Centroid-based Approaches Research Articles

Patterns of Failure After Intensity Modulated Radiation Therapy in Head and Neck Squamous Cell Carcinoma of Unknown Primary: Implication of Elective Nodal and Mucosal Dose Coverage

PurposeWe evaluated the geometric and dosimetric-based distribution of mucosal and nodal recurrences in patients with metastatic head and neck squamous cell carcinoma to cervical lymph nodes of unknown primary after intensity modulated radiation therapy using validated typology-indicative taxonomy. Methods and MaterialsWe reviewed the data of 260 patients who were irradiated between 2000 and 2015 and had a median follow-up time for surviving patients of 61 months. The mucosal and nodal recurrences were manually delineated on computed tomography images demonstrating the recurrences. The images were overlaid on the treatment plan using deformable image registration. The locations of the recurrences were determined relative to the original planning target volumes and doses using centroid-based approaches. Subsequently, the pattern of failures were classified into 5 types based on combined spatial and dosimetric criteria: A (central high dose), B (peripheral high dose), C (central elective dose), D (peripheral elective dose), and E (extraneous dose). For patients with type A failure with simultaneous nontype A lesions, the overall pattern of failures was defined as type A. ResultsThirty-two patients had mucosal or nodal recurrences. The most common clinical nodal stage was N2b (66%). Preradiation therapy neck dissections were performed in 6 patients. The median dose delivered to clinical tumor volume 1 was 66 Gy. The majority (84%) had total/partial pharyngeal mucosa elective irradiation. Twenty-three patients had nodal recurrences, 8 had mucosal recurrences, and 1 had both nodal and mucosal recurrences. Twenty-one patients (91%) had type A nodal failure, and 7 of the mucosal failures (89%) were type C. ConclusionsThe majority of nodal recurrences occurred within the high-dose area, demanding the need for identification of radioresistant areas within malignant nodes. Future studies should focus on either dose escalation of high-risk volumes or novel radiosensitizers.

Hybrid Models for Adaptive Allocation of Electricity for Households

In this paper, we analyze, model, predict and cluster Global Active Power, i.e., a time series data obtained at one minute intervals from electricity sensors of a household. We analyze changes in seasonality and trends to model the data. We then compare various forecasting methods such as SARIMA and LSTM to forecast sensor data for the household and combine them to achieve a hybrid model that captures nonlinear variations better than either SARIMA or LSTM used in isolation. Finally, we cluster slices of time series data effectively using a novel clustering algorithm that is a combination of density-based and centroid-based approaches, to discover relevant subtle clusters from sensor data. Our experiments have yielded meaningful insights from the data at both a micro, day-to-day granularity, as well as a macro, weekly to monthly granularity.

Boosting Search Performance Using Query Variations

Rank fusion is a powerful technique that allows multiple sources of information to be combined into a single result set. Query variations covering the same information need represent one way in which different sources of information might arise. However, when implemented in the obvious manner, fusion over query variations is not cost-effective, at odds with the usual web-search requirement for strict per-query efficiency guarantees. In this work, we propose a novel solution to query fusion by splitting the computation into two parts: one phase that is carried out offline, to generate pre-computed centroid answers for queries addressing broadly similar information needs, and then a second online phase that uses the corresponding topic centroid to compute a result page for each query. To achieve this, we make use of score-based fusion algorithms whose costs can be amortized via the pre-processing step and that can then be efficiently combined during subsequent per-query re-ranking operations. Experimental results using the ClueWeb12B collection and the UQV100 query variations demonstrate that centroid-based approaches allow improved retrieval effectiveness at little or no loss in query throughput or latency and within reasonable pre-processing requirements. We additionally show that queries that do not match any of the pre-computed clusters can be accurately identified and efficiently processed in our proposed ranking pipeline.

Times Series Averaging and Denoising from a Probabilistic Perspective on Time–Elastic Kernels

Abstract In the light of regularized dynamic time warping kernels, this paper re-considers the concept of a time elastic centroid for a set of time series. We derive a new algorithm based on a probabilistic interpretation of kernel alignment matrices. This algorithm expresses the averaging process in terms of stochastic alignment automata. It uses an iterative agglomerative heuristic method for averaging the aligned samples, while also averaging the times of their occurrence. By comparing classification accuracies for 45 heterogeneous time series data sets obtained by first nearest centroid/medoid classifiers, we show that (i) centroid-based approaches significantly outperform medoid-based ones, (ii) for the data sets considered, our algorithm, which combines averaging in the sample space and along the time axes, emerges as the most significantly robust model for time-elastic averaging with a promising noise reduction capability. We also demonstrate its benefit in an isolated gesture recognition experiment and its ability to significantly reduce the size of training instance sets. Finally, we highlight its denoising capability using demonstrative synthetic data. Specifically, we show that it is possible to retrieve, from few noisy instances, a signal whose components are scattered in a wide spectral band.

Patterns of Failure After Radiation Therapy in Head and Neck Squamous Cell Carcinoma of Unknown Primary: Implication of Elective Nodal and Mucosal Dose Coverage

There is a debate about the optimal mucosal and nodal dose coverage in head and neck squamous cell carcinoma of unknown primary (HN-SCCUP). We evaluated the geometric and dosimetric-based distribution of mucosal and regional failures (MRFs) in patients with HN-SCCUP using validated typology-indicative taxonomy. Out of 249 patients treated definitively with radiotherapy (RT) for HN-SCCUP with intact imaging data, 23 patients had mucosal and/or regional failures. This retrospective study is approved by the institutional IRB. A total of 28 (recurrent gross mucosal/nodal volumes) rGM/NVs were delineated on follow-up images showing the failures. The images were overlaid on the treatment plan using deformable image registration. rGM/NVs were determined relative to original planning target volumes and dose using a centroid-based approaches. The pattern of failures (POF) were then classified into five types based on combined spatial and dosimetric criteria; A (central high dose), B (peripheral high dose), C (central elective dose), D (peripheral elective dose), and E (extraneous dose). For patients with type A failure with simultaneous non-type A lesions, the overall POF was defined as type A. Fifteen patients had regional failure (RF), 4 had mucosal failure (MF), and 4 had MRF. Median age was 60 (range: 47 -84) years. Tonsillectomy was performed in 15 (60%) patients. The most common clinical nodal stage was N2b (40%). Neck dissections were performed in 14 patients – 5 pre-RT and 9 post-RT. Seven patients had induction and 8 patients had concurrent chemotherapy with RT. The median dose delivered was 66 (range: 60 – 70) Gy. IMRT was utilized in 21 (91%) patients, with lower neck split field technique used in 9 patients. The majority (10, 43%) had total pharyngeal mucosa irradiation, followed by nasopharyngeal and oropharyngeal coverage in 6 (26%) patients. Four out of the 8 MF were not covered by the elective mucosal dose. Median time to failure was 16.2(range: 1.7-113.23) months. Mean rGM/N volume was 9.959 cm3 (range 0.4–63.5).For spatial and dosimetric analysis of POF, 17 (72%) patients had type A failure. Non-type A failures were distributed as follows: 2 (8%) type C, 1 (4%) type D and 4 (16%) type E. In those who had MF, the primary lesion was detected within the oropharynx in 3, larynx in 4 patients and 1 in oral cavity. 7 mucosal and 14 regional failures received subsequent salvage treatment. Median overall survival was 46.56(range: 6.8-145.46) months. The majority of RF occurred within the high-dose area and MF occurred within the pharyngeal axis, suggesting that the dose coverage to the mucosa and lymph node regions should be carefully designed for CUP patients. In addition, given the low incidence of geographical misses, future studies should be directed towards dose escalation of high-risk volumes.

Patterns of locoregional failure following post-operative intensity-modulated radiotherapy to oral cavity cancer: quantitative spatial and dosimetric analysis using a deformable image registration workflow

BackgroundWe sought to identify spatial/dosimetric patterns of failure for oral cavity cancer patients receiving post-operative IMRT (PO-IMRT).MethodsTwo hundred eighty-nine OCC patients receiving PO-IMRT were retrospectively reviewed from 2000 to 2012. Diagnostic CT documenting recurrence (rCT) was co-registered with planning CT (pCT) using a validated deformable image registration software. Manually segmented recurrent gross disease (rGTV) was deformed to co-registered pCTs. Mapped rGTVs were compared dosimetrically to planned dose and spatially to planning target volumes using centroid-based approaches. Failures types were classified using combined spatial/dosimetric criteria: A (central high-dose), B (peripheral high-dose), C (central intermediate/low-dose), D (peripheral intermediate/low-dose), and E (extraneous-dose).ResultsFifty-four patients with recurrence were analyzed; 26 local recurrence, 19 regional recurrence, and 9 both local and regional recurrence. Median time to recurrence was 4 months (range 0–71). Median rGTVs volume was 3.7 cm3 (IQR 1.4–10.6). For spatial and dosimetric analysis of the patterns of failure, 30 patients (55.5%) were classified as type A (central high-dose). Non-central high dose failures were distributed as follows: 2 (3.7%) type B, 10 (18.5%) type C, 1 (1.8%) type D, and 9 (16.7%) type E. Non-IMRT failure in the matching low-neck field was seen in two patients. No failures were noted at the IMRT-supraclavicular field match-line.ConclusionsApproximately half of patients with local/regional failure had non-central high dose recurrence. Peripheral high dose misses were uncommon reflecting adequate delineation and dose delivery. Future strategies are needed to reduce types C and E failures.

Decentralized Clustering by Finding Loose and Distributed Density Cores

Centroid-based clustering approaches fail to recognize extremely complex patterns that are non-isotropic. We analyze the underlying causes and find some inherent flaws in these approaches, including Shape Loss, False Distances and False Peaks, which typically cause centroid-based approaches to fail when applied to complex patterns. As an alternative to current methods, we propose a hybrid decentralized approach named DCore, which is based on finding density cores instead of centroids, to overcome these flaws. The underlying idea is that we consider each cluster to have a shrunken density core that roughly retains the shape of the cluster. Each such core consists of a set of loosely connected local density peaks of higher density than their surroundings. Borders, edges and outliers are distributed around the outsides of these cores in a hierarchical structure. Experiments demonstrate that the promise of DCore lies in its power to recognize extremely complex patterns and its high performance in real applications, for example, image segmentation and face clustering, regardless of the dimensionality of the space in which the data are embedded.

Methodology for analysis and reporting patterns of failure in the Era of IMRT: head and neck cancer applications.

BackgroundThe aim of this study is to develop a methodology to standardize the analysis and reporting of the patterns of loco-regional failure after IMRT of head and neck cancer.Material and MethodsTwenty-one patients with evidence of local and/or regional failure following IMRT for head-and-neck cancer were retrospectively reviewed under approved IRB protocol. Manually delineated recurrent gross disease (rGTV) on the diagnostic CT documenting recurrence (rCT) was co-registered with the original planning CT (pCT) using both deformable (DIR) and rigid (RIR) image registration software. Subsequently, mapped rGTVs were compared relative to original planning target volumes (TVs) and dose using a centroid-based approaches. Failures were then classified into five types based on combined spatial and dosimetric criteria; A (central high dose), B (peripheral high dose), C (central elective dose), D (peripheral elective dose), and E (extraneous dose).ResultsA total of 26 recurrences were identified. Using DIR, recurrences were assigned to more central TVs compared to RIR as detected using the spatial centroid-based method (p = 0.0002). rGTVs mapped using DIR had statistically significant higher mean doses when compared to rGTVs mapped rigidly (mean dose 70 vs. 69 Gy, p = 0.03). According to the proposed classification 22 out of 26 failures were of type A (central high dose) as assessed by DIR method compared to 18 out of 26 for the RIR because of the tendencey of RIR to assign failures more peripherally.ConclusionsRIR tends to assigns failures more peripherally. DIR-based methods showed that the vast majority of failures originated in the high dose target volumes and received full prescribed doses suggesting biological rather than technology-related causes of failure. Validated DIR-based registration is recommended for accurate failure characterization and a novel typology-indicative taxonomy is recommended for failure reporting in the IMRT era.Electronic supplementary materialThe online version of this article (doi:10.1186/s13014-016-0678-7) contains supplementary material, which is available to authorized users.

A GENETIC GRAPH-BASED APPROACH FOR PARTITIONAL CLUSTERING

Clustering is one of the most versatile tools for data analysis. In the recent years, clustering that seeks the continuity of data (in opposition to classical centroid-based approaches) has attracted an increasing research interest. It is a challenging problem with a remarkable practical interest. The most popular continuity clustering method is the spectral clustering (SC) algorithm, which is based on graph cut: It initially generates a similarity graph using a distance measure and then studies its graph spectrum to find the best cut. This approach is sensitive to the parameters of the metric, and a correct parameter choice is critical to the quality of the cluster. This work proposes a new algorithm, inspired by SC, that reduces the parameter dependency while maintaining the quality of the solution. The new algorithm, named genetic graph-based clustering (GGC), takes an evolutionary approach introducing a genetic algorithm (GA) to cluster the similarity graph. The experimental validation shows that GGC increases robustness of SC and has competitive performance in comparison with classical clustering methods, at least, in the synthetic and real dataset used in the experiments.