Micrima announces 2 poster presentations at ECR 2017

1. Automatic labelling of lesions using radiofrequency feature discrimination
Authors N. Ridley1, A. Iriarte2, L. Tsui2, C. Bore2, M. Shere2, I. Lyburn3, P. Bannister2; 1Swindon/UK, 2Bristol/UK, 3Cheltenham/UK

Aims and objectives
The MARIA™ system is a novel CE­marked breast imaging system that exploits the scattering contrast between breast tissue types across a 5GHz spectrum to detect lesions in both dense and lucent tissue. This early work demonstrates the potential of whole­breast radiofrequency imaging to automatically label lesions according to their clinical type based on intrinsic dielectric properties measured in­vivo.

Methods and materials
Patients attending symptomatic clinics at 3 sites were identified by clinicians as having a palpable lump. Following informed consent eligible patients underwent this prone imaging technique. Bilateral radiofrequency data was recorded along with clinical information and other imaging studies including ultrasound and/ or mammography and, when relevant, core biopsy results to determine lesion type. 36 studies exhibiting unifocal lesions with a histological diagnosis of malignant (Ca) or benign (Fa, cyst) were analysed by plotting their average radiofrequency response up to 6.3Ghz against the response calculated at higher frequencies.

Results
The evaluated studies comprised 16 Ca, 13 cyst and 7 Fa. Based on the clustering of datasets, it can be seen that the analysis groups Ca into a well­behaved region according to its radiofrequency response characteristics. These cases are generally welldiscriminated from the benign cases with a PPV for Ca of 76.19%.Plot of average banded radio­frequency response (high vs low) for clinical studies scanned using MARIA™.

Conclusion
While further work is needed to understand the behaviour of sub­types within the benign cohort, this analysis builds on previous ex­vivo studies which indicated that lesions can be automatically classified based on their frequency­dependent dielectric response.

2. Cancer detection in dense tissue using radiofrequency imaging ­ a clincial evaluation
Authors N. Ridley1, M. Shere2, I. Lyburn3, P. Bannister2; 1Swindon/UK, 2Bristol/UK, 3Cheltenham/UK

Aims and objectives
As part of a multi­center symptomatic trial of the CE Marked MARIA™. radiofrequency breast imaging system, detection performance was analysed for cases where the patient exhibited dense tissue and where a diagnosis of cancer was confirmed by histology.

Methods and materials
Patients attending symptomatic clinics at 3 sites were identified by clinicians as having a palpable lump. Following informed consent eligible patients underwent this prone imaging technique. The bilateral reconstructed 3D images were correlated with clinical information and other imaging studies including ultrasound, mammography and core biopsy results to determine sensitivity scores.

Results
Of 21 cases evaluated that were graded either BI­RAD c or d for breast density and where biopsy confirmed a diagnosis of cancer, a sensitivity of 90% (19/21) was obtained for lesion detection using MARIA™ (mean age 55.8, age range 35­80).

Conclusion
Initial results indicate the potential that the MARIA™ system offers as a well­tolerated non­ionising imaging modality that has been shown to be effective at detecting cancers in women with dense breasts. MARIA™ may contribute to overcoming some of the challenges posed by trying to optimise the balance between benefit and harm of screening in women of younger age or who otherwise exhibit dense breast tissue. The results of this study are very encouraging. We are continuing to evaluate this exciting novel imaging technique.

Share

  • Latest news

  • WordPress Image Lightbox Plugin