SleepNet: automated sleep analysis via dense convolutional neural network using physiological time series.

OBJECTIVE: In this work, a dense recurrent convolutional neural network (DRCNN) was constructed to detect sleep disorders including arousal, apnea and hypopnea using polysomnography (PSG) measurement channels provided in the 2018 PhysioNet Challenge database. APPROACH: Our model structure is composed of multiple dense convolutional units (DCU) followed by a bidirectional long-short term memory (LSTM) layer followed by a softmax output layer. The sleep events, including sleep stages, arousal regions and multiple types of apnea and hypopnea, are manually annotated by experts, which enables us to train our proposed network using a multi-task learning mechanism. Three binary cross-entropy loss functions, corresponding to sleep/wake, target arousal and apnea-hypopnea/normal detection tasks, are summed up to generate our overall network loss function that is optimized using the Adam method. Our model performance was evaluated using two metrics: the area under the precision-recall curve (AUPRC) and the area under the receiver operating characteristic curve (AUROC). To measure our model generalization, 4-fold cross-validation was also performed. For training, our model was applied to full night recording data. MAIN RESULTS: Finally, the average AUPRC and AUROC values associated with the arousal detection task were 0.505 and 0.922, respectively, on our testing dataset. An ensemble of four models trained on different data folds improved the AUPRC and AUROC to 0.543 and 0.931, respectively. SIGNIFICANCE: Our proposed algorithm achieved the first place in the official stage of the 2018 PhysioNet Challenge for detecting sleep arousals with an AUPRC of 0.54 on the blind testing dataset.

Analysis of radioactive waste generated during the cyclotron production of 99mTc.

Past and prospective shortages of medical radioisotopes have driven recent developments in the direct production of 99mTc via the 100Mo(p,2n)99mTc reaction. The cyclotron-based production method has been shown to successfully produce 99mTc, however trace impurities present in the enriched molybdenum target can also lead to the unintended creation of other radioisotopes which constitute waste. The isotopic composition of the waste has to be investigated in order to determine how it can be handled, transported and safely stored. In this article, we report which waste radioisotopes are created alongside 99mTc during target irradiation. Results are based on the gamma spectroscopy of waste produced. Significant complexities in the emission spectra made automated identification of radioisotopes inaccurate; complexities were resolved using a manual radioisotope identification procedure. The impact of target composition, integrated beam current and duration of target irradiation on the waste produced was studied. Results indicate that an average of 0.059 ± 0.003 GBq of waste is generated per 1 GBq of 99mTc produced. Two-thirds of the total waste activity produced was attributed to 99Mo (T 1/2 = 66 h) alone, while a total of fifty radioisotopes were found in the waste. Long-lived isotopes (T 1/2 > 2 months) constituted only 1% of the total waste activity at end of beam (EOB). In conclusion, it was determined that the waste generated during cyclotron-based 99mTc production was acceptably low for routine clinical production.

Two bifunctional desferrioxamine chelators for bioorthogonal labeling of biovectors with zirconium-89.

We report two bifunctional chelators, DFO-Cys and DFO-CBT, to label biovectors with zirconium-89 according to the 2-cyanobenzothiazole/1,2-aminothiol cycloaddition. Their features are high labeling yields, rapid and efficient bioconjugation, metabolically stable luciferin-based end products, and applicability to orthogonal two-step labeling of sensitive biomolecules.

Imaging study of using radiopharmaceuticals labeled with cyclotron-produced 99mTc.

Cyclotron-produced 99mTc (CPTc) has been recognized as an attractive and practical substitution of reactor/generator based 99mTc. However, the small amount of 92-98Mo in the irradiation of enriched 100Mo could lead to the production of other radioactive technetium isotopes (Tc-impurities) which cannot be chemically separated. Thus, these impurities could contribute to patient dose and affect image quality. The potential radiation dose caused by these Tc-impurities produced using different targets, irradiation conditions, and corresponding to different injection times have been investigated, leading us to create dose-based limits of these parameters for producing clinically acceptable CPTc. However, image quality has been not considered. The aim of the present work is to provide a comprehensive and quantitative analysis of image quality for CPTc. The impact of Tc-impurities in CPTc on image resolution, background noise, and contrast is investigated by performing both Monte-Carlo simulations and phantom experiments. Various targets, irradiation, and acquisition conditions are employed for investigating the image-based limits of CPTc production parameters. Additionally, the relationship between patient dose and image quality of CPTc samples is studied. Only those samples which meet both dose- and image-based limits should be accepted in future clinical studies.

Antimicrobial compounds from root, stem bark and seeds of Melia volkensii.

Three compounds, toosendanin (1), kulactone (2) and scopoletin (3), were isolated from either the root bark and/or the stem bark of Melia volkensii. Their structures were determined on the basis of spectroscopic data generated and by comparison with data from the literature. 1 and 2, isolated for the first time from M. volkensii, exhibited significant (p < 0.05) activity against Escherichia coli with minimum inhibitory concentration of 12.5 µg/mL, close to that of neomycin (6.25 µg/mL). The compounds also exhibited high activity against Aspergillus niger (MIC 6.25 µg/mL compared to 2.5 µg/mL for clotrimazole). Dichloromethane and methanol seed, hexane stem bark and methanol root bark extracts exhibited activities towards Escherichia coli, Staphylococcus aureus, Aspergillus niger and Plasmodium falciparum, respectively. Antimicrobial activity of the plant towards A. niger, P. falciparum and S. aureus is reported for the first time in the current work.

Molybdenum target specifications for cyclotron production of 99mTc based on patient dose estimates.

In response to the recognized fragility of reactor-produced (99)Mo supply, direct production of (99m)Tc via (100)Mo(p,2n)(99m)Tc reaction using medical cyclotrons has been investigated. However, due to the existence of other Molybdenum (Mo) isotopes in the target, in parallel with (99m)Tc, other technetium (Tc) radioactive isotopes (impurities) will be produced. They will be incorporated into the labeled radiopharmaceuticals and result in increased patient dose. The isotopic composition of the target and beam energy are main factors that determine production of impurities, thus also dose increases. Therefore, they both must be considered when selecting targets for clinical (99m)Tc production. Although for any given Mo target, the patient dose can be predicted based on complicated calculations of production yields for each Tc radioisotope, it would be very difficult to reverse these calculations to specify target composition based on dosimetry considerations. In this article, a relationship between patient dosimetry and Mo target composition is studied. A simple and easy algorithm for dose estimation, based solely on the knowledge of target composition and beam energy, is described. Using this algorithm, the patient dose increase due to every Mo isotope that could be present in the target is estimated. Most importantly, a technique to determine Mo target composition thresholds that would meet any given dosimetry requirement is proposed.

A fast and simple dose-calibrator-based quality control test for the radionuclidic purity of cyclotron-produced (99m)Tc.

Cyclotron production of 99mTc through the (100)Mo(p,2n)99mTc reaction channel is actively being investigated as an alternative to reactor-based (99)Mo generation by nuclear fission of (235)U. Like most radioisotope production methods, cyclotron production of 99mTc will result in creation of unwanted impurities, including Tc and non-Tc isotopes. It is important to measure the amounts of these impurities for release of cyclotron-produced 99mTc (CPTc) for clinical use. Detection of radioactive impurities will rely on measurements of their gamma (γ) emissions. Gamma spectroscopy is not suitable for this purpose because the overwhelming presence of 99mTc and the count-rate limitations of γ spectroscopy systems preclude fast and accurate measurement of small amounts of impurities. In this article we describe a simple and fast method for measuring γ emission rates from radioactive impurities in CPTc. The proposed method is similar to that used to identify (99)Mo breakthrough in generator-produced 99mTc: one dose calibrator (DC) reading of a CPTc source placed in a lead shield is followed by a second reading of the same source in air. Our experimental and theoretical analysis show that the ratio of DC readings in lead to those in air are linearly related to γ emission rates from impurities per MBq of 99mTc over a large range of clinically-relevant production conditions. We show that estimates of the γ emission rates from Tc impurities per MBq of 99mTc can be used to estimate increases in radiation dose (relative to pure 99mTc) to patients injected with CPTc-based radiopharmaceuticals. This enables establishing dosimetry-based clinical-release criteria that can be tested using commercially-available dose calibrators. We show that our approach is highly sensitive to the presence of 93gTc, 93mTc, 94gTc, 94mTc, 95mTc, 95gTc, and 96gTc, in addition to a number of non-Tc impurities.

Quantitative analysis of relationships between irradiation parameters and the reproducibility of cyclotron-produced (99m)Tc yields.

Cyclotron production of (99m)Tc through the (100)Mo(p,2n) (99m)Tc reaction channel is actively being investigated as an alternative to reactor-based (99)Mo generation by nuclear fission of (235)U. An exciting aspect of this approach is that it can be implemented using currently-existing cyclotron infrastructure to supplement, or potentially replace, conventional (99m)Tc production methods that are based on aging and increasingly unreliable nuclear reactors. Successful implementation will require consistent production of large quantities of high-radionuclidic-purity (99m)Tc. However, variations in proton beam currents and the thickness and isotopic composition of enriched (100)Mo targets, in addition to other irradiation parameters, may degrade reproducibility of both radionuclidic purity and absolute (99m)Tc yields. The purpose of this article is to present a method for quantifying relationships between random variations in production parameters, including (100)Mo target thicknesses and proton beam currents, and reproducibility of absolute (99m)Tc yields (defined as the end of bombardment (EOB) (99m)Tc activity). Using the concepts of linear error propagation and the theory of stochastic point processes, we derive a mathematical expression that quantifies the influence of variations in various irradiation parameters on yield reproducibility, quantified in terms of the coefficient of variation of the EOB (99m)Tc activity. The utility of the developed formalism is demonstrated with an example. We show that achieving less than 20% variability in (99m)Tc yields will require highly-reproducible target thicknesses and proton currents. These results are related to the service rate which is defined as the percentage of (99m)Tc production runs that meet the minimum daily requirement of one (or many) nuclear medicine departments. For example, we show that achieving service rates of 84.0%, 97.5% and 99.9% with 20% variations in target thicknesses requires producing on average 1.2, 1.5 and 1.9 times the minimum daily activity requirement. The irradiation parameters that would be required to achieve these service rates are described. We believe the developed formalism will aid in the development of quality-control criteria required to ensure consistent supply of large quantities of high-radionuclidic-purity cyclotron-produced (99m)Tc.

Graphical user interface for yield and dose estimations for cyclotron-produced technetium.

The cyclotron-based (100)Mo(p,2n)(99m)Tc reaction has been proposed as an alternative method for solving the shortage of (99m)Tc. With this production method, however, even if highly enriched molybdenum is used, various radioactive and stable isotopes will be produced simultaneously with (99m)Tc. In order to optimize reaction parameters and estimate potential patient doses from radiotracers labeled with cyclotron produced (99m)Tc, the yields for all reaction products must be estimated. Such calculations, however, are extremely complex and time consuming. Therefore, the objective of this study was to design a graphical user interface (GUI) that would automate these calculations, facilitate analysis of the experimental data, and predict dosimetry. The resulting GUI, named Cyclotron production Yields and Dosimetry (CYD), is based on Matlab®. It has three parts providing (a) reaction yield calculations, (b) predictions of gamma emissions and (c) dosimetry estimations. The paper presents the outline of the GUI, lists the parameters that must be provided by the user, discusses the details of calculations and provides examples of the results. Our initial experience shows that the proposed GUI allows the user to very efficiently calculate the yields of reaction products and analyze gamma spectroscopy data. However, it is expected that the main advantage of this GUI will be at the later clinical stage when entering reaction parameters will allow the user to predict production yields and estimate radiation doses to patients for each particular cyclotron run.

Theoretical dosimetry estimations for radioisotopes produced by proton-induced reactions on natural and enriched molybdenum targets.

This study presents a summary of the dosimetry calculations performed for three technetium agents most commonly used in nuclear medicine diagnostic studies, namely sestamibi™, phosphonates and pertechnetate, labeled with cyclotron-produced technetium. Calculated patient doses were compared to those that would be delivered by the same radiotracers labeled with technetium obtained from a generator produced in a reactor. The main difference is that technetium from a generator is pure, i.e. contains only (99m)Tc and its decay product (99g)Tc, while in a cyclotron a large number of other stable and radioactive isotopes are created. In our calculations only technetium radioisotopes (ground and isomeric states) were considered as they will be included in the radiotracer labeling process and will contribute to the patient dose. Other elements should be removed by chemical purification. These dose estimates are based on our theoretical calculations of the proton-induced reaction cross sections and radioisotope production yields. Thick targets of enriched (three different compositions) and natural molybdenum, and three initial beam energies (16, 19 and 24 MeV) were considered for irradiation times of 3, 6 and 12 h with a beam current of 200 µA. The doses were calculated for injection times corresponding to 0, 2, 8, 12 and 24 h after the end of beam.

Theoretical modeling of yields for proton-induced reactions on natural and enriched molybdenum targets.

Recent acute shortage of medical radioisotopes prompted investigations into alternative methods of production and the use of a cyclotron and ¹°°Mo(p,2n)(99m)Tc reaction has been considered. In this context, the production yields of (99m)Tc and various other radioactive and stable isotopes which will be created in the process have to be investigated, as these may affect the diagnostic outcome and radiation dosimetry in human studies. Reaction conditions (beam and target characteristics, and irradiation and cooling times) need to be optimized in order to maximize the amount of (99m)Tc and minimize impurities. Although ultimately careful experimental verification of these conditions must be performed, theoretical calculations can provide the initial guidance allowing for extensive investigations at little cost. We report the results of theoretically determined reaction yields for (99m)Tc and other radioactive isotopes created when natural and enriched molybdenum targets are irradiated by protons. The cross-section calculations were performed using a computer program EMPIRE for the proton energy range 6-30 MeV. A computer graphical user interface for automatic calculation of production yields taking into account various reaction channels leading to the same final product has been created. The proposed approach allows us to theoretically estimate the amount of (99m)Tc and its ratio relative to (99g)Tc and other radioisotopes which must be considered reaction contaminants, potentially contributing to additional patient dose in diagnostic studies.

Chronic inflammation is negatively associated with prostate cancer and high-grade prostatic intraepithelial neoplasia on needle biopsy.

Tissue inflammation has been linked to cancer in several disease models. We tested the association between chronic inflammation and prostate cancer (PCa), as well as high-grade prostatic intraepithelial neoplasia (HGPIN), in prostatic needle biopsy specimens. Tissues from 4526 men, who underwent systematic ultrasound-guided sextant needle biopsies of the prostate, were classified in the following order as PCa, or HGPIN, or chronic inflammation or benign. PCa was diagnosed in 1633 (36.1%), HGPIN in 535 (11.8%) and chronic inflammation in 347 (7.7%). Chronic inflammation conferred a protective effect from PCa: odds ratio (OR) = 0.20, 95% confidence interval (CI) = 0.15-0.28. Chronic inflammation was also inversely associated with HGPIN: OR = 0.11, 95% CI = 0.05-0.22. The ORs remained virtually unchanged after adjustment for age, serum prostate-specific antigen (PSA), digital rectal examination (DRE) and gland volume. Chronic inflammation is more frequent in the presence of benign histology than it is in the presence of PCa or HGPIN.

French-Canadian linguistic validation of the NIH Chronic Prostatitis Symptom Index.

INTRODUCTION: The NIH Chronic Prostatitis Symptom Index (CPSI) is recommended in the clinical evaluation of men with chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS). However, its use is not possible in French speakers, as it has not been validated in this population. We performed a linguistic validation of the CPSI. METHODS: Linguistic translation followed the forward-backward-forward technique and relied on professional medical translators, bilingual health professionals, and patient input. Along with the SF-12, the translated version was administered to a convenience sample of men presenting for pre-vasectomy visits (controls) and to consecutive patients with established CP/CPPS (cases). Men with CP/CPPS were subsequently asked to complete a 14-day retest questionnaire. Psychometric testing addressed standard reliability and validity characteristics. RESULTS: Thirty-six cases and 38 controls with respective mean ages of 46.5 and 44.0 years participated and 33 (91.2%) cases completed the retest questionnaire. Pain (p<0.001), urinary (p<0.001) and quality-of-life (QOL) scale (p<0.001) score means differed between cases and controls. For the same scales, Cronbach's alphas for cases were respectively 0.70, 0.72 and 0.79 versus 0.80, 0.57, and 0.88 for controls. The retest product-moments were 0.83 for pain, 0.55 for urinary, and 0.83 for QOL scales. In cases, strong correlation was noted between QOL and pain scales (r=0.7), and between urinary and pain scales (r=0.6), versus moderate correlation between QOL and urinary scales (r=0.4). Negative correlation was recorded between CPSI scales and SF-12 scales, which ranged from -0.2 to -0.4. CONCLUSIONS: When applied to CPPS and control subjects, the French Canadian CPSI translation demonstrates excellent discriminant properties. Moreover, its reliability and validity characteristics confirm the qualities of the CPSI as a standard evaluative tool for men with CPPS.

Erythrinaline alkaloids from the flowers and pods of Erythrina lysistemon and their DPPH radical scavenging properties.

Fourteen different erythrinaline alkaloids have been isolated from the flowers and pods of Erythrina lysistemon with four being reported for the first time in nature and five for the first time in this species and the rest having been re-isolated. The new compounds are (+)-11beta-hydroxyerysotramidine (1), (+)-11beta-methoxyerysotramidine (2), (+)-11beta-hydroxyerysotrine N-oxide (4) and (+)-11beta-hydroxyerysotrine (8). (+)-11alpha-Hydroxyerysotrine N-oxide (3), earlier misidentified as erythrartine N-oxide (beta-hydroxyerysotrine N-oxide 4), was also re-isolated along with four other alkaloids. Correct identification of compounds 4 and 8 was aided by the fact that the two sets of C-11 epimers 3, 4 and 8, 9 were both isolated in this study thus making it easier to identify and assign the individual epimers. (+)-Erythristemine (14) was found distributed in most of the plant parts investigated. Preliminary work on the crude chloroform/methanol (1:1) showed moderate toxicity to brine shrimp (LC50 23 ppm) and moderate (IC50 86 microg/ml) radical scavenging properties against stable 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical. The DPPH radical scavenging properties of the isolated compounds were assessed using TLC autographic and spectrophotometric assays whereupon only compounds 11 (1 microg; 90 microg/ml) and 12 (0.1 microg; 160 microg/ml) showed any notable activity. It appears the two compounds are slow reacting and do not reach steady state conditions within the standard half an hour time frame but only seemed to have reached steady state conditions after 4 h.

Erythrinaline alkaloids and antimicrobial flavonoids from Erythrina latissima.

The seed pods of Erythrina latissima yielded erysotrine, erysodine, syringaresinol, vanillic acid, a new erythrina alkaloid, (+)-10,11-dioxoerysotrine, which was lethal to brine shrimp and 2-(5'-hydroxy-3'-methoxy phenyl)-6-hydroxy-5-methoxybenzofuran, which showed strong antimicrobial activity against the yeast spores, Gram-positive and Gram-negative bacteria. The root bark gave four known pterocarpans which showed moderate to strong antifungal activity against the yeast spores and three known flavonoids showed antimicrobial activity against all test microorganisms.

Superiority of iodine-123 compared with iodine-131 scanning for thyroid remnants in patients with differentiated thyroid cancer.

PURPOSE: Iodine-123 is a pure gamma emitter and has excellent characteristics for imaging with modern scintillation cameras. The goal of this study was to compare the performance of I-123 and I-131 as imaging agents for whole-body scanning in patients with differentiated thyroid cancer undergoing ablation for thyroid remnants after initial surgery. METHODS: Fourteen patients with differentiated thyroid cancer who had undergone near-total thyroidectomy and had serum thyroid-stimulating hormone levels greater than 40 mU/I underwent diagnostic scanning 5 hours after administration of 48 to 56 MBq (1.3 to 1.5 mCi) I-123 and 48 hours after administration of 111 MBq (3 mCi) I-131. After receiving ablative I-131 therapy, they also underwent whole-body planar imaging 7 days later. The diagnostic I-123 and I-131 scans were compared with each other and with the post-therapy images by two nuclear medicine physicians and one endocrinologist. RESULTS: The diagnostic scans revealed 35 foci in the thyroid bed and neck. The I-123 images showed all 35 foci, but only 32 of the 35 foci (91 %) were seen on the I-131 scans. The findings of pre- and post-therapy scans were concordant in 11 of 13 patients, and the same general sites of uptake (left and right thyroid bed, midline) were revealed on both sets of images. In one patient, a focus seen on the diagnostic I-123 and I-131 images was not visualized on the post-therapy scan and was thought to represent possible stunning. An additional area of uptake in the lower right neck and upper mediastinum was present on the post-therapy scan of another patient, but it was not seen on diagnostic images. CONCLUSION: These results show improved quality of imaging with 50 MBq (1.5 mCi) I-123 compared with 111 MBq (3 mCi) I-131 for whole-body scanning in patients with differentiated thyroid cancer undergoing thyroid remnant ablation. I-123 imaging may prove to be the preferred procedure in such settings in patients with differentiated thyroid cancer.

Extensive F-18 FDG uptake in metastatic spindle cell carcinoma of the lung.

A 77-year-old man with stage IIB squamous cell carcinoma of the lung underwent right upper lobectomy. One month later he was examined for right chest pain, dyspnea, cough, and weakness. A roentgenogram showed nondiagnostic diffuse opacification of his right lung cavity. An F-18 FDG positron emission tomographic (PET) study revealed extensive uptake in the right pleural area, left adrenal gland, right axilla, and soft tissues consistent with extensive local recurrence and metastatic disease. Biopsy of a right chest soft tissue lesion showed spindle cell carcinoma, a rare variant of squamous cell carcinoma.

Intracavernosal alprostadil is effective for the treatment of erectile dysfunction in diabetic men.

The efficacy and safety of intracavernosal alprostadil was evaluated for the treatment of erectile dysfunction in men with type I or type II diabetes mellitus. This was an open-label, flexible dose-escalating study involving 336 men (77% of whom were Asian/Oriental) enrolled by 15 centres in Australia, Canada and seven countries in Asia. The effective alprostadil dose, ie the dose producing penile rigidity adequate for intercourse and lasting up to 60 min, was established by titration at the clinic prior to entry into the 6 month self-treatment home phase. All men were fully trained in the self-injection technique before entry into the home phase. Efficacy and safety were assessed using patient and partner diaries and by interview at clinic visits during the titration phase and after 1, 3 and 6 months of treatment. An effective home dose was established by titration for 94% of the 336 men (median dose 20 microg, range 2.5-60 microg). Of 278 (83%) men who entered the home phase, 277 men (247 with type II diabetes and 30 with type I diabetes) had evaluable data for alprostadil dosage and clinical response. During the home phase, a satisfactory erectile response was achieved after 99% of injections, and the median alprostadil dose remained unchanged. The initial home dose and clinical response were similar in type I and type II diabetic men. Treatment was generally well tolerated with a low incidence of penile pain (24%) In conclusion, intracavernosal alprostadil was effective and well tolerated in type I and type II diabetic men with erectile dysfunction of mixed aetiology.

Long term follow-up of intravesical Bacillus Calmette-Guerin for the treatment of bladder transitional cell carcinoma.

OBJECTIVE: To review the long-term follow-up, in terms of recurrence and progression, of transitional cell carcinoma of the bladder treated with intravesical BCG with the following indications: CIS, Ta and T1. MATERIALS AND METHODS: Ninety-two patients who had received complete course of BCG between 1987 and 1993 were included in the study and followed for an average of 59 months (range 12 to 102). RESULTS: The recurrence and progression were looked at. Patients treated with BCG for Carcinoma in situ, 11 of 19 (53%) remained tumor-free after 1 or 2 courses of BCG for the duration of the follow-up (mean 4.9 years, range 1.5 to 8.5 years). For patients treated for recurring tumors, 17 of 50 (34%) had no recurrences after 1 or 2 courses of BCG with the same follow-up. When facing multiple tumors, 10 of 23 (43%) patients did not experience recurrences. Therefore, in the 92 patients treated, 38 presented no recurrences after 1 or 2 courses of BCG, for a success rate of 41%. In terms of progression, of the 19 patients treated with BCG for CIS, 4 (21%) went on to develop muscle invasive disease. Of the 50 patients treated for recurrent tumors, 2 (4%) eventually developed lamina propria invasion (initial lesion was a Ta tumor), 4 (8%) carcinoma in situ and 7 (14%) muscle invasive disease, for an overall progression rate of 26% in this group. Of the 25 patients treated for multiple tumors, 1 (4%) developed CIS and 3 (12%) presented with muscle invasive disease, for an overall progression rate of 16% for the duration of the follow-up. Therefore, 21 of 92 (23%) patients had progression of their disease following BCG therapy. No prognostic factors for recurrence or progression could be identified in these tumors. CONCLUSION: When indications warrant its use, BCG is effective in reducing recurrences and limiting progression in TCC of the bladder. Recurrence within 2 years of treatment is, however, a sign of poor prognosis and other therapeutic options should be sought.

Impact of attenuation correction on the accuracy of FDG-PET in patients with abdominal tumors: a free-response ROC analysis.

The aim of this study was to evaluate image quality and lesion detectability with and without attenuation correction in patients with abdominal tumors, using a free-response receiver operating characteristic (FROC) methodology. Thirty-four patients with various abdominal tumors were evaluated (11 men, 23 women, median age 48 years). Whole-body emission scans were performed 68 min (35-102 min) after intravenous injection of 4.3 MBq/kg fluorine-18 fluorodeoxyglucose (FDG). Images were reconstructed using the OS-EM algorithm and corrected for attenuation either using postinjection singles transmission (n=27) or by calculation and body outline (n=7). Total scan duration did not exceed 70 min. Studies were read independently by four observers unaware of any clinical data. The uncorrected (UC) images were systematically read before the attenuation-corrected (AC) images. All studies were given an image quality score ranging from 1 (unreadable) to 5 (excellent). Each focus of increased activity was then localized and given a probability of malignancy using a five-point scale. The average image quality score was similar for both UC and AC images. At the time of the positron emission tomography (PET) scans, 127 lesions (63 liver metastases, 9 retroperitoneal lesions, 50 peritoneal or bowel lesions, and 5 pancreatic carcinomas) were revealed by pathological or correlative studies. The areas under the FROC curves were consistently greater for AC images (range 0.8663-0.8867) than for UC images (range 0.7774 -0.8613). Overall, the difference between the AC images and the UC images was significant (P=0.019). In particular, correction for attenuation increased the sensitivity regardless of the location of the lesions. In conclusion, correction for attenuation significantly improves the diagnostic accuracy of FDG-PET for abdominal staging of neoplasms, without impairing the image quality.