (96-98%). 2D T2-FLAIR and 3D multi-echo GRE (ME-GRE) were also moderately re-identifiable (44-45%), and our derived T2* from ME-GRE (comparable to a typical 2D T2*) matched at only 10%. Finally, diffusion, functional and ASL images were each minimally re-identifiable (0-8%). Applying de-facing with mri_reface version 0.3 reduced successful re-identification to ≤8%, while differential effects on popular quantitative pipelines for cortical volumes and thickness, white matter hyperintensities (WMH), and quantitative susceptibility mapping (QSM) measurements were all either comparable with or smaller than scan-rescan estimates. Consequently, high-quality de-facing software can greatly reduce the risk of re-identification for identifiable MRI sequences with only negligible effects on automated intracranial measurements. The current-generation echo-planar and spiral sequences (dMRI, fMRI, and ASL) each had minimal match rates, suggesting that they have a low risk of re-identification and can be shared without de-facing, but this conclusion should be re-evaluated if they are acquired without fat suppression, with a full-face scan coverage, or if newer developments reduce the current levels of artifacts and distortion around the face.
| S-EPMC10389782 | biostudies-literature
Project description:Achieving the upper limits of face identification accuracy in forensic applications can minimize errors that have profound social and personal consequences. Although forensic examiners identify faces in these applications, systematic tests of their accuracy are rare. How can we achieve the most accurate face identification: using people and/or machines working alone or in collaboration? In a comprehensive comparison of face identification by humans and computers, we found that forensic facial examiners, facial reviewers, and superrecognizers were more accurate than fingerprint examiners and students on a challenging face identification test. Individual performance on the test varied widely. On the same test, four deep convolutional neural networks (DCNNs), developed between 2015 and 2017, identified faces within the range of human accuracy. Accuracy of the algorithms increased steadily over time, with the most recent DCNN scoring above the median of the forensic facial examiners. Using crowd-sourcing methods, we fused the judgments of multiple forensic facial examiners by averaging their rating-based identity judgments. Accuracy was substantially better for fused judgments than for individuals working alone. Fusion also served to stabilize performance, boosting the scores of lower-performing individuals and decreasing variability. Single forensic facial examiners fused with the best algorithm were more accurate than the combination of two examiners. Therefore, collaboration among humans and between humans and machines offers tangible benefits to face identification accuracy in important applications. These results offer an evidence-based roadmap for achieving the most accurate face identification possible.
| S-EPMC6004481 | biostudies-literature
Project description:While there have been decades of clinical and theoretical interest in developmental and acquired face recognition difficulties, very little work has examined their remediation. Here, we report two studies that examined the efficacy of an existing face training programme in improving face-processing skills in adults and children with developmental face recognition impairments. The programme has only been trialled in typical children to date, where 2 weeks of perceptual training (modelled on an adapted version of the popular family game Guess Who?) resulted in face-specific improvements for memory but not perception after 2 weeks of training. In Study 1, we performed a randomized, parallel groups, placebo-controlled trial of the same programme in 20 adults with a pre-existing diagnosis of developmental prosopagnosia. Assessment tasks were administered immediately before and after training, and 2 weeks later. Face-specific gains in memory (but not perception) were observed in the experimental group and were greatest in those with the poorest face recognition skills at entry. These gains persisted 2 weeks after training ceased. In Study 2, a case-series approach was used to administer the experimental version of the training programme to four children who presented with difficulties in face recognition. Improvements in face memory were observed in three of the participants; while one also improved at face perception, there was mixed evidence for the face specificity of these gains. Together, these findings suggest plasticity in the human face recognition system through to at least mid-adulthood and also pave the way for longer-term implementations of the face training programme that will likely elicit greater gains in both adults and children.
| S-EPMC8977649 | biostudies-literature
Project description:The ability to recognize an individual face is essential to human social interaction. Even subtle errors in this process can have huge implications for the way we relate to social partners. Because autism spectrum disorder (ASD) is characterized by deficits in social interaction, researchers have theorized about the potential role of atypical face identity processing to the symptom profile of ASD for more than 40 years. We conducted an empirical meta-analysis of this large literature to determine whether and to what extent face identity processing is atypical in ASD compared to typically developing (TD) individuals. We also tested the hypotheses that the deficit is selective to face identity recognition, not perception, and that methodological variation across studies moderates the magnitude of the estimated deficit. We identified 112 studies (5,390 participants) that generated 172 effect sizes from both recognition (k = 119) and discrimination (k = 53) paradigms. We used state-of-the-art approaches for assessing the validity and robustness of the analyses. We found comparable and large deficits in ASD for both face identity recognition (Hedge's g = -0.86) and discrimination (Hedge's g = -0.82). This means that the score of an average ASD individual is nearly 1 SD below the average TD individual on tasks assessing both aspects of face identity processing. These deficits generalize across age groups, sex, IQ scores, and task paradigms. These findings suggest that deficits in face identity processing may represent a core deficit in ASD. (PsycInfo Database Record (c) 2021 APA, all rights reserved).
| S-EPMC8961473 | biostudies-literature
Project description:In the last decade, a novel individual differences approach has emerged across the face recognition literature. While the field has long been concerned with prosopagnosia (the inability to recognise facial identity), it has more recently become clear that there are vast differences in face recognition ability within the typical population. "Super-recognisers" are those individuals purported to reside at the very top of this spectrum. On one hand, these people are of interest to cognitive neuropsychologists who are motivated to explore the commonality of the face recognition continuum, whereas on the other hand, researchers from the forensic face matching field evaluate the implementation of super-recognisers into real-world police and security settings. These two rather different approaches have led to discrepancies in the definition of super-recognisers, and perhaps more fundamentally, the approach to identifying them, resulting in a lack of consistency that prohibits theoretical progress. Here, we review the protocols used in published work to identify super-recognisers, and propose a common definition and screening recommendations that can be adhered to across fields.
| S-EPMC8531948 | biostudies-literature
Project description:Face recognition strategies do not generalize across individuals. Many studies have reported robust cultural differences between West Europeans/North Americans and East Asians in eye movement strategies during face recognition. The social orientation hypothesis posits that individualistic vs. collectivistic (IND/COL) value systems, respectively defining West European/North American and East Asian societies, would be at the root of many cultural differences in visual perception. Whether social orientation is also responsible for such cultural contrast in face recognition remains to be clarified. To this aim, we conducted two experiments with West European/North American and Chinese observers. In Experiment 1, we probed the existence of a link between IND/COL social values and eye movements during face recognition, by using an IND/COL priming paradigm. In Experiment 2, we dissected the latter relationship in greater depth, by using two IND/COL questionnaires, including subdimensions to those concepts. In both studies, cultural differences in fixation patterns were revealed between West European/North American and East Asian observers. Priming IND/COL values did not modulate eye movement visual sampling strategies, and only specific subdimensions of the IND/COL questionnaires were associated with distinct eye-movement patterns. Altogether, we show that the typical contrast between IND/COL cannot fully account for cultural differences in eye movement strategies for face recognition. Cultural differences in eye movements for faces might originate from mechanisms distinct from social orientation.
| S-EPMC10721205 | biostudies-literature