1
|
Warren BE, Bilbily A, Gichoya JW, Conway A, Li B, Fawzy A, Barragán C, Jaberi A, Mafeld S. An Introductory Guide to Artificial Intelligence in Interventional Radiology: Part 1 Foundational Knowledge. Can Assoc Radiol J 2024:8465371241236376. [PMID: 38445497 DOI: 10.1177/08465371241236376] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [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: 03/07/2024] Open
Abstract
Artificial intelligence (AI) is rapidly evolving and has transformative potential for interventional radiology (IR) clinical practice. However, formal training in AI may be limited for many clinicians and therefore presents a challenge for initial implementation and trust in AI. An understanding of the foundational concepts in AI may help familiarize the interventional radiologist with the field of AI, thus facilitating understanding and participation in the development and deployment of AI. A pragmatic classification system of AI based on the complexity of the model may guide clinicians in the assessment of AI. Finally, the current state of AI in IR and the patterns of implementation are explored (pre-procedural, intra-procedural, and post-procedural).
Collapse
Affiliation(s)
- Blair Edward Warren
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Alexander Bilbily
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- 16 Bit Inc., Toronto, ON, Canada
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
| | | | - Aaron Conway
- Prince Charles Hospital, Queensland University of Technology, Brisbane, QLD, Australia
| | - Ben Li
- Division of Vascular Surgery, Department of Surgery, University of Toronto, Toronto, ON, Canada
| | - Aly Fawzy
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
| | - Camilo Barragán
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Arash Jaberi
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Sebastian Mafeld
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| |
Collapse
|
2
|
Warren BE, Bilbily A, Gichoya JW, Chartier LB, Fawzy A, Barragán C, Jaberi A, Mafeld S. An Introductory Guide to Artificial Intelligence in Interventional Radiology: Part 2: Implementation Considerations and Harms. Can Assoc Radiol J 2024:8465371241236377. [PMID: 38445517 DOI: 10.1177/08465371241236377] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [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: 03/07/2024] Open
Abstract
The introduction of artificial intelligence (AI) in interventional radiology (IR) will bring about new challenges and opportunities for patients and clinicians. AI may comprise software as a medical device or AI-integrated hardware and will require a rigorous evaluation that should be guided based on the level of risk of the implementation. A hierarchy of risk of harm and possible harms are described herein. A checklist to guide deployment of an AI in a clinical IR environment is provided. As AI continues to evolve, regulation and evaluation of the AI medical devices will need to continue to evolve to keep pace and ensure patient safety.
Collapse
Affiliation(s)
- Blair Edward Warren
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Alexander Bilbily
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- 16 Bit Inc., Toronto, ON, Canada
- Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada
| | | | - Lucas B Chartier
- Division of Emergency Medicine, Department of Medicine, University of Toronto, Toronto, ON, Canada
- Department of Emergency Medicine, University Health Network, Toronto, ON, Canada
| | - Aly Fawzy
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
| | - Camilo Barragán
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Arash Jaberi
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| | - Sebastian Mafeld
- Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
- Joint Department of Medical Imaging, University Health Network, Toronto, ON, Canada
| |
Collapse
|
3
|
Warren BE, Roche-Nagle G, Zhu J, Wang G, Eisenberg N, Rajan DK, Mafeld S. Endovascular community response to mortality data in use of paclitaxel devices for peripheral vascular disease. J Vasc Surg 2021; 74:2006-2013.e2. [PMID: 34182026 DOI: 10.1016/j.jvs.2021.05.058] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/08/2020] [Accepted: 05/21/2021] [Indexed: 11/18/2022]
Abstract
OBJECTIVES Examine the endovascular community response to data demonstrating increased mortality in paclitaxel devices for the treatment of peripheral arterial disease in femoropopliteal lesions. METHODS A retrospective observational study using the Vascular Quality Initiative Peripheral Vascular Intervention registry dataset was performed examining paclitaxel device use for peripheral arterial disease in femoropopliteal arteries treated from 2017-2019. A total of 41707 patients and 52208 procedures were analyzed during the study period. Post-hoc analysis was performed to examine use during selected time periods in 2019. RESULTS Total femoropopliteal procedures in 2017, 2018, and 2019 were 17458, 21140, and 21322, respectively. Paclitaxel devices were used for 8852 arteries in 2017, 10691 in 2018, and 6732 in 2019, which was significantly reduced, when comparing 2019 volumes to 2017 or 2018 (p < .0001) and 2019 versus 2018 + 2017 volumes (p < .0001). Post-hoc analysis of selected times in 2019 demonstrated variable use throughout 2019. CONCLUSIONS Following publication of data with concerns of mortality associated with paclitaxel device use in 2018, a rapid reduction in overall paclitaxel device use was observed in 2019.
Collapse
Affiliation(s)
- Blair Edward Warren
- Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada.
| | - Graham Roche-Nagle
- Division of Vascular Surgery, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada
| | - Jiachen Zhu
- Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada
| | - Guan Wang
- Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada
| | - Naomi Eisenberg
- Division of Vascular Surgery, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada
| | - Dheeraj K Rajan
- Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada; Joint Department of Medical Imaging, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada
| | - Sebastian Mafeld
- Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada; Joint Department of Medical Imaging, Toronto General Hospital, University Health Network, University of Toronto, Toronto, Ontario, Canada
| |
Collapse
|
4
|
Atiq-Ur-Rehma N, Mackintosh JB, Warren BE, Lindsay DR. Revegetated Saline Pastures as a Forage Reserve for Sheep: 1. Effects of Season and Grazing on Morphology and Nutritive Value of Saltbush. Rangel J 1999. [DOI: 10.1071/rj9990003] [Citation(s) in RCA: 10] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [What about the content of this article? (0)] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/23/2022]
Abstract
This study determined the impact of grazing on the kind of material selected by sheep from a saltbush (Atriplex amnicola) plantation and the changes in the chemical composition of saltbush plants associated with season.
Saltbush plants in four plots of about 0.9 hectares each were scored from 1 to 5 on the basis of leafiness. Ten Merino wethers were grazed on each plot, giving a stocking rate of about 11 per hectare.
The results demonstrated that sheep selected only stem material of less than 1.5 mm diameter, while the grazing pressure on saltbush plants, as described by the number of stems eaten per 0.1 m2, increased from less than one to an average of 18 to 20 stems in 5 to 7 weeks.
Grazing had a significant effect on dry matter digestibility (DMD) and nitrogen concentration of whole plant samples. During six weeks of grazing the DMD of whole plant samples cut 10 cm from the tips of the branches dropped from 0.53 to 0.25 (P<0.05), whereas the nitrogen content declined from 11 g/kg to 8 g/kg (P<0.05). At the end of grazing the nutritional value of whole plant samples was very poor and sheep refused to eat stem that was thicker than 1.5 mm. These findings question the grazable fraction reported in the literature for saltbush plantation. when the stem diameters used in calculations are not reported or thicker stems were assumed grazable.
Season also had a significant effect on the ratio of leaf to stem and the mineral content. The concentration of sodium in saltbush leaf was negatively correlated (r = -0.93) with both nitrogen and potassium.
Collapse
|
5
|
Abstract
It was shown by Paterson that in a cold�worked face-centred cubic metal, deformation faults produce a peak shift, and twin faults produce a peak asymmetry. If the peak shape is represented by a Fourier series, the asymmetry is expressed by the sine terms. However, the evaluation of the sine coefficients is difficult, due to overlapping of the tails of neighbouring refleotions and uncertainty in the peak origin. In this paper, a method is developed which combines the tails of the (Ill) and (200) reflections. This allows a determination of the twin fault probability in face-centred cubic metals which is more nearly independent of overlapping of the tails and choice of the peak origin.
Collapse
|
6
|
Abstract
The small angle intensity from a cold-worked polygrained metal is treated in terms of double Bragg reflections. It is assumed that cold work breaks up the original grains into subgrains with varying orientations, such that a first reflection in one subgrain can be followed by a second reflection in another subgrain of the original grain. For an initial sample which is polygrained with random orientation, the small angle intensity resulting from cold work is given in absolute units in terms of two parameters; an average initial grain size G, and a subgrain correlation function p( ep) giving the distribution in orientation of subgrains with respect to any average subgrain of the original grain. Under favourable conditions, both of these quantities can be obtained from the small angle intensity curve. For polygrained metals, small angle scattering is a useful tool -for obtaining the subgrain correlation function peep), a quantity not obtainable from the usual high angle measurements.
Collapse
|