Limits...
Characterizing Heterogeneity within Head and Neck Lesions Using Cluster Analysis of Multi-Parametric MRI Data.

Borri M, Schmidt MA, Powell C, Koh DM, Riddell AM, Partridge M, Bhide SA, Nutting CM, Harrington KJ, Newbold KL, Leach MO - PLoS ONE (2015)

Bottom Line: Cumulative distributions of voxels, containing pre and post-treatment data and including both primary tumours and lymph nodes, were partitioned into k clusters (k = 2, 3 or 4).Principal component analysis and cluster validation were employed to investigate data composition and to independently determine the optimal number of clusters.Partitioning with the optimal number of clusters (k = 4), determined with cluster validation, produced the best separation between reducing and non-reducing clusters.

View Article: PubMed Central - PubMed

Affiliation: CR-UK Cancer Imaging Centre, The Royal Marsden NHS Foundation Trust and The Institute of Cancer Research, London, United Kingdom.

ABSTRACT

Purpose: To describe a methodology, based on cluster analysis, to partition multi-parametric functional imaging data into groups (or clusters) of similar functional characteristics, with the aim of characterizing functional heterogeneity within head and neck tumour volumes. To evaluate the performance of the proposed approach on a set of longitudinal MRI data, analysing the evolution of the obtained sub-sets with treatment.

Material and methods: The cluster analysis workflow was applied to a combination of dynamic contrast-enhanced and diffusion-weighted imaging MRI data from a cohort of squamous cell carcinoma of the head and neck patients. Cumulative distributions of voxels, containing pre and post-treatment data and including both primary tumours and lymph nodes, were partitioned into k clusters (k = 2, 3 or 4). Principal component analysis and cluster validation were employed to investigate data composition and to independently determine the optimal number of clusters. The evolution of the resulting sub-regions with induction chemotherapy treatment was assessed relative to the number of clusters.

Results: The clustering algorithm was able to separate clusters which significantly reduced in voxel number following induction chemotherapy from clusters with a non-significant reduction. Partitioning with the optimal number of clusters (k = 4), determined with cluster validation, produced the best separation between reducing and non-reducing clusters.

Conclusion: The proposed methodology was able to identify tumour sub-regions with distinct functional properties, independently separating clusters which were affected differently by treatment. This work demonstrates that unsupervised cluster analysis, with no prior knowledge of the data, can be employed to provide a multi-parametric characterization of functional heterogeneity within tumour volumes.

No MeSH data available.


Related in: MedlinePlus

Slice with the largest tumour cross-section of a representative primary tumour.(a) ROI (red) contoured on the most enhanced frame of the DCE series, (b) ROI (blue) contoured on b = 1000 image of the DWI series, (c) IAUG60 map, (d) ADC map, with the two ROIs overlapped, and (e) cluster analysis map pre and post-induction chemotherapy (IC), in the simplest case of k = 2. The red ROI also encompasses the region where cluster analysis was performed; the blue ROIs represent the area of restricted diffusion independently contoured on DWI data, almost coincident with Cluster 1, which is characterized by both low ADC and high IAUGC60 values.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4580650&req=5

pone.0138545.g001: Slice with the largest tumour cross-section of a representative primary tumour.(a) ROI (red) contoured on the most enhanced frame of the DCE series, (b) ROI (blue) contoured on b = 1000 image of the DWI series, (c) IAUG60 map, (d) ADC map, with the two ROIs overlapped, and (e) cluster analysis map pre and post-induction chemotherapy (IC), in the simplest case of k = 2. The red ROI also encompasses the region where cluster analysis was performed; the blue ROIs represent the area of restricted diffusion independently contoured on DWI data, almost coincident with Cluster 1, which is characterized by both low ADC and high IAUGC60 values.

Mentions: The ROIs contouring the area of enhancement on DCE data were employed in the analysis. For large PTs (an example in Fig 1) this area is larger than the core of restricted diffusion contoured on DWI data (a total of 970 and 703 voxels for all central ROIs at baseline, for DCE and DWI respectively), while for the LNs, which are generally encapsulated structures, there is minimal difference in size between DCE and DWI ROIs (a total of 475 and 490 voxels for all central ROIs at baseline, for DCE and DWI respectively).


Characterizing Heterogeneity within Head and Neck Lesions Using Cluster Analysis of Multi-Parametric MRI Data.

Borri M, Schmidt MA, Powell C, Koh DM, Riddell AM, Partridge M, Bhide SA, Nutting CM, Harrington KJ, Newbold KL, Leach MO - PLoS ONE (2015)

Slice with the largest tumour cross-section of a representative primary tumour.(a) ROI (red) contoured on the most enhanced frame of the DCE series, (b) ROI (blue) contoured on b = 1000 image of the DWI series, (c) IAUG60 map, (d) ADC map, with the two ROIs overlapped, and (e) cluster analysis map pre and post-induction chemotherapy (IC), in the simplest case of k = 2. The red ROI also encompasses the region where cluster analysis was performed; the blue ROIs represent the area of restricted diffusion independently contoured on DWI data, almost coincident with Cluster 1, which is characterized by both low ADC and high IAUGC60 values.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4580650&req=5

pone.0138545.g001: Slice with the largest tumour cross-section of a representative primary tumour.(a) ROI (red) contoured on the most enhanced frame of the DCE series, (b) ROI (blue) contoured on b = 1000 image of the DWI series, (c) IAUG60 map, (d) ADC map, with the two ROIs overlapped, and (e) cluster analysis map pre and post-induction chemotherapy (IC), in the simplest case of k = 2. The red ROI also encompasses the region where cluster analysis was performed; the blue ROIs represent the area of restricted diffusion independently contoured on DWI data, almost coincident with Cluster 1, which is characterized by both low ADC and high IAUGC60 values.
Mentions: The ROIs contouring the area of enhancement on DCE data were employed in the analysis. For large PTs (an example in Fig 1) this area is larger than the core of restricted diffusion contoured on DWI data (a total of 970 and 703 voxels for all central ROIs at baseline, for DCE and DWI respectively), while for the LNs, which are generally encapsulated structures, there is minimal difference in size between DCE and DWI ROIs (a total of 475 and 490 voxels for all central ROIs at baseline, for DCE and DWI respectively).

Bottom Line: Cumulative distributions of voxels, containing pre and post-treatment data and including both primary tumours and lymph nodes, were partitioned into k clusters (k = 2, 3 or 4).Principal component analysis and cluster validation were employed to investigate data composition and to independently determine the optimal number of clusters.Partitioning with the optimal number of clusters (k = 4), determined with cluster validation, produced the best separation between reducing and non-reducing clusters.

View Article: PubMed Central - PubMed

Affiliation: CR-UK Cancer Imaging Centre, The Royal Marsden NHS Foundation Trust and The Institute of Cancer Research, London, United Kingdom.

ABSTRACT

Purpose: To describe a methodology, based on cluster analysis, to partition multi-parametric functional imaging data into groups (or clusters) of similar functional characteristics, with the aim of characterizing functional heterogeneity within head and neck tumour volumes. To evaluate the performance of the proposed approach on a set of longitudinal MRI data, analysing the evolution of the obtained sub-sets with treatment.

Material and methods: The cluster analysis workflow was applied to a combination of dynamic contrast-enhanced and diffusion-weighted imaging MRI data from a cohort of squamous cell carcinoma of the head and neck patients. Cumulative distributions of voxels, containing pre and post-treatment data and including both primary tumours and lymph nodes, were partitioned into k clusters (k = 2, 3 or 4). Principal component analysis and cluster validation were employed to investigate data composition and to independently determine the optimal number of clusters. The evolution of the resulting sub-regions with induction chemotherapy treatment was assessed relative to the number of clusters.

Results: The clustering algorithm was able to separate clusters which significantly reduced in voxel number following induction chemotherapy from clusters with a non-significant reduction. Partitioning with the optimal number of clusters (k = 4), determined with cluster validation, produced the best separation between reducing and non-reducing clusters.

Conclusion: The proposed methodology was able to identify tumour sub-regions with distinct functional properties, independently separating clusters which were affected differently by treatment. This work demonstrates that unsupervised cluster analysis, with no prior knowledge of the data, can be employed to provide a multi-parametric characterization of functional heterogeneity within tumour volumes.

No MeSH data available.


Related in: MedlinePlus