Radiomics modelling in rectal cancer to predict disease-free survival: evaluation of different approaches.

BACKGROUND: Radiomics may be useful in rectal cancer management. The aim of this study was to assess and compare different radiomics approaches over qualitative evaluation to predict disease-free survival (DFS) in patients with locally advanced rectal cancer treated with neoadjuvant therapy. METHODS: Patients from a phase II, multicentre, randomized study (GRECCAR4; NCT01333709) were included retrospectively as a training set. An independent cohort of patients comprised the independent test set. For both time points and both sets, radiomic features were extracted from two-dimensional manual segmentation (MS), three-dimensional (3D) MS, and from bounding boxes. Radiomics predictive models of DFS were built using a hyperparameters-tuned random forests classifier. Additionally, radiomics models were compared with qualitative parameters, including sphincter invasion, extramural vascular invasion as determined by MRI (mrEMVI) at baseline, and tumour regression grade evaluated by MRI (mrTRG) after chemoradiotherapy (CRT). RESULTS: In the training cohort of 98 patients, all three models showed good performance with mean(s.d.) area under the curve (AUC) values ranging from 0.77(0.09) to 0.89(0.09) for prediction of DFS. The 3D radiomics model outperformed qualitative analysis based on mrEMVI and sphincter invasion at baseline (P = 0.038 and P = 0.027 respectively), and mrTRG after CRT (P = 0.017). In the independent test cohort of 48 patients, at baseline and after CRT the AUC ranged from 0.67(0.09) to 0.76(0.06). All three models showed no difference compared with qualitative analysis in the independent set. CONCLUSION: Radiomics models can predict DFS in patients with locally advanced rectal cancer.

Conditional risk of diverticulitis after non-operative management.

BACKGROUND: The objective of this study was to describe conditional recurrence-free survival (RFS) of patients after an index episode of diverticulitis managed without surgery, and to estimate the difference in conditional RFS for diverticulitis according to specific risk factors. METHODS: This was a multicentre retrospective cohort study including all patients managed without surgery for acute sigmoid diverticulitis at two university-affiliated hospitals in Montreal, Quebec, Canada. Conditional RFS for diverticulitis was estimated over 10 years of follow-up. A Cox proportional hazards model was performed at the index episode and again 2 years later. RESULTS: In total, 991 patients were included for analysis. The 1, 2- and 3-year actuarial diverticulitis RFS rates were 81·1, 71·5 and 67·5 per cent respectively. Compared with the 1-year actuarial RFS rate of 81·1 per cent, the 1-year conditional RFS increased with each additional year survived recurrence-free, reaching 96·0 per cent after surviving the first 4 years recurrence-free. A similar phenomenon was observed for 2-year diverticulitis conditional RFS. Lower age (hazard ratio (HR) 0·98, 95 per cent c.i. 0·98 to 0·99), Charlson Co-morbidity Index score of 2 or above (HR 1·78, 1·32 to 2·39) and immunosuppression (HR 1·85, 1·38 to 2·48) were independently associated with recurrence of diverticulitis from the index episode. At 2 years from the index episode, immunosuppression was no longer associated with diverticulitis recurrence (HR 1·02, 0·50 to 2·09). CONCLUSION: The conditional RFS of patients with diverticulitis improved with each year that was survived recurrence-free. Although several factors at index presentation may be associated with early recurrence, the conditional probability of recurrence according to many of these risk factors converged with time.

ANTECEDENTES: El objetivo de este estudio fue describir la supervivencia condicional libre de recidiva de diverticulitis (diverticulitis recurrence-free survival, Div-RFS) en pacientes tras un episodio de diverticulitis tratado de forma conservadora, y calcular la diferencia en la Div-RFS condicional de acuerdo con factores de riesgo específicos. MÉTODOS: Estudio de cohorte retrospectivo multicéntrico que incluyó a todos los pacientes tratados de forma no quirúrgica por diverticulitis sigmoidea aguda en dos hospitales afiliados a la universidad en Montreal, Quebec, Canadá. La supervivencia condicional libre de recidiva de la diverticulitis se calculó durante 10 años de seguimiento. Se realizó un análisis mediante un modelo de riesgos proporcionales de Cox en el episodio índice y nuevamente 2 años después. RESULTADOS: En total, se incluyeron 991 pacientes en el análisis. La Div-RFS actuarial a 1, 2 y 3 años fue del 81,1%, 71,5% y 67,5%, respectivamente. En comparación con la Div-RFS actuarial a 1 año del 81,1%, la Div-RFS condicional a 1 año aumentó con cada año adicional sobrevivido sin recidiva, alcanzando el 96,0% después de sobrevivir los primeros 4 años sin recidiva. Se observó un fenómeno similar para Div-RFS condicional a los 2 años. Una menor edad (cociente de riesgos instantáneos, hazard ratio, HR: 0,98; i.c. del 95%: 0,98 a 0,99), la puntuación de comorbilidad de Charlson ≥ 2 (HR: 1,78; i.c. del 95%: 1,32 a 2,39) y la inmunosupresión (HR: 1,85; i.c. del 95%: 1,38 a 2,48) se asociaron de forma independiente con la recidiva de la diverticulitis desde el episodio índice. En la regresión de Cox a los 2 años del episodio índice, la inmunosupresión ya no se asoció con recidiva de diverticulitis (HR: 1,02; i.c. del 95% 0,50-2,09). CONCLUSIÓN: La Div-RFS condicional mejoró con cada año de supervivencia sin recidiva. Si bien varios factores en la presentación del episodio índice pueden estar asociados con una recidiva precoz, la probabilidad condicional de recidiva en relación con muchos de estos factores de riesgo coincidió con el tiempo.

Performance comparison of modified ComBat for harmonization of radiomic features for multicenter studies.

Multicenter studies are needed to demonstrate the clinical potential value of radiomics as a prognostic tool. However, variability in scanner models, acquisition protocols and reconstruction settings are unavoidable and radiomic features are notoriously sensitive to these factors, which hinders pooling them in a statistical analysis. A statistical harmonization method called ComBat was developed to deal with the "batch effect" in gene expression microarray data and was used in radiomics studies to deal with the "center-effect". Our goal was to evaluate modifications in ComBat allowing for more flexibility in choosing a reference and improving robustness of the estimation. Two modified ComBat versions were evaluated: M-ComBat allows to transform all features distributions to a chosen reference, instead of the overall mean, providing more flexibility. B-ComBat adds bootstrap and Monte Carlo for improved robustness in the estimation. BM-ComBat combines both modifications. The four versions were compared regarding their ability to harmonize features in a multicenter context in two different clinical datasets. The first contains 119 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging and positron emission tomography imaging. In that case ComBat was applied with 3 labels corresponding to each center. The second one contains 98 locally advanced laryngeal cancer patients from 5 centers with contrast-enhanced computed tomography. In that specific case, because imaging settings were highly heterogeneous even within each of the five centers, unsupervised clustering was used to determine two labels for applying ComBat. The impact of each harmonization was evaluated through three different machine learning pipelines for the modelling step in predicting the clinical outcomes, across two performance metrics (balanced accuracy and Matthews correlation coefficient). Before harmonization, almost all radiomic features had significantly different distributions between labels. These differences were successfully removed with all ComBat versions. The predictive ability of the radiomic models was always improved with harmonization and the improved ComBat provided the best results. This was observed consistently in both datasets, through all machine learning pipelines and performance metrics. The proposed modifications allow for more flexibility and robustness in the estimation. They also slightly but consistently improve the predictive power of resulting radiomic models.

Impact of the T2-weighted axial oblique MRI sequence in the assessment of peroneal tendons.

AIM: To define the role of the T2-weighted axial oblique sequence for the magnetic resonance imaging (MRI) assessment of peroneal tendon pathologies. MATERIALS AND METHODS: Two radiologists interpreted 180 ankle MRI examinations using standard sequences alone and then in combination with an axial oblique sequence. The readers indicated how likely a peroneal pathology was present using a five-level confidence scale. Diagnostic confidence, interobserver agreement, and clinical correlation were compared. Changes in diagnosis were recorded. RESULTS: For both readers, the diagnostic confidence was significantly higher using the axial oblique sequence for tendinosis and inframalleolar tenosynovitis for both tendons and for peroneus brevis partial and longitudinal split tears (p<0.001). For reader 1, the diagnostic confidence was also higher using the axial oblique sequence for peroneus longus partial tears (p=0.007). Changes in diagnosis were seen for tendinosis and tenosynovitis of both tendons and for peroneus brevis partial and longitudinal split tears in 0.6-10.8% of cases. Inter-rater reliability was significantly higher with the axial oblique sequence for the diagnosis of tendinosis, inframalleolar tenosynovitis, and partial tear for both tendons, and for peroneus brevis longitudinal split tear. Amongst 105 examinations with clinical information, peroneal pathologies were most frequently diagnosed as present in cases with lateral symptoms (17% versus 14%) and absent in cases without lateral symptoms (92% versus 86%) on the axial oblique sequence. CONCLUSION: The axial oblique sequence for the assessment of peroneal tendons allows for higher diagnostic confidence, inter-rater reliability, and clinical correlation and can lead to changes in diagnosis.

How to differentiate uterine leiomyosarcoma from leiomyoma with imaging.

Uterine leiomyomas, the most frequent benign myomatous tumors of the uterus, often cannot be distinguished from malignant uterine leiomyosarcomas using clinical criteria. Furthermore, imaging differentiation between both entities is frequently challenging due to their potential overlapping features. Because a suspected leiomyoma is often managed conservatively or with minimally invasive treatments, the misdiagnosis of leiomyosarcoma for a benign leiomyoma could potentially result in significant treatment delays, therefore increasing morbidity and mortality. In this review, we provide an overview of the differences between leiomyoma and leiomyosarcoma, mainly focusing on imaging characteristics, but also briefly touching upon their demographic, histopathological and clinical differences. The main indications and limitations of available cross-sectional imaging techniques are discussed, including ultrasound, computed tomography, magnetic resonance imaging (MRI) and positron emission tomography/computed tomography. A particular emphasis is placed on the review of specific MRI features that may allow distinction between leiomyomas and leiomyosarcomas according to the most recent evidence in the literature. The potential contribution of texture analysis is also discussed. In order to help guide-imaging diagnosis, we provide an MRI-based diagnostic algorithm which takes into account morphological and functional features, both individually and in combination, in an attempt to optimize radiologic differentiation of leiomyomas from leiomyosarcomas.

Ovarian cancer: An update on imaging in the era of radiomics.

Tumor heterogeneity in ovarian cancer has been reported at the histological and genetic levels and is associated with adverse clinical outcomes. Tumor evaluation using standard computed tomography or magnetic resonance imaging techniques does not account for the intra- or inter-tumoral heterogeneity in advanced ovarian cancer with peritoneal carcinomatosis. As such, computational approaches in assessing tumor heterogeneity have been proposed using radiomics and radiogenomics in order to analyze the whole tumor heterogeneity as opposed to single biopsy sampling. As part of radiomics, texture analysis, which includes the extraction of multiple data from images has been proposed recently to evaluate advanced ovarian tumor heterogeneity. In this short review, we explain the basics of radiomics, how to perform texture analysis, and its applications to ovarian cancer imaging.

Enhancement of breast cancer on pre-treatment dynamic contrast-enhanced MRI using computer-aided detection is associated with response to neo-adjuvant chemotherapy.

PURPOSE: To evaluate whether enhancement of breast cancer on pre-treatment dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) as evaluated semi-quantitatively using computer-aided detection (CAD) is associated with response to neo-adjuvant chemotherapy. MATERIALS AND METHODS: A total of 84 women, (mean age, 51±10 [SD] years; range: 30-73 years) with 84 breast cancers who underwent MRI before neo-adjuvant chemotherapy were included in this retrospective study. The proportion of each type of signal intensity-time curve (SITC) (type 1: persistent; type 2: plateau; Type 3: washout) within the tumor volume was quantified semi-automatically using a CAD system (Aegis®, Sentinelle medical, Toronto, Canada) and was compared to histological features of the tumors and to pathological response to neo-adjuvant chemotherapy. RESULTS: Pathological complete response was obtained in 29 patients (35%). Proportion of SITC type 1 was greater in non-responders (P=0.019) while proportion of SITC type 3 was greater in responders (P=0.04). Sensitivity, specificity, and accuracy of proportion of SITC type 1 for the identification of incomplete response on pathology were 42% (95% CI: 29%-56%), 90% (95% CI: 73%-98%), and 59% (95% CI: 48%-70%), respectively. CONCLUSION: Proportion of SITC type 1 (persistent) in breast cancers on pre-treatment MRI as semi-automatically quantified using a CAD system is associated with absence of pathological complete response to neo-adjuvant chemotherapy with good specificity.

[Dyspnea on exertion, angina pectoris and loud buzzing in the right groin in an 82-year-old man]. / Belastungsdyspnoe, Angina pectoris und lautes Schwirren über der rechten Leiste bei einem 82-Jährigen.

We report on the case of an 82-year-old man who was suffering from chest pain and dyspnea. Acute cardiac ischemia could be excluded. Cardiac catheterization also revealed an aneurysm of the right common iliac artery. In addition, an arteriovenous fistula between the iliac artery and vein was detected by computer tomography angiography. Apparently, these symptoms were caused by a high output heart failure with known coronary heart disease. The patient was treated by implantation of prosthesis and oversewing the fistula which led to full recovery.

Randomised clinical trial: MRCP-first vs. ERCP-first approach in patients with suspected biliary obstruction due to bile duct stones.

BACKGROUND: The preferred initial investigation with either magnetic resonance (MRCP) or endoscopic retrograde cholangiopancreatography (ERCP) in patients with suspected biliary obstruction remains controversial in many clinical settings. AIM: To assess the effectiveness of an initial MRCP vs. ERCP in the work-up of patients at moderate likelihood of a suspected biliary obstruction. METHODS: Patients with an unconfirmed benign biliary obstruction, based on laboratory and ultrasound findings, were randomised to an ERCP-first or MRCP-first strategy, stratified by level of obstruction. The primary outcome was the occurrence of a disease or procedure-related bilio-pancreatic adverse events within the next 12 months. Secondary outcomes were the number of subsequent bilio-pancreatic procedures, duration of hospitalisation, days away from activities of daily living (ADL), quality of life (SF-36) and mortality. RESULTS: We randomised 126 patients to ERCP-first and 131 to MRCP-first (age 54 ± 18 years, 62% female, 39% post-cholecystectomy). In follow-up, 18/126 (14.3%) ERCP-first and 25/131 (19.1%) MRCP-first patients experienced a procedure- or disease-related complication (P = 0.30) (disease-related in 13 and 18 patients, and procedure-related in 5 and 7 patients respectively). A cause of biliary obstruction was found in 39.7% vs. 49.6% of patients (P = 0.11). Sixty-six (50%) patients in the MRCP-first group ended up avoiding an ERCP in follow-up. ERCP-first and MRCP-first patients were away from usual activities for 3.4 ± 7.7 days and 2.0 ± 4.8 days respectively (P < 0.001). CONCLUSION: A strategy of MRCP-first decreased the need for subsequent MRCPs, but not complications. Further study is required to define factors influencing the eventual use of MRCP vs. ERCP in appropriately selected patients (ClinicalTrial.gov: NCT01424657).

Creating and transporting Trojan wave packets.

Nondispersive localized Trojan wave packets with n(i) ~ 305 moving in near-circular Bohr-like orbits are created and transported to localized near-circular Trojan states of higher n, n(f) ~ 600, by driving with a linearly polarized sinusoidal electric field whose period is slowly increased. The protocol is remarkably efficient with over 80% of the initial atoms being transferred to the higher n states, a result confirmed by classical trajectory Monte Carlo simulations.

Analysis of circular wave packets generated by pulsed electric fields.

We demonstrate that circular wave packets in high Rydberg states generated by a pulsed electric field applied to extreme Stark states are characterized by a position-dependent energy gradient that leads to a correlation between the principal quantum number n and the spatial coordinate. This correlation is rather insensitive to the initial state and can be seen even in an incoherent mix of states such as is generated experimentally allowing information to be placed into, and extracted from, such wave packets. We show that detailed information on the spatial distribution of a circular wave packet can be extracted by analyzing the complex phase of its expansion coefficients.

Multidisciplinary management of vascular access for haemodialysis: from the preparation of the initial access to the treatment of stenosis and thrombosis.

In February 2008 a multidisciplinary study group was established in Germany to improve the treatment of patients with potential vascular access problems. As one of the first results of their work interdisciplinary recommendations for the management of vascular access were provided, from the creation of the initial access to the treatment of complications. As a rule the wrist arteriovenous fistula (AVF) is the access of choice due to its lower complication rate when compared to other types of access. The AVF should be created 3 months prior to the expected start of haemodialysis to allow for sufficient maturation. Second and third choice accesses are arteriovenous grafts (AVG) and central venous catheters (CVC). Ultrasound is a reliable tool for vessel selection before access creation, and also for the diagnosis of complications in AVF and grafts. Access stenosis and thrombosis can be treated surgically and interventionally. The comparison of both methods reveals advantages and disadvantages for each. The therapeutic decision should be based on the individual patients' constitution, and also on the availability and experience of the involved specialists.

Electric-field-induced dissociation of heavy Rydberg ion-pair states.

A classical trajectory Monte Carlo approach is used to simulate the dissociation of H(+)⋅⋅⋅F(-) and K(+)⋅⋅⋅Cl(-) heavy Rydberg ion pairs induced by a ramped electric field, a technique used experimentally to detect and probe ion-pair states. Simulations that include the effects of the strong short-range repulsive interaction associated with ion-pair scattering are in good agreement with experimental results for Stark wavepackets probed by a ramped field, demonstrating that many of the characteristics of field-induced dissociation can be well described using a quasi-classical model. The data also show that states with a given value of principal quantum number (i.e., binding energy) can dissociate over a broad range of applied fields, the exact field being governed by the initial orbital angular momentum and orientation of the state.

Lifetimes of heavy-Rydberg ion-pair states formed through Rydberg electron transfer.

The lifetimes of K(+)..Cl(-), K(+)..CN(-), and K(+)..SF(6)(-) heavy-Rydberg ion-pair states produced through Rydberg electron transfer reactions are measured directly as a function of binding energy using electric field induced detachment and the ion-pair decay channels discussed. The data are interpreted using a Monte Carlo collision code that models the detailed kinematics of electron transfer reactions. The lifetimes of K(+)..Cl(-) ion-pair states are observed to be very long, >100 micros, and independent of binding energy. The lifetimes of strongly bound (>30 meV) K(+)..CN(-) ion pairs are found to be similarly long but begin to decrease markedly as the binding energy is reduced below this value. This behavior is attributed to conversion of rotational energy in the CN(-) ion into translational energy of the ion pair. No long-lived K(+)..SF(6)(-) ion pairs are observed, their lifetimes decreasing with increasing binding energy. This behavior suggests that ion-pair loss is associated with mutual neutralization as a result of charge transfer.

Realization of localized Bohr-like wave packets.

We demonstrate a protocol to create localized wave packets in very-high-n Rydberg states which travel in nearly circular orbits around the nucleus. Although these wave packets slowly dephase and eventually lose their localization, their motion can be monitored over several orbital periods. These wave packets represent the closest analog yet achieved to the original Bohr model of the hydrogen atom, i.e., an electron in a circular classical orbit around the nucleus. The possible extension of the approach to create "planetary atoms" in highly correlated stable multiply excited states is discussed.

Transporting Rydberg electron wave packets with chirped trains of pulses.

A protocol for steering Rydberg electrons towards targeted final states is realized with the aid of a chirped train of half-cycle pulses (HCPs). Its novel capabilities are demonstrated experimentally by transporting potassium atoms excited to the lowest-lying quasi-one-dimensional states in the n(i)=350 Stark manifold to a narrow range of much higher-n states. We demonstrate that this coherent state transfer is, to a high degree, reversible. The protocol allows for remarkable selectivity and is highly efficient, with typically over 80% of the parent atoms surviving the HCP sequence.

Electric dipole echoes in Rydberg atoms.

We report the first observation of echoes in the electric dipole moment of an ensemble of Rydberg atoms precessing in an external electric field F. Rapid reversal of the field direction is shown to play a role similar to that of a pi pulse in NMR in rephasing a dephased ensemble of electric dipoles resulting in the buildup of an echo. The mechanisms responsible for this are discussed with the aid of classical trajectory Monte Carlo simulations.

Charge transfer and ionisation by intermediate-energy heavy ions.

The use of heavy ion beams for microbeam studies of mammalian cell response leads to a need to better understand interaction cross sections for collisions of heavy ions with tissue constituents. For ion energies of a few MeV u(-1) or less, ions capture electrons from the media in which they travel and undergo subsequent interactions as partially 'dressed' ions. For example, 16 MeV fluorine ions have an equilibrium charge of 7(+), 32 MeV sulphur ions have an equilibrium charge of approximately 11(+), and as the ion energies decrease the equilibrium charge decreases dramatically. Data for interactions of partially dressed ions are extremely rare, making it difficult to estimate microscopic patterns of energy deposition leading to damage to cellular components. Such estimates, normally obtained by Monte Carlo track structure simulations, require a comprehensive database of differential and total ionisation cross sections as well as charge transfer cross sections. To provide information for track simulation, measurement of total ionisation cross sections have been initiated at East Carolina University using the recoil ion time-of-flight method that also yields cross sections for multiple ionisation processes and charge transfer cross sections; multiple ionisation is prevalent for heavy ion interactions. In addition, measurements of differential ionisation cross sections needed for Monte Carlo simulation of detailed event-by-event particle tracks are under way. Differential, total and multiple ionisation cross sections and electron capture and loss cross sections measured for C(+) ions with energies of 100 and 200 keV u(-1) are described.

Navigating localized wave packets in phase space.

The ability to localize and to steer Rydberg wave packets in phase space using tailored sequences of half-cycle pulses is demonstrated. Classical phase-space portraits are used to explain the method and to illustrate the level of control that can be achieved. This is confirmed experimentally by positioning a phase-space-localized wave packet at the center of a stable island or navigating it around its periphery. This work provides a valuable starting point for further engineering of electronic wave functions.