The road to optimal acceleration of Dixon imaging and quantitative T2-mapping in the ankle using compressed sensing and parallel imaging.

OBJECTIVE: This study aimed to find the optimal acceleration factor achievable with CS-SENSE for a clinical ankle protocol while maintaining comparable image quality. METHODS: We explored the optimal acceleration achievable with factor CS-SENSE, for an ankle protocol with T2-weighted, PD-weighted TSE-Dixon (coronal, axial and sagittal) and T2-mapping (sagittal) sequences, on a 3 T MRI-scanner. This study contained three steps: (1) phantom test, (2) pilot test on healthy volunteers, (3) anatomical assessment on a cohort of healthy volunteers and a quantitative analysis. CS-SENSE images (acceleration factors between 2.0× and 12.0×) were compared to reference SENSE images (acceleration factor 2.0×). Three blinded radiologists evaluated the image quality and provided an anatomical assessment using a five-point Likert scale of 25 anatomical regions. RESULTS: The total acquisition time of the TSE-Dixon sequence was reduced by 45 % from 13'38″ to 7'37″ (acceleration factor between 3.6× and 4.0×), the T2-mapping scan time was reduced by 31 % from 5'28″ to 3'47″ (acceleration factor of 3.0×), while maintaining comparable image quality. The results from the anatomical assessment of SENSE 2.0× versus CS-SENSE 3.6× were comparable in 88.7 % as shown by the 5-point Likert scale measurements. The T2-relaxation measurements had a good correlation of ρ = 0.7 between SENSE and CS-SENSE. CONCLUSION: We found an optimum acceleration factor with CS-SENSE between 3.6× and 4.0× for TSE-Dixon and 3.0× for T2-mapping sequences in a clinical MR imaging protocol of the ankle. The total scan time was reduced by 41 % while maintaining adequate image quality.

Donor characteristics do not influence transfusion-related acute lung injury incidence in a secondary analysis of two case-control studies.

OBJECTIVE: To investigate the relation between donor characteristics and TRALI incidence. BACKGROUND: Transfusion-related acute lung injury (TRALI) is a potentially fatal complication of transfusion. In pre-clinical studies and several clinical studies, TRALI has been related to loss of product quality during red blood cell (RBC) storage, called the "storage lesion". Donor characteristics, as for example age, genetics and life style choices influence this "storage lesion". We hypothesized that donor sex, age and blood type is related to TRALI incidence. METHODS/MATERIALS: We performed a secondary analysis of two cohort studies, designed to identify TRALI risk factors by matching TRALI patients to transfused controls. We obtained donor sex, age and blood type from the Dutch Blood Bank Sanquin and investigated TRALI incidence in patients who were exposed to a certain donor characteristic. We used Kruskal-Wallis testing to compare the number of transfused products and Chi2 testing to compare proportions of TRALI patients and transfused control. RESULTS: After implementation of the male-donor only plasma strategy, patients received more transfusion products from male donors. However, we did not detect a relation between TRALI incidence and donor sex. Both TRALI patients and transfused controls received mainly products from donors over 41 years old, but donor age did not influence TRALI risk. Donor blood type, the transfusion of blood type-compatible and blood type-matched products also had no influence on TRALI incidence. CONCLUSION: We conclude that in two cohorts of TRALI patients, donor age, donor sex and donor blood type are unrelated to TRALI.

Compressed Sensing-Sensitivity Encoding (CS-SENSE) Accelerated Brain Imaging: Reduced Scan Time without Reduced Image Quality.

BACKGROUND AND PURPOSE: Compressed sensing-sensitivity encoding is a promising MR imaging acceleration technique. This study compares the image quality of compressed sensing-sensitivity encoding accelerated imaging with conventional MR imaging sequences. MATERIALS AND METHODS: Patients with known, treated, or suspected brain tumors underwent compressed sensing-sensitivity encoding accelerated 3D T1-echo-spoiled gradient echo or 3D T2-FLAIR sequences in addition to the corresponding conventional acquisition as part of their clinical brain MR imaging. Two neuroradiologists blinded to sequence and patient information independently evaluated both the accelerated and corresponding conventional acquisitions. The sequences were evaluated on 4- or 5-point Likert scales for overall image quality, SNR, extent/severity of artifacts, and gray-white junction and lesion boundary sharpness. SNR and contrast-to-noise ratio values were compared. RESULTS: Sixty-six patients were included in the study. For T1-echo-spoiled gradient echo, image quality in all 5 metrics was slightly better for compressed sensing-sensitivity encoding than conventional images on average, though it was not statistically significant, and the lower bounds of the 95% confidence intervals indicated that compressed sensing-sensitivity encoding image quality was within 10% of conventional imaging. For T2-FLAIR, image quality of the compressed sensing-sensitivity encoding images was within 10% of the conventional images on average for 3 of 5 metrics. The compressed sensing-sensitivity encoding images had somewhat more artifacts (P = .068) and less gray-white matter sharpness (P = .36) than the conventional images, though neither difference was significant. There was no significant difference in the SNR and contrast-to-noise ratio. There was 25% and 35% scan-time reduction with compressed sensing-sensitivity encoding for FLAIR and echo-spoiled gradient echo sequences, respectively. CONCLUSIONS: Compressed sensing-sensitivity encoding accelerated 3D T1-echo-spoiled gradient echo and T2-FLAIR sequences of the brain show image quality similar to that of standard acquisitions with reduced scan time. Compressed sensing-sensitivity encoding may reduce scan time without sacrificing image quality.

The practice of diagnosing and reporting transfusion-associated circulatory overload: a national survey among physicians and haemovigilance officers.

OBJECTIVES: This study aims at identifying factors that disciplines consider when diagnosing and reporting transfusion-associated circulatory overload ('TACO'). BACKGROUND: TACO is a clinical diagnosis based mainly on subjective factors. Therefore, TACO could be an underreported complication of blood transfusion. METHODS: A survey was conducted among critical care physicians, anaesthesiologists, haematologists, transfusion medicine physicians and haemovigilance officers using case vignettes and a questionnaire. Factors that may affect diagnosing TACO were investigated using conjoint analysis. A positive B-coefficient indicates a positive preference for diagnosing TACO. Participants rated factors influencing reporting TACO on a 0- to 100-point scale. RESULTS: One hundred and seven surveys were returned (62%). Vignettes showed preferences in favour of diagnosing TACO with the onset of symptoms within 2 h [ß 0·4(-0·1-1·0)], positive fluid balance [ß 0·9(0·4-1·5)] and history of renal failure [ß 0·6(0·1-1·2)]. Compared with transfusion of a single unit of red blood cells (RBC), respondents showed a preference for diagnosing TACO following a single unit of solvent/detergent (S/D) plasma or pooled platelet concentrate (PPC) [ß 0·3(-0·2-0·7) resp. 0·5(-0·1-1·2)]. Multiple transfusion (6 RBC + 4 S/D plasma) was a strong preference for diagnosing TACO compared to 1 RBC and 1 S/D plasma [ß 0·3(-0·8-1·3)]. Respondents did not fully take into account new hypertension and tachycardia when reporting TACO [median 70 (IQR 50-80) resp. 60 (IQR 50-80)]. No differences were observed between disciplines involved. CONCLUSION: When diagnosing and reporting TACO, physicians and haemovigilance officers do consider known risk factors for TACO. Reporting could be improved by increasing the awareness of haemodynamic variables in future education programmes.

The practice of platelet transfusion prior to central venous catheterization in presence of coagulopathy: a national survey among clinicians.

BACKGROUND: Correction of coagulopathy prior to central venous catheter (CVC) placement is advocated by guidelines, while retrospective studies support restrictive use of transfusion products. STUDY DESIGN AND METHODS: We conducted a mixed vignette and questionnaire web survey to investigate current practice and preferences for CVC placement. Clinical vignettes were used to quantify the tendency to administer platelet concentrate. A positive ß-coefficient is in favour of administering platelet concentrate. RESULTS: Ninety-seven physicians answered the survey questions (36 critical care physicians, 14 haematologists, 20 radiologists and 27 anaesthesiologist). Eighty-six physicians subsequently completed the clinical vignettes (response rate 71%). Preferences in favour of correcting thrombocytopenia prior CVC placement were platelet counts of 10 × 109 /L and 20 × 109 /L (ß = 3·9; ß = 3·2, respectively), the subclavian insertion site (ß = 0·8). An elevated INR (INR = 3; ß = 0·6) and an elevated aPTT (aPTT = 60 s; ß = 0·4) showed a positive trend towards platelet transfusion. Platelet transfusion was less likely in an emergency setting (ß = -0·4). Reported transfusion thresholds for CVC placement varied from <10 × 109 /L to 80 × 109 /L for platelet count, from 1·0 to 10·0 for INR and from 25 s to 150 s for aPTT. Implementation of ultrasound guidance as standard practice was limited. CONCLUSION: Current transfusion practice prior to CVC placement is highly variable. Physicians adjust the decision to correct coagulopathy prior CVC placement based on clinical parameters, insertion site and technique applied.

Neural network output optimization using interval analysis.

The problem of output optimization within a specified input space of neural networks (NNs) with fixed weights is discussed in this paper. The problem is (highly) nonlinear when nonlinear activation functions are used. This global optimization problem is encountered in the reinforcement learning (RL) community. Interval analysis is applied to guarantee that all solutions are found to any degree of accuracy with guaranteed bounds. The major drawbacks of interval analysis, i.e., dependency effect and high-computational load, are both present for the problem of NN output optimization. Taylor models (TMs) are introduced to reduce these drawbacks. They have excellent convergence properties for small intervals. However, the dependency effect still remains and is even made worse when evaluating large input domains. As an alternative to TMs, a different form of polynomial inclusion functions, called the polynomial set (PS) method, is introduced. This new method has the property that the bounds on the network output are tighter or at least equal to those obtained through standard interval arithmetic (IA). Experiments show that the PS method outperforms the other methods for the NN output optimization problem.