Page 1 de 17 Forensic investigations in the case of Mr Louvet: Radioactivity determination in Mr Louvet belongings To : Pr Mangin (CURML/CHUV) PD Dr Pascal Froidevaux Authors : Dr S?bastien Baechler Pr Fran?ois Bochud Version of : 11.05.2012 1. Introduction The Institute of Radiation Physics (thereafter IRA) received the personal belongings of Mr Louvet by the University Centre of Legal Medicine (CURML) for investigating a potential radioactivity contamination with a particular attention to polonium-210 (210Po). Because more than 20 half-lives have passed between a potential contamination by 210Po and the present measurements, it is necessary to apply low level techniques to the determination of unsupported grandmother 210Pb). 210 Po (210Po not supported by its Materials and Methods Surface contamination measurements Surface contamination measurements were performed using a Berthold LB-122 monitor in ?/? mode and a Como monitor in ? mode. All the set of personal belongings was scanned manually for radioactive contamination. Gamma spectrometry measurements For gamma spectrometry measurements, personal belongings were pooled into 11 samples as shown in Figure 1. Each sample was then measured on an HPGe gamma spectrometer (measuring time > 24h). Artificial radionuclides, such as 60Co, 134Cs and 137Cs were investigated, as well as 210Pb. rapport_mesure_po_2012_05_11.docx Page 2 de 17 Figure 1. Pictures of Mr Louvet belongings pooled in 11 samples for gamma spectrometry measurements. Page 3 de 17 Alpha spectrometry measurements Sampling: we have proceeded to the 210 Po determination based on the fact that some of them might have been worn shortly before his death. Thus the first samples are: a sample taken from the collar of a sportswear with strong dirt stains, a sample from the front tight of the same sportswear, a sample from the front tight of a woolen pant, 2 pills each of a mixture of different medicines (about 38 g) and two samples taken from an underwear that appeared obviously worn. One of the two samples has been taken from a stained part of the front of the underwear, identified possibly as a urine stain. Following the results of the 210Po determination (see results) we have proceeded to the sampling of 6 aliquots from the remaining of this underwear. These samples, as the two before, contain about 3-4 g of cotton wool. The sportswear is a mixture between cotton wool and synthetic fibers. We have then taken a sample from a hospital cap. This sample has been taken by cutting around a blood stain and weighted 0.7041 g. This 3rd samples batch contained also a sample from another underwear (probably not worn), a second sample from the collar of the sportswear, a sample made of the bristles of two toothbrushes (one manifestly used) and a reactants blank (H2SO4: 10 ml; HNO3: 25 ml; NH4OH: 40 ml, Fe:20 mg). Page 4 de 17 The forth batch of samples contained two aliquots of a third underwear, an aliquot from a kefieh taken around a stain (probably blood stain, 2.2540 g) and a sample from a "Russian" chapka (internal band in contact with the head) known to have been worn by Mr Louvet shortly before his death. The 5th batch of samples contained 3 underwears from 3 different IRA collaborators. The 6th batch of samples contained an aliquot from a sock possibly worn, a large piece from a long johns (112.7529 g), a reactants blank and two new underwears bought directly from a shop (Bon Genie, Lausanne) of the brand Zimmerli (the same as the one of Mr Louvet's underwear) and one underwear of the brand Hanro. The 7th batch of samples contained an aliquot (far from the blood stain) of the hospital cap, a supplementary aliquot of the Russian chapka, a sample formed by the bristles of the toothbrush of an IRA collaborator and three aliquots of long johns (taken along the leg). The 8th batch contained a sample formed by 4 pieces of cotton wool taken from a child's drawing sampled around stains (possibly saliva, vomit and blood, total 1.4631 g), a sample taken from the interior and from the superior band (stained) of an old slipper, a sample from a new sock (not worn, still attached by thread) and a sample of reactants blank. The 9th batch contained two more samples from the collar of the sportswear, an aliquot of the sportswear taken from the back, around a stain (possibly blood, 0.7565 g) and three more aliquots from the interior band in contact with the skull of the Russian chapka. The 10th and last batch contained smear of personal belongings such as glasses, a blank smear, a samples formed by the bristles of a toothbrush of an IRA collaborator, a sample of the hospital cap and an aliquot of the child drawing (out of the stains) and a reactants blank. All the sampling has been carried out with the aim to find contaminated samples, thus the choice of "stained" samples. 210 Nevertheless cotton wool may contain a certain amount (not known) of Pb/210Po because of exposition of cotton wool to atmospheric deposition during cultivation because cotton wool balls have a large adsorption surface. We therefore decided to measure some blank cotton (not known to have been exposed to unsupported 210Po). 210 Po determination: the method is presented in detail in supplementary material. Briefly, the aliquots 209 are traced with 50 mBq of Po and digested by conc. H2SO4 at about 70?C to provoke the dehydration of the cotton wool (charred sample), then the charred sample is oxidized cautiously with portions of conc. HNO3 (strong emission of NOx vapors). When the NOx production has reduced, the oxidation is pursued during two more hours under a glass watch on a hot plate. After this treatment aiming to remove a large part of organics by wet ashing, the complete removing of the organic matter is obtained through a digestion in a pressurized microwave digester (MLS Ultraclave IV). A clear solution is obtained. If synthetic fibers are present, the solution must be filtered because they are not soluble. 210 Po is co-precipitated from the acidic solution along with iron hydroxide. After centrifugation, the precipitate is dissolved in 80 ml 1 M HCl , 500 mg of ascorbic acid is added and Page 5 de 17 polonium is spontaneously electrodeposited on a silver disc during 4 hours at 50?C (or overnight at room temperature). Determination of 210 210 Po supported by Pb in the most active samples: the method is presented in 210 detail in supplementary material. Briefly, after Po electrodeposition the solution is evaporated to dryness, HNO3 portions are added to destroy all remaining ascorbic acid. The residue is dissolved in 10 ml 9 M HCl and this solution is passed through an anionic chromatography column (2 g of AG1x8) to extract [PoCl6]2-. re-growth of 210 210 Pb will pass in the elution solution which is evaporated to dryness. 209 The residue is dissolved in 80 ml 1M HCl and let in the refrigerator during at least 3 months (36.6% Po from 210 Pb) afterwards, 50 mBq of Po is added and polonium is electrodeposited on a silver disc as before. Biokinetic model of 210Po and urinary excretion The systemic biokinetic model of polonium proposed by Leggett and Eckerman [1] (see Figure 2) has been implemented in the simulation modeling tool Ecolego in order to calculate the typical retention of 210 Po in organs and tissues in case of poisoning. In the latter situation, ingestion is the most 210 probable route of intake. Therefore, the systemic biokinetic model was coupled to the human alimentary tract model of ICRP 100. Absorption of Po is assumed to occur exclusively from the small intestine and is characterized by the fractional intestinal absorption f1=0.1 (inorganic form) or f1=0.5 (organic form). The f1 value is defined as the fraction of the activity leaving the stomach that is subsequently transferred to blood by absorption from the small intestine. Using this model with f1=0.1, the daily urinary excretion after acute ingestion of 1 Bq of activity estimates of ingested 210 210 Po was determined. Based on Po in case of the poisoning of Mr. Litvinenko, the level of activity that we might find in urine and belongings, especially underwear, in case of Mr Louvet poisoning was estimated. Page 6 de 17 Ingestion Figure 2. Compartments of the systemic model of polonium and connection with the model of the gastrointestinal tract. Results and discussion Surface contamination measurements Surface contamination measurements did not show the presence of radioactive contamination in the personal belongings. Gamma spectrometry measurements Results of gamma spectrometry measurements are given in Table 1. None of the samples did show a measurable activity. We therefore give the detection limits of the measuring facilities for 60Co, 134Cs, 137 Cs and 210Pb. Table 1. Results of activity measurements of Mr Louvet belongings using gamma spectrometry. The symbol "<" means lower than the detection limit ot the measuring facility. # 1 2 3 4 5 7 8 9 10 11 12 Description of the sample Kefieh Sportswear Hat Clothing, including underwear Sports bag (full) Clothing, including underwear Russian chapka (3) and cap Drugs - part 1 Drugs - part 2 Slippers and various objects Various objets (glasses) Co-60 < 0.6 < 0.2 < 0.2 < 0.3 < 2 < 0.7 < 0.5 < 0.4 < 0.6 < 0.7 < 0.3 Activity in Bq Cs-134 Cs-137 < 0.6 < 0.2 < 0.2 < 0.3 < 2 < 0.7 < 0.5 < 0.4 < 0.7 < 0.7 < 0.3 < 0.6 < 0.2 < 0.2 < 0.3 < 2 < 0.7 < 0.5 < 0.4 < 0.7 < 0.7 < 0.4 Pb-210 < 19 < 3 < 3 < 4 < 30 < 20 < 3 < 5 < 10 < 10 < 5 Page 7 de 17 Since 210 Po emits a gamma ray of 803 keV with a low intensity of 0.00107%, its activity can be measured in urine using gamma spectrometry if the level of intake is high and the measurements are performed early enough after intake. Therefore, the results of both measurements of urine performed on 8.11.2004 by the "Laboratoire de contr?le radiotoxicologique des Arm?es" were reassessed for the presence of 210 Po. This complementary analysis of both spectra acquired during 15 hours did not 210 reveal the presence Po. Based on the detection limit of 54 Mn (? of 834.8 keV) provided in the report, we estimated the detection limit of 210Po for this facility around 25 kBq/l. Alpha spectrometry measurements The results are presented in the Table 2. After the measure of the first batch, it appeared that an aliquot of the underwear with urine stain was particularly high in 210 Po (49 mBq/g). We then measured 6 others aliquots and another sample presented an enhanced 210Po activity (9.7 mBq/g). All the others aliquots of the underwear have an activity that may be considered as background activity of 0-4-1.0 mBq/g cotton wool. Page 8 de 17 Table 2. Activity of 210 Po (mBq/g) for aliquots of different samples of Mr Louvet belongings and reference 210 samples (not contaminated) taken from IRA collaborators, classified by order of activity. Description Bristles of two toothbrushes Underwear (A) urine stain Hospital cap, aliquot 3 Hospital cap with blood stain Russian chapka-1 Underwear (A) ,aliquot 3 Sportswear, back, blood stain Hospital cap, aliquot 2 Russian chapka-4 Old slipper, interior Child's drawing, stains Sportswear, collar, aliquot 1 Sock (worn) Russian chapka-2 Kefieh, blood stain Sportswear, collar, aliquot 4 Sportswear, collar, aliquot 5 Smear of personal belongings Russian chapka-3 underwear C, aliquot 1 Child's drawing, without stains Russian chapka-5 Underwear C, aliquot 0 Underwear A, aliquot 2 Cotton wool pant Old slipper, superior band Uinderwear A, aliquot 4 Underwear B, aliquot 2 Sportswear, collar without cotton wool phase Underwear A, aliquot 3 Sportswear, front tight Underwear A, aliquot 5 Underwear A, aliquot 6 Underwear A, aliquot 1 Long johns, aliquot 3 New sock, not worn Underwear B, aliquot 1 Long johns, aliquot 1 Long johns, aliquot 2 Medicine mixture (2 pills each) Cotton and reference samples Bristles of toothbrush collaborator IRA 1 Underwear collaborator IRA 2, aliquot 1 Underwear new, brand Zimmerli A Underwear collaborator IRA 2, aliquot 2 Underwear collaborator IRA 3, aliquot 1 Underwear collaborator IRA 1, aliquot 1 Underwear collaborator IRA 1, aliquot 2 Underwear collaborator IRA 3, aliquot 2 Underwear new, brand Zimmerli A Underwear new, brand Hanro m (g) in the analysis 0.4 3.69 0.8193 0.704 3.636 6.2667 0.7565 0.2895 1.9403 0.8906 1.4631 3.3723 2.1693 0.3049 2.254 3.9710 4.1180 0.63 1.7320 2.5885 2.5781 2.7055 3.2342 5.476 2.7707 5.2198 3.3384 4.6926 2.9461 2.3731 2.4764 4.9678 4.3517 3.425 3.5286 1.6164 4.0858 2.2687 4.1529 39 0.4776 3.0715 3.3725 3.2315 3.7615 3.1571 3.5817 3.3311 3.5805 4.2094 Po (mBq/g) 54 49 20 19.2 9.7 9.7 6.8 6.0 5.8 3.9 3.5 3.2 2.2 2.1 1.6 1.4 1.2 1.2 1.1 1.1 1.0 1.0 0.9 0.9 0.8 0.8 0.7 0.7 0.6 0.6 0.6 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 0.1 21 6.7 2.1 1.5 1.0 0.9 0.9 0.5 0.5 0.3 Page 9 de 17 underwear 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 4.7 4.9 5.1 Energy (MeV) 5.3 urine stain sample-1 sample-2 sample-4 urine stain-3 sample-5 sample-6 sample-7 Figure 2. Alpha spectra of polonium normalized to the activity of the 209Po tracer (left peak) for 7 aliquots of an underwear belonging to Mr Louvet. 210Po is estimated by the surface under the right peak. Aliquot named "urine stain" clearly showed stains attributed to urine. Figure 2 shows the alpha spectra of polonium sources obtained from all the samples taken from the underwear A. We can see that two aliquots are strongly contaminated with 210 counts (a.U.) Po. The most contaminated aliquot is a sample from a part of the underwear showing a stain attributed to urine. Then several aliquots of underwear B and C and also taken from a long johns have been analyzed; none presented an activity above 1.1 mBq/g of cotton wool. This is coherent with the fact that these clothes looked as having not been worn. We suspect these samples as having been bought in an airport free shop during Mr Louvet's transport to Paris, because the Zimmerli brand of underwear is not commonly found out of luxury shops. Several aliquots of the sportswear are manifestly contaminated by 210 Po. The sportswear showed many dirt stains, especially to the collar, thus we suspect it has been worn by Mr Louvet before his death. Alpha spectra are presented on the Figure 3. We can see that 3 of the 4 aliquots from the collar have 210 Po activities much higher than the one sampled from the tight and that the aliquot of the back presenting a suspected blood stain has an activity higher that the two supposedly uncontaminated aliquots. Page 10 de 17 sportswear 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 4.7 4.9 5.1 5.3 counts (a.u.) pants collard, stained collard collard collard back, blood stain Energy (MeV) Figure 3. Alpha spectra of polonium normalized to the activity of the 209Po tracer for 5 aliquots of a sportswear belonging to Mr Louvet. In the bag of personal belonging, there was a small hospital cap presenting a blood stain. We strongly suspect that this cap has been worn during Mr Louvet hospitalization. We cut a piece of cotton around this blood stain and two others from the cap. Results show that the blood stain and one of the other samples are strongly contaminated by 210Po (Figure 4.) Page 11 de 17 hospital cap 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 4.7 4.9 5.1 Energy (MeV) 5.3 counts (a.u.) blood stain sample-1 sample-2 Figure 4. Alpha spectra of polonium normalized to the activity of the 209Po tracer for 3 aliquots of a hospital cap which is strongly suspected to have been worn before Mr Louvet's death. On a witness photograph of Mr Louvet taken shortly before his transfer to Paris, we can see that he was wearing a Russian chapka. This chapka, along with two others was found in the belongings that we received. We cut several pieces of the internal band in close contact with the skull near the temples. Two of these samples contain significantly more 210Po than the 3 others (Figure 5). After having measured more than 40 aliquots from Mr Louvet belongings, it was necessary to determine a background value for different samples of cotton wool. These measurements show that most of the samples have a 210Po activity between 0.5-1 mBq/g (n=6). Nevertheless, 3 aliquots taken from two different underwears show activities between 1.5 and 6.7 mBq/g. The value of 6.7 mBq/g of underwear collaborator IRA 2 (aliquot 1) is not explained up to now. Nevertheless we observe that the aliquots sampled from Mr Louvet belongings that have been obviously not worn do not contain significant 210Po activity (0.2-0.9 mBq/g, n=18). Thus clearly an activity close to 1.0 mBq/g can be considered as a background value for MrLouvet belongings. Page 12 de 17 russian chapka 1 0.8 0.6 0.4 0.2 0 4.5 4.7 4.9 5.1 Energy (MeV) 5.3 sample-1 sample-2 sample-3 sample-3 sample-5 Figure 5. Alpha spectra of polonium normalized to the activity of the 209Po tracer for 5 aliquots of a Russian Chapka which has been worn before Mr Louvet's death. Bristles from toothbrush present another problem. They appeared clearly contaminated when sampled from Mr Louvet belongings (54 mBq/g). Nevertheless a control measurement carried out on a toothbrush from IRA collaborator 1 also shows a significant activity (21 mBq/g) while a second control measurement on a toothbrush from IRA collaborator 3 presents only a background value (1.0 mBq/g). Thus the measurement of Mr Louvet toothbrush is not conclusive and only the further determination of supported toothbrush by unsupported 210 counts (a.u.) Po will give an answer concerning a possible contamination of the 210 Po. The measurements of reactants blank (all the products used in an analysis without the aliquot) yield activities between 0.2 and 2 mBq/sample. The highest activities (1.2 and 2.0 mBq/sample) have been observed using a glass H2SO4 bottle. When a new and plastic H2SO4 bottle is used we observe a significant decrease in reactants blank activity (< 1 mBq/sample). Nevertheless we reached the limit of the low activity measurements and some of the aliquots have lower activity than blank samples. In addition, glassware was used all along the chemical process and glass is known to contain significant 210 Pb activity. Nevertheless the presence of 210Po in reactants and glassware used in this work cannot account for the 210Po activities found in some of Mr Louvet belongings. Page 13 de 17 all measurments 60 50 40 mBq/g 30 20 10 0 1 6 11 16 21 26 31 36 41 46 Figure 6. 210Po activity of all the samples measured in this work according to increasing activity. Red bars represent samples taken from IRA collaborator 1,2 and 3 and underwear from a luxury shop. Activities up to 2 mBq/g can be considered as background activities. On figure 6 we presented the activities of all the samples measured in this work. Activities up to 2 mBq/g can be considered as background activities (here about 33 of 50 measures). We observe clearly samples with increased activities. Nevertheless 2 of the 17 potentially contaminated samples do not come from Mr Louvet belongings, so that the comparison is not fully conclusive. Only the determination of unsupported by 210Po. We thus chemically separated 210 210 Po can settle the question about a possible poisoning of Mr Louvet 210 Pb to determine supported 210 Po in 18 of the most contaminated samples and let them settle for more than 3 months for the samples along with the 210 Po re-growth in the refrigerator. The list of Po current activity is given in Table 3. We plan to measure the supported 210Po at the beginning of June 2012. Page 14 de 17 Table 3. Activity of supported 210Po (mBq/g), after chemical separation of 210Pb and a re-growth period of 210Po of 3 months. (re-growth level=36.6%). 210 210 Type Bristles of two toothbrushes Underwear A, urine stain Underwear A, aliquot 3 Hospital cap. Blood stain Sportwear, collar, aliquot 1 Russian chapka-1 Underwear, IRA collaborator 2, aliquot 1 Toothbrush, IRA collaborator 1 Child'drawing, stains Sportswear, back, stain Russian chapka -4 Hospital cap, Aliquot 2 Underwear B aliquot 1 Underwear B, aliquot 2 Sock (worn) Underwear C, aliquot 1 Old slipper, interior Kefieh, stain N? Po-12-023 Po-12-011 Po-12-015 Po-12-019 Po-12-008 Po-12-028 Po-12-030 Po-12-043 Po-12-047 Po-12-054 Po-12-056 Po-12-061 Po-12-020 Po-12-021 Po-12-032 Po-12-025 Po-12-048 Po-12-026 Po total (mBq/g) 54 49 9.7 19.2 3.2 9.7 6.7 21 3.5 6.8 5.8 19.9 0.3 0.7 2.2 1.1 3.9 1.6 Po supported (mBq/g) % supported Biokinetic model of 210Po and urinary excretion The cumulative urinary excretion and the daily urinary excretion after acute intake by ingestion of 210 Po are shown in Figure 7 and 8. A good agreement was found between the values of cumulative urinary excretion reported by Harrison et al.[2] and those calculated in our study (see Figure 6). Regarding the poisoning of Mr. Litvinenko by 210Po in November 2006 in London, Harrison et al. [2] concluded that 0.1-0.3 GBq or more absorbed to blood of an adult male is likely to be fatal within 1 month. This range would correspond to an intake of 1-3 GBq or more, assuming f1=01. Based on their study, Li et al. [3] found that the estimated amount of f1=0.5 to 1.4 GBq assuming f1=0.1. Considering a poisoning by ingestion of 1 GBq of 210 210 Po ranged from 27 MBq assuming Po (f1=0.1), it is expected to find about 500 kBq/day in urinary excretion the first 10 days after intake, about 250 kBq/day between 10 and 20 days and about 150 kBq/day between 20 and 30 days, according to Figure 8. Note that for a daily urinary excretion of about 1000 ml/day, such a concentration of 210 Po might have been observed in the gamma spectrometry analysis of 2004 (we estimated the detection limit at 25 kBq/l). Then, assuming that about 2 ml of urine might be found in the underwear with urine stains, we expect that 1 kBq can be found in the underwear if it is worn during the first 10 days after intake. This activity of 1 kBq in October 2004 corrected for radioactive decay of 210 Po (T1/2 = 138.4 d) gives about 1.4 Page 15 de 17 mBq in February 2012. For comparison, an activity of 180 mBq was found for the urine stain of underwear (A) (see Table 2). Figure 7. Cumulative urinary excretion of 210Po after a single ingestion (f1=0.1). Figure 8. Daily urinary excretion after a single ingestion of 1 Bq of 210Po (f1=0.1). Page 16 de 17 Conclusions All the belongings of Mr Louvet were first measured with surface contamination monitors and HPGe gamma spectrometry. We observed no significant activities. We then evaluated a potential contamination by a lethal quantity of polonium-210 (210Po). Because of the long time between the death (2004) and the present measurements (2012), the potential quantity of 210 Po in Mr Louvet's belonging had decreased by a factor of one million (20 radioactive half-lives). We considered the values of the literature and ran our own biokinetic and urinary excretion model. We concluded that, even in case of a poisoning similar to the Litvinenko case, only traces of the order of few mBq (one decay every 1,000 seconds) were expected to be found in year 2012. We therefore chose to conduct 210Po measurements on samples that had been manifestly worn by Mr Louvet and where stains of residual biological liquids could be suspected by direct eye observation. Some of these samples show 210 Po activities that are clearly above the values measured on other samples that had either not been worn or that were not containing visible suspect stains. This observation alone is however not sufficient to draw a final conclusion because out of the 10 measurements performed on local samples totally unrelated with Mr Louvet's belonging, two show 210 Po activities above the value of 2 mBq/g, which can be set as a limit for background value. To clarify the origin of the measured 210Po, we should take into account that this nuclide is naturally present in the environment as a decay product coming from lead-210 (210Pb): 210 Po is said to be supported by 210Pb. Therefore, we propose to wait until the beginning of June 2012 and measure the samples again. If we observe a significant amount of 210Po in the samples with high activities in the first measurements, this would show that 210Po is supported by 210Pb. This would therefore be a very strong argument in favor of a natural origin of the observed 210Po, although such a quantity would be very uncommon. Alternatively, if we do not measure a significant amount of conclude that the high activity measured the first time does not come from the therefore is not explainable by a known natural phenomenon. Finally, and in addition to our alpha-spectrometric analysis, we re-analyzed the raw data of the gamma spectrometry performed on urine samples by the French government before Mr Louvet's death. We did not detect any abnormal gamma activities. In particular, we checked the area around the energy 803 keV that is (weakly) emitted by 210Po. We did not find any evidence of this nuclide in the spectra. 210 Po, we should Pb decay and 210 Page 17 de 17 References 1. Leggett, R.W. and K.F. Eckerman, A systemic biokinetic model for polonium. Sci Total Environ, 2001. 275(1-3): p. 109-25. 2. Harrison, J., et al., Polonium-210 as a poison. J Radiol Prot, 2007. 27(1): p. 17-40. 3. Li, W.B., et al., Internal dose assessment of 210Po using biokinetic modeling and urinary excretion measurement. Radiat Environ Biophys, 2008. 47(1): p. 101-10. Privat Docent (Adjunct Professor) Dr Pascal Froidevaux Leader of the Radioecology Group Institute of Radiation Physics (IRA) Lausanne University Hospital (CHUV) Lausanne, Switzerland Dr S?bastien Baechler Leader of the Radiation Protection Group Institute of Radiation Physics (IRA) Lausanne University Hospital (CHUV) Lausanne, Switzerland Associate Professor Dr Fran?ois Bochud Director of the Institute of Radiation Physics (IRA) Lausanne University Hospital (CHUV) Lausanne, Switzerland