1
|
Trimpl MJ, Primakov S, Lambin P, Stride EPJ, Vallis KA, Gooding MJ. Beyond automatic medical image segmentation-the spectrum between fully manual and fully automatic delineation. Phys Med Biol 2022; 67. [PMID: 35523158 DOI: 10.1088/1361-6560/ac6d9c] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/11/2022] [Accepted: 05/06/2022] [Indexed: 12/19/2022]
Abstract
Semi-automatic and fully automatic contouring tools have emerged as an alternative to fully manual segmentation to reduce time spent contouring and to increase contour quality and consistency. Particularly, fully automatic segmentation has seen exceptional improvements through the use of deep learning in recent years. These fully automatic methods may not require user interactions, but the resulting contours are often not suitable to be used in clinical practice without a review by the clinician. Furthermore, they need large amounts of labelled data to be available for training. This review presents alternatives to manual or fully automatic segmentation methods along the spectrum of variable user interactivity and data availability. The challenge lies to determine how much user interaction is necessary and how this user interaction can be used most effectively. While deep learning is already widely used for fully automatic tools, interactive methods are just at the starting point to be transformed by it. Interaction between clinician and machine, via artificial intelligence, can go both ways and this review will present the avenues that are being pursued to improve medical image segmentation.
Collapse
Affiliation(s)
- Michael J Trimpl
- Mirada Medical Ltd, Oxford, United Kingdom
- Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, United Kingdom
- Oxford Institute for Radiation Oncology, University of Oxford, Oxford, United Kingdom
| | - Sergey Primakov
- The D-Lab, Department of Precision Medicine, GROW-School for Oncology, Maastricht University, Maastricht, NL, The Netherlands
| | - Philippe Lambin
- The D-Lab, Department of Precision Medicine, GROW-School for Oncology, Maastricht University, Maastricht, NL, The Netherlands
| | - Eleanor P J Stride
- Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, United Kingdom
| | - Katherine A Vallis
- Oxford Institute for Radiation Oncology, University of Oxford, Oxford, United Kingdom
| | | |
Collapse
|
2
|
Brucato G, Appelbaum PS, Hesson H, Shea EA, Dishy G, Lee K, Pia T, Syed F, Villalobos A, Wall MM, Lieberman JA, Girgis RR. Psychotic symptoms in mass shootings v. mass murders not involving firearms: findings from the Columbia mass murder database. Psychol Med 2021; 52:1-9. [PMID: 33595428 DOI: 10.1017/s0033291721000076] [Citation(s) in RCA: 8] [Impact Index Per Article: 2.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
Abstract
BACKGROUND Mass shootings account for a small fraction of annual worldwide murders, yet disproportionately affect society and influence policy. Evidence suggesting a link between mass shootings and severe mental illness (i.e. involving psychosis) is often misrepresented, generating stigma. Thus, the actual prevalence constitutes a key public health concern. METHODS We examined global personal-cause mass murders from 1900 to 2019, amassed by review of 14 785 murders publicly described in English in print or online, and collected information regarding perpetrator, demographics, legal history, drug use and alcohol misuse, and history of symptoms of psychiatric or neurologic illness using standardized methods. We distinguished whether firearms were or were not used, and, if so, the type (non-automatic v. semi- or fully-automatic). RESULTS We identified 1315 mass murders, 65% of which involved firearms. Lifetime psychotic symptoms were noted among 11% of perpetrators, consistent with previous reports, including 18% of mass murderers who did not use firearms and 8% of those who did (χ2 = 28.0, p < 0.01). US-based mass shooters were more likely to have legal histories, use recreational drugs or misuse alcohol, or have histories of non-psychotic psychiatric or neurologic symptoms. US-based mass shooters with symptoms of any psychiatric or neurologic illness more frequently used semi-or fully-automatic firearms. CONCLUSIONS These results suggest that policies aimed at preventing mass shootings by focusing on serious mental illness, characterized by psychotic symptoms, may have limited impact. Policies such as those targeting firearm access, recreational drug use and alcohol misuse, legal history, and non-psychotic psychopathology might yield more substantial results.
Collapse
Affiliation(s)
- Gary Brucato
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Paul S Appelbaum
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Hannah Hesson
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Eileen A Shea
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Gabriella Dishy
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Kathryn Lee
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Tyler Pia
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Faizan Syed
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Alexandra Villalobos
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Melanie M Wall
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Jeffrey A Lieberman
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| | - Ragy R Girgis
- Department of Psychiatry, Columbia University Irving Medical Center, New York State Psychiatric Institute, 1051 Riverside Drive Unit 31, New York, NY10032
| |
Collapse
|
3
|
Vo Chieu VD, Wacker F, Rieder C, Pöhler GH, Schumann C, Ballhausen H, Ringe KI. Ablation zone geometry after CT-guided hepatic microwave ablation: evaluation of a semi-automatic software and comparison of two different ablation systems. Int J Hyperthermia 2020; 37:533-541. [PMID: 32468872 DOI: 10.1080/02656736.2020.1766704] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/24/2022] Open
Abstract
Purpose: The aims of this study were to evaluate a semi-automatic segmentation software for assessment of ablation zone geometry in computed tomography (CT)-guided microwave ablation (MWA) of liver tumors and to compare two different MWA systems.Material and Methods: 27 patients with 40 hepatic tumors (primary liver tumor n = 20, metastases n = 20) referred for CT-guided MWA were included in this retrospective IRB-approved study. MWA was performed using two systems (system 1: 915 MHz; n = 20; system 2: 2.45 GHz; n = 20). Ablation zone segmentation and ellipticity index calculations were performed using SAFIR (Software Assistant for Interventional Radiology). To validate semi-automatic software calculations, results (2 perpendicular diameters, ellipticity index, volume) were compared with those of manual analysis (intraclass correlation, Pearson's correlation, Mann-Whitney U test; p < 0.05 deemed significant.Results: Manual measurements of mean maximum ablation zone diameters were 43 mm (system 1) and 34 mm (system 2), respectively. Correlations between manual and semi-automatic measurements were r = 0.72 and r = 0.66 (both p < 0.0001) for perpendicular diameters, and r = 0.98 (p < 0.001) for volume. Manual analysis demonstrated that ablation zones created with system 2 had a significantly lower ellipticity index compared to system 1 (mean 1.17 vs. 1.86, p < 0.0001). Results correlated significantly with semi-automatic software measurements (r = 0.71, p < 0.0001).Conclusion: Semi-automatic assessment of ablation zone geometry using SAFIR is feasible. Software-assisted evaluation of ablation zones may prove beneficial with complex ablation procedures, especially for less experienced operators. The 2.45 GHz MWA system generated a significantly more spherical ablation zone compared to the 915 MHz system. The choice of a specific MWA system significantly influences ablation zone geometry.
Collapse
Affiliation(s)
- Van Dai Vo Chieu
- Department of Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany
| | - Frank Wacker
- Department of Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany
| | - Christian Rieder
- Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany
| | - Gesa H Pöhler
- Department of Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany
| | | | - Hanne Ballhausen
- Fraunhofer Institute for Digital Medicine MEVIS, Bremen, Germany
| | - Kristina I Ringe
- Department of Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany
| |
Collapse
|
4
|
McCarthy CS, Ramprashad A, Thompson C, Botti JA, Coman IL, Kates WR. A comparison of FreeSurfer-generated data with and without manual intervention. Front Neurosci 2015; 9:379. [PMID: 26539075 PMCID: PMC4612506 DOI: 10.3389/fnins.2015.00379] [Citation(s) in RCA: 87] [Impact Index Per Article: 9.7] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/14/2015] [Accepted: 09/29/2015] [Indexed: 01/18/2023] Open
Abstract
This paper examined whether FreeSurfer-generated data differed between a fully-automated, unedited pipeline and an edited pipeline that included the application of control points to correct errors in white matter segmentation. In a sample of 30 individuals, we compared the summary statistics of surface area, white matter volumes, and cortical thickness derived from edited and unedited datasets for the 34 regions of interest (ROIs) that FreeSurfer (FS) generates. To determine whether applying control points would alter the detection of significant differences between patient and typical groups, effect sizes between edited and unedited conditions in individuals with the genetic disorder, 22q11.2 deletion syndrome (22q11DS) were compared to neurotypical controls. Analyses were conducted with data that were generated from both a 1.5 tesla and a 3 tesla scanner. For 1.5 tesla data, mean area, volume, and thickness measures did not differ significantly between edited and unedited regions, with the exception of rostral anterior cingulate thickness, lateral orbitofrontal white matter, superior parietal white matter, and precentral gyral thickness. Results were similar for surface area and white matter volumes generated from the 3 tesla scanner. For cortical thickness measures however, seven edited ROI measures, primarily in frontal and temporal regions, differed significantly from their unedited counterparts, and three additional ROI measures approached significance. Mean effect sizes for edited ROIs did not differ from most unedited ROIs for either 1.5 or 3 tesla data. Taken together, these results suggest that although the application of control points may increase the validity of intensity normalization and, ultimately, segmentation, it may not affect the final, extracted metrics that FS generates. Potential exceptions to and limitations of these conclusions are discussed.
Collapse
Affiliation(s)
- Christopher S McCarthy
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| | - Avinash Ramprashad
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| | - Carlie Thompson
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| | - Jo-Anna Botti
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| | - Ioana L Coman
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| | - Wendy R Kates
- Department of Psychiatry and Behavioral Sciences, Center for Psychiatric Neuroimaging, State University of New York at Upstate Medical University Syracuse, NY, USA
| |
Collapse
|