- Email: [email protected]
[I256] A review of medical laser accidents: A simple burn to death by laser
Abstracts / Physica Medica 52 (2018) 1–98
paediatric DRLs, mostly for a limited set of examinations. The existing paediatric DRLs are often adopted from old European Commission recommendations or from rarely updated surveys performed by other countries. Furthermore, the lack of a standardized methodology for establishing DRLs in children prevents the comparison among those already available. Recently, an EC project on new European Guidelines for DRLs in Paediatric Imaging (PiDRL) was awarded to a consortium lead by ESR with the involvement of ESPR, EFRS, EFOMP, and STUK. This project resulted in a series of recommendations concerning the methods to establish new DRLs in children. These recommendations include the definition of local, national, and European DRLs, the set of examinations for which DRLs should be established, grouping of children, dosimetric parameters, and instructions on how to make good use of DRLs. The PiDRL guidelines also provide a set of European DRLs (EDRLs), based on the median value of the distribution of national DRLs for a defined clinical imaging task surveyed for standardised patient groupings. These EDRLs provide an interim solution for countries without national DRLs (NDRLs), until such NDRLs become available. Future steps to be taken are the establishment and use of new EDRLs according to the PiDRL recommendations. The involvement and joint efforts of authoritative bodies, scientific societies, and stakeholders are pivotal to this aim. http://www.eurosafeimaging.org/wp/wp-content/uploads/2014/ 02/European-Guidelines-on-DRLs-for-Paediatric-Imaging_Revised_ 18-July-2016_clean.pdf. https://doi.org/10.1016/j.ejmp.2018.06.324
[I253] Basics of deep learning Jonas Teuwen * Radboud University Medical Center, Radiology and Nuclear Medicine, Nijmegen, Netherlands ⇑ Corresponding author. Deep learning has attracted much interest from the medical community due to its successful application to medical imaging problems which were thought to be purely within the human realm. The growing amount of well-curated medical image data and the increasing availability of affordable resources has allowed to apply deep learning to many different imaging modalities. In this talk, we will cover what deep learning is, its different flavors and some recent state-of-the-art applications to medical imaging. We will first look at Convolutional Neural Networks, the main workhorse in medical imaging and look at relevant applications in mammography, chest CT and digital pathology. Finally we will show some applications of deep generative models to the reconstruction of images.
95
Along with the technical imaging modality development, the optimisation process has transformed into more demanding and multiprofessional challenge where the image quality metrics should evolve accordingly from technical towards clinical presentation. In the parameter level, this development may include clinical task function and observer related parameters supplementing the traditional MTF and NPS parameters. New methods enable model observer based detectability and diagnostic accuracy estimates. Ultimately, we should aim beyond the concept of technical quality, to extend our methods and knowledge towards measuring and optimising the diagnostic value in terms of care outcome. Modern radiological imaging technology, reconstruction and post-processing techniques provide new and mostly non-linear image output. In part, this explains the need for more complicated analysis of image quality compared to the traditional and more linear image output. Improvement in radiological optimisation requires also patient-specific and indication-specific adjustment of imaging parameters and analysis methods. One size and purpose simply does not fit all patients and applications. Both of these aspects – nonlinearity and patient/indication specificity – aim to improve diagnostic information content and representation of indication-specific image features in radiology. Improved optimisation process and more consistent imaging quality (evaluated by target value, its uncertainty and precision) require objective and quantitative connections from diagnostic and technical parameters to clinical outcome parameters. Comprehensive methodology to enable this approach involves combining several types of data together, as described in a recent publication from an international summit [1]. Artificial intelligence (AI) based deep learning methods – including data quality control and validation – are prerequisites for this kind of data analysis, due to inherent non-linearity of the problem and large amount of heterogeneous data which is not equitable by traditional methods. Our medical physicist professional role should follow this development and incorporate AI & deep learning topics accordingly into our educational programs.
Reference [1] Samei E, Jarvinen H, Kortesniemi M, Simantirakis G, Goh C, Wallace A, Vano E, Bejan A, Rehani MM, Vassileva J. Medical imaging dose optimization from ground up: expert opinion of an international summit. J Radiol Prot 2018 May 17. https://doi.org/ 10.1088/1361-6498/aac575. [Epub ahead of print]. https://doi.org/10.1016/j.ejmp.2018.06.326
[I255]
https://doi.org/10.1016/j.ejmp.2018.06.325
Abstract not available. [I254] From image quality to care outcome Mika Kortesniemi * Hus Medical Imaging Center, Radiology, Helsinki, Finland Corresponding author.
[I256] A review of medical laser accidents: A simple burn to death by laser Ayakkanu Manivannan *
⇑
Medical physicists have a long tradition of measuring image quality with objective metrics including contrast, noise and resolution, and their frequency-based derivatives. These methods have supported our main tasks related to quality assurance and optimisation.
NHS Ayrshire & Arran Crosshouse Hospital, Medical Physics, Kilmarnock, United Kingdom ⇑ Corresponding author. After the invention of the laser by Maiman in 1960, Ophthalmology took advantage of the laser immediately and reported the first
96
Abstracts / Physica Medica 52 (2018) 1–98
use of a medical laser in 1963 to replace the filament light source and improved the accuracy of photocoagulation for the treatment of diabetic retinopathy. In the past 60 years, a variety of lasers are used in many medical specialties such as ophthalmology, gynaecology, urology ENT and maxilo-facial surgery. Millions of safe laser procedures are carried out annually throughout the world. Despite international guidelines, regulations and local safety measures and procedures, adverse events (also known as accidents) do happen and result in undesirable consequences. In recent years, due to medical lasers becoming inexpensive, there is a massive increase in their use for cosmetic applications such as skin rejuvenation, hair removal and tattoo removal. In many countries, use of cosmetic lasers is unregulated and is carried out mostly by non-medical personnel. The main cause of laser accidents is due to human error, particularly not following safety procedures. Both patients and laser users become the victims of tissue damage due to unwanted laser exposure. Most laser accidents in literature report damage to skin and eye. As lasers do not penetrate much into the skin, most injuries occur as burns leading to swelling and pain followed by scars. As the eye is transparent to visible and some invisible light (UV and near infrared), lasers can cause irreversible and permanent damage in the eye and lead to blindness. Simple laser burns are usually unreported due to the fear of being blamed for the mistake. Litigation law suits against cosmetic laser treatments such as laser hair removal and tattoo removal are on the rise. Furthermore, few laser accidents report fire due to combustible material in their path. High voltage employed in some lasers such as CO2 and copper vapour lasers can result in electric shock. Even though rare, there are few reported cases of laser accidents resulting in death. This presentation is a review of all reported and unreported events. https://doi.org/10.1016/j.ejmp.2018.06.328
[OA257] Influence of nonionizing millimeter electromagnetic radiation on tumor and healthy DNA Vitali Kalantaryan a,*, Radik Martirosyan a, Yura Babayan b, Pogos Vardevanyan c a
Yerevan State University, Microwave Radiophysics, Yerevan, Armenia Yerevan State University of Architecture and Construction, Physics and Electrotecnics, Yerevan, Armenia c Yerevan State University, Biophysics, Yerevan, Armenia ⇑ Corresponding author. b
Purpose. The present study was undertaken to investigate whether millimeter range electromagnetic nonionizing radiation (NIR) of selective frequencies can suppress tumor cells growth in vivowithout cytostatic agents. Methods. The course of influence of NIR started 3 days before transplantation in order to raise activity of the animals immune system. On the fourth day animals were injected by sarcoma-37 and daily exposure was continued during 15 days. For study of the effect of irradiation on the secondary structure of DNA, in the experiments DNA isolated from the liver of healthy mice (hDNA) as well as from the tumor sarcoma 37 (tDNA) was used. Results. After 15 sessions of exposure without cytostatics, at animals of the irradiated 0,5 h was observed an inhibition of tumor growth by 33.5% compared with a control group and a sharp suppression of the level of DNA methylation in 2.1 times. The tDNA has the high level of methylation (4.7 mol%), which after 0.5 h daily exposure becomes (2.2 mol%) close to the corresponding value for hDNA (1.9 mol%). Differential melting curves (DMC) of tDNA are shifted relatively DMC of the hDNA to lower temperatures, and in the DMC of tDNA the additional peaks are appeared. The obtained
results are correlated with the spectrophotometric data. Under the influence of millimeter NIR the values of temperature and interval of melting of tDNA are changed and approach to the corresponding values of hDNA. Conclusions. Hypermethylation of tumor DNA may cause a selective sensitivity of malignant cells toward the influence of the NIR which allows an expressed antitumor effect. It is possible, that 30 min daily exposure leads to the activation of specific molecular mechanisms of cells, resulting in a decrease of undesirable structural changes in the tumor DNA and inhibition of tumor growth. Presented preliminary results have demonstrated the promising application of millimeter NIR for clinical oncology in the treatment of malignancies without damaging other tissues and antitumoral drugs and without harmful ionizing radiotheraphy. https://doi.org/10.1016/j.ejmp.2018.06.329
[OA258] Handheld MED testers: An inter-comparison to inform requirements for acceptance testing and calibration Michael Manley a,*, Jackie McCavana b a
Breastcheck, Medical Physics, Cork, Ireland St. Vincent’s University Hospital, Medical Physics and Clinical Engineering, Dublin, Ireland ⇑ Corresponding author. b
Purpose. Hand held Minimal Erythema Dose (MED) testers are designed to measure the MED for a patient receiving UVB phototherapy. The devices have a series of apertures with a superimposed foil grid of varying attenuation so that 10 incremental doses of UVB are delivered to a patient simultaneously. Introducing these devices into clinical service has presented challenges from a validation and output measurement perspective because of the device’s aperture size, the smaller bulb size, exposure geometry (device is in contact with the patient’s skin) and warm up characteristics of the bulb. In this study an inter-comparison is ‘performed on a number of units/models, to evaluate their performance characteristics and determine the minimum testing required before introducing such units into clinical service. Methods. A piece of EBT3 Gafchromic film calibrated to UVB dose was placed on device exposure area and was used to measure both the absolute and relative output of the apertures of each device. A four minute exposure was acquired after the warm up time specified by the manufacturer. This measurement was repeated for a 20 min exposure from switch on with no warm-up. The output characteristics were compared to manufacturer’s specification and an inter comparison was carried out between same model devices. Warm up and spectral characteristics of each device were also measured using a spectral radiometer. Results. Outputs measured were within 13% of that quoted by the manufacturer at the specified warm up time. Results show that the measured relative outputs agreed with that of the manufacturer within 15% and range of relative outputs for a number of units of the same model spanned over a range of up to 30%. Conclusions. Minimum acceptance testing of such units should include evaluation of warm-up characteristics; and simultaneous output measurements of all apertures over patient exposure time interval. This will allow a table of planned aperture exposures to be confirmed/determined. Periodic output measurement can then be carried out with EBT3 Gafchromic film or radiometer that has a detector with a suitable sized input optic. https://doi.org/10.1016/j.ejmp.2018.06.330
paediatric DRLs, mostly for a limited set of examinations. The existing paediatric DRLs are often adopted from old European Commission recommendations or from rarely updated surveys performed by other countries. Furthermore, the lack of a standardized methodology for establishing DRLs in children prevents the comparison among those already available. Recently, an EC project on new European Guidelines for DRLs in Paediatric Imaging (PiDRL) was awarded to a consortium lead by ESR with the involvement of ESPR, EFRS, EFOMP, and STUK. This project resulted in a series of recommendations concerning the methods to establish new DRLs in children. These recommendations include the definition of local, national, and European DRLs, the set of examinations for which DRLs should be established, grouping of children, dosimetric parameters, and instructions on how to make good use of DRLs. The PiDRL guidelines also provide a set of European DRLs (EDRLs), based on the median value of the distribution of national DRLs for a defined clinical imaging task surveyed for standardised patient groupings. These EDRLs provide an interim solution for countries without national DRLs (NDRLs), until such NDRLs become available. Future steps to be taken are the establishment and use of new EDRLs according to the PiDRL recommendations. The involvement and joint efforts of authoritative bodies, scientific societies, and stakeholders are pivotal to this aim. http://www.eurosafeimaging.org/wp/wp-content/uploads/2014/ 02/European-Guidelines-on-DRLs-for-Paediatric-Imaging_Revised_ 18-July-2016_clean.pdf. https://doi.org/10.1016/j.ejmp.2018.06.324
[I253] Basics of deep learning Jonas Teuwen * Radboud University Medical Center, Radiology and Nuclear Medicine, Nijmegen, Netherlands ⇑ Corresponding author. Deep learning has attracted much interest from the medical community due to its successful application to medical imaging problems which were thought to be purely within the human realm. The growing amount of well-curated medical image data and the increasing availability of affordable resources has allowed to apply deep learning to many different imaging modalities. In this talk, we will cover what deep learning is, its different flavors and some recent state-of-the-art applications to medical imaging. We will first look at Convolutional Neural Networks, the main workhorse in medical imaging and look at relevant applications in mammography, chest CT and digital pathology. Finally we will show some applications of deep generative models to the reconstruction of images.
95
Along with the technical imaging modality development, the optimisation process has transformed into more demanding and multiprofessional challenge where the image quality metrics should evolve accordingly from technical towards clinical presentation. In the parameter level, this development may include clinical task function and observer related parameters supplementing the traditional MTF and NPS parameters. New methods enable model observer based detectability and diagnostic accuracy estimates. Ultimately, we should aim beyond the concept of technical quality, to extend our methods and knowledge towards measuring and optimising the diagnostic value in terms of care outcome. Modern radiological imaging technology, reconstruction and post-processing techniques provide new and mostly non-linear image output. In part, this explains the need for more complicated analysis of image quality compared to the traditional and more linear image output. Improvement in radiological optimisation requires also patient-specific and indication-specific adjustment of imaging parameters and analysis methods. One size and purpose simply does not fit all patients and applications. Both of these aspects – nonlinearity and patient/indication specificity – aim to improve diagnostic information content and representation of indication-specific image features in radiology. Improved optimisation process and more consistent imaging quality (evaluated by target value, its uncertainty and precision) require objective and quantitative connections from diagnostic and technical parameters to clinical outcome parameters. Comprehensive methodology to enable this approach involves combining several types of data together, as described in a recent publication from an international summit [1]. Artificial intelligence (AI) based deep learning methods – including data quality control and validation – are prerequisites for this kind of data analysis, due to inherent non-linearity of the problem and large amount of heterogeneous data which is not equitable by traditional methods. Our medical physicist professional role should follow this development and incorporate AI & deep learning topics accordingly into our educational programs.
Reference [1] Samei E, Jarvinen H, Kortesniemi M, Simantirakis G, Goh C, Wallace A, Vano E, Bejan A, Rehani MM, Vassileva J. Medical imaging dose optimization from ground up: expert opinion of an international summit. J Radiol Prot 2018 May 17. https://doi.org/ 10.1088/1361-6498/aac575. [Epub ahead of print]. https://doi.org/10.1016/j.ejmp.2018.06.326
[I255]
https://doi.org/10.1016/j.ejmp.2018.06.325
Abstract not available. [I254] From image quality to care outcome Mika Kortesniemi * Hus Medical Imaging Center, Radiology, Helsinki, Finland Corresponding author.
[I256] A review of medical laser accidents: A simple burn to death by laser Ayakkanu Manivannan *
⇑
Medical physicists have a long tradition of measuring image quality with objective metrics including contrast, noise and resolution, and their frequency-based derivatives. These methods have supported our main tasks related to quality assurance and optimisation.
NHS Ayrshire & Arran Crosshouse Hospital, Medical Physics, Kilmarnock, United Kingdom ⇑ Corresponding author. After the invention of the laser by Maiman in 1960, Ophthalmology took advantage of the laser immediately and reported the first
96
Abstracts / Physica Medica 52 (2018) 1–98
use of a medical laser in 1963 to replace the filament light source and improved the accuracy of photocoagulation for the treatment of diabetic retinopathy. In the past 60 years, a variety of lasers are used in many medical specialties such as ophthalmology, gynaecology, urology ENT and maxilo-facial surgery. Millions of safe laser procedures are carried out annually throughout the world. Despite international guidelines, regulations and local safety measures and procedures, adverse events (also known as accidents) do happen and result in undesirable consequences. In recent years, due to medical lasers becoming inexpensive, there is a massive increase in their use for cosmetic applications such as skin rejuvenation, hair removal and tattoo removal. In many countries, use of cosmetic lasers is unregulated and is carried out mostly by non-medical personnel. The main cause of laser accidents is due to human error, particularly not following safety procedures. Both patients and laser users become the victims of tissue damage due to unwanted laser exposure. Most laser accidents in literature report damage to skin and eye. As lasers do not penetrate much into the skin, most injuries occur as burns leading to swelling and pain followed by scars. As the eye is transparent to visible and some invisible light (UV and near infrared), lasers can cause irreversible and permanent damage in the eye and lead to blindness. Simple laser burns are usually unreported due to the fear of being blamed for the mistake. Litigation law suits against cosmetic laser treatments such as laser hair removal and tattoo removal are on the rise. Furthermore, few laser accidents report fire due to combustible material in their path. High voltage employed in some lasers such as CO2 and copper vapour lasers can result in electric shock. Even though rare, there are few reported cases of laser accidents resulting in death. This presentation is a review of all reported and unreported events. https://doi.org/10.1016/j.ejmp.2018.06.328
[OA257] Influence of nonionizing millimeter electromagnetic radiation on tumor and healthy DNA Vitali Kalantaryan a,*, Radik Martirosyan a, Yura Babayan b, Pogos Vardevanyan c a
Yerevan State University, Microwave Radiophysics, Yerevan, Armenia Yerevan State University of Architecture and Construction, Physics and Electrotecnics, Yerevan, Armenia c Yerevan State University, Biophysics, Yerevan, Armenia ⇑ Corresponding author. b
Purpose. The present study was undertaken to investigate whether millimeter range electromagnetic nonionizing radiation (NIR) of selective frequencies can suppress tumor cells growth in vivowithout cytostatic agents. Methods. The course of influence of NIR started 3 days before transplantation in order to raise activity of the animals immune system. On the fourth day animals were injected by sarcoma-37 and daily exposure was continued during 15 days. For study of the effect of irradiation on the secondary structure of DNA, in the experiments DNA isolated from the liver of healthy mice (hDNA) as well as from the tumor sarcoma 37 (tDNA) was used. Results. After 15 sessions of exposure without cytostatics, at animals of the irradiated 0,5 h was observed an inhibition of tumor growth by 33.5% compared with a control group and a sharp suppression of the level of DNA methylation in 2.1 times. The tDNA has the high level of methylation (4.7 mol%), which after 0.5 h daily exposure becomes (2.2 mol%) close to the corresponding value for hDNA (1.9 mol%). Differential melting curves (DMC) of tDNA are shifted relatively DMC of the hDNA to lower temperatures, and in the DMC of tDNA the additional peaks are appeared. The obtained
results are correlated with the spectrophotometric data. Under the influence of millimeter NIR the values of temperature and interval of melting of tDNA are changed and approach to the corresponding values of hDNA. Conclusions. Hypermethylation of tumor DNA may cause a selective sensitivity of malignant cells toward the influence of the NIR which allows an expressed antitumor effect. It is possible, that 30 min daily exposure leads to the activation of specific molecular mechanisms of cells, resulting in a decrease of undesirable structural changes in the tumor DNA and inhibition of tumor growth. Presented preliminary results have demonstrated the promising application of millimeter NIR for clinical oncology in the treatment of malignancies without damaging other tissues and antitumoral drugs and without harmful ionizing radiotheraphy. https://doi.org/10.1016/j.ejmp.2018.06.329
[OA258] Handheld MED testers: An inter-comparison to inform requirements for acceptance testing and calibration Michael Manley a,*, Jackie McCavana b a
Breastcheck, Medical Physics, Cork, Ireland St. Vincent’s University Hospital, Medical Physics and Clinical Engineering, Dublin, Ireland ⇑ Corresponding author. b
Purpose. Hand held Minimal Erythema Dose (MED) testers are designed to measure the MED for a patient receiving UVB phototherapy. The devices have a series of apertures with a superimposed foil grid of varying attenuation so that 10 incremental doses of UVB are delivered to a patient simultaneously. Introducing these devices into clinical service has presented challenges from a validation and output measurement perspective because of the device’s aperture size, the smaller bulb size, exposure geometry (device is in contact with the patient’s skin) and warm up characteristics of the bulb. In this study an inter-comparison is ‘performed on a number of units/models, to evaluate their performance characteristics and determine the minimum testing required before introducing such units into clinical service. Methods. A piece of EBT3 Gafchromic film calibrated to UVB dose was placed on device exposure area and was used to measure both the absolute and relative output of the apertures of each device. A four minute exposure was acquired after the warm up time specified by the manufacturer. This measurement was repeated for a 20 min exposure from switch on with no warm-up. The output characteristics were compared to manufacturer’s specification and an inter comparison was carried out between same model devices. Warm up and spectral characteristics of each device were also measured using a spectral radiometer. Results. Outputs measured were within 13% of that quoted by the manufacturer at the specified warm up time. Results show that the measured relative outputs agreed with that of the manufacturer within 15% and range of relative outputs for a number of units of the same model spanned over a range of up to 30%. Conclusions. Minimum acceptance testing of such units should include evaluation of warm-up characteristics; and simultaneous output measurements of all apertures over patient exposure time interval. This will allow a table of planned aperture exposures to be confirmed/determined. Periodic output measurement can then be carried out with EBT3 Gafchromic film or radiometer that has a detector with a suitable sized input optic. https://doi.org/10.1016/j.ejmp.2018.06.330