SVM-based classification of breast tumour phantoms using a UWB radar prototype system

Conceição, Raquel C.; Medeiros, Hugo; O’Halloran, Martin; Rodriguez-Herrera, Diego; Flores-Tapia, Daniel; Pistorius, Stephen

doi:10.1109/ursigass.2014.6930131

Cited by 11 publications

(11 citation statements)

References 11 publications

(26 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Benign tumours are roughly elliptical and usually have well circumscribed margins, and malignant tumours have irregular shapes and are surrounded by a radiating pattern of spikes, commonly referred to as spicules [ 22 , 23 , 24 , 25 , 26 ]. Previous studies have already shown how microwave backscattered signals may change if tumours of different sizes or shapes are present within the breast [ 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 ]. These studies have also demonstrated that classification and machine learning algorithms are able to learn from the shape differences in backscattered signals, albeit in relatively simple datasets.…”

Section: Introductionmentioning

confidence: 99%

“…Experimental datasets have also been used to assess the performance of diagnosis systems, namely by using principal component analysis in combination with support vector machines, linear discriminant analysis and quadratic discriminant analysis. In [ 36 , 37 ], tumour phantoms with various sizes and shapes were immersed in a breast phantom with dielectric properties matching those of adipose tissue. Importantly, the experimental results presented in these studies are in general agreement with previous numerical data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Diagnosing Breast Cancer with Microwave Technology: Remaining Challenges and Potential Solutions with Machine Learning

et al. 2018

View full text Add to dashboard Cite

Currently, breast cancer often requires invasive biopsies for diagnosis, motivating researchers to design and develop non-invasive and automated diagnosis systems. Recent microwave breast imaging studies have shown how backscattered signals carry relevant information about the shape of a tumour, and tumour shape is often used with current imaging modalities to assess malignancy. This paper presents a comprehensive analysis of microwave breast diagnosis systems which use machine learning to learn characteristics of benign and malignant tumours. The state-of-the-art, the main challenges still to overcome and potential solutions are outlined. Specifically, this work investigates the benefit of signal pre-processing on diagnostic performance, and proposes a new set of extracted features that capture the tumour shape information embedded in a signal. This work also investigates if a relationship exists between the antenna topology in a microwave system and diagnostic performance. Finally, a careful machine learning validation methodology is implemented to guarantee the robustness of the results and the accuracy of performance evaluation.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Diagnosing Breast Cancer with Microwave Technology: Remaining Challenges and Potential Solutions with Machine Learning

et al. 2018

View full text Add to dashboard Cite

show abstract

“…The latter three classifiers -NB, DT, and kNNwere used to assess all signals, expanding and complementing prior studies. 25,26 These classifiers are well-suited to analyse data such as those collected in this study: continuous time domain signals with labeled data. For example, NB is based on probabilities, DT is based on decision analysis and kNN is based on pattern recognition.…”

Section: B3 Classification Algorithmsmentioning

confidence: 99%

“…To the best of our knowledge, this was the first study to report experimental results using a MWI prototype system for tumor classification based on tumor shape. The study was later expanded to include SVM as a classification algorithm which, despite the good accuracy results, required high computational power in order to provide an efficient hyperparameters optimization.…”

Section: Introductionmentioning

confidence: 99%

Classification of breast tumor models with a prototype microwave imaging system

et al. 2020

Self Cite

View full text Add to dashboard Cite

Purpose: The assessment of the size and shape of breast tumors is of utter importance to the correct diagnosis and staging of breast cancer. In this paper, we classify breast tumor models of varying sizes and shapes using signals collected with a monostatic ultra-wideband radar microwave imaging prototype system with machine learning algorithms specifically tailored to the collected data. Methods: A database comprising 13 benign and 13 malignant tumor models with sizes between 13 and 40 mm was created using dielectrically representative tissue mimicking materials. These tumor models were placed inside two breast phantoms: a homogeneous breast phantom and a breast phantom with clusters of fibroglandular mimicking tissue, accounting for breast heterogeneity. The breast phantoms with tumors were imaged with a monostatic microwave imaging prototype system, over a 1-6 GHz frequency range. The classification of benign and malignant tumors embedded in the two breast phantoms was completed, and tumor classification was evaluated with Principal Component Analysis as a feature extraction method, and tuned Na€ ıve Bayes (NB), decision trees (DT), and knearest neighbours (kNN) as classifiers. We further study which antenna positions are better placed to classify tumors, discuss the feature extraction method and optimize classification algorithms, by tuning their hyperparameters, to improve sensitivity, specificity and the receiver operating characteristic curve, while ensuring maximum generalization and avoiding overfitting and data contamination. We also added a realistic synthetic skin response to the collected signals and examined its global effect on classification of benign vs malignant tumors. Results: In terms of global classification performance, kNN outperformed DT and NB machine learning classifiers, achieving a classification accuracy of 96.2% when classifying between benign and malignant tumor phantoms in a homogeneous breast phantom (both when the skin artifact is and is not considered). Conclusions:We experimentally classified tumor models as benign or malignant with a microwave imaging system, and we showed a methodology that can potentially assess the shape of breast tumors, which will give further insight into the correct diagnosis and staging of breast cancer.

show abstract

“…Although clinical trials were performed by several groups Meaney et al, 2000;Porter et al, 2016), problems-such as the difference between the dielectric properties between tumor and glandular tissues being less than that in the existing data (Bourqui, Garrett, et al, 2012;J. Garrett & Fear, 2014; J. D. Garrett & Fear, 2015;Winters et al, 2006) and classification-remain (Conceição et al, 2014;McGinley et al, 2010;O'Halloran et al, 2011;O'Loughlin et al, 2016).…”

Section: Introductionmentioning

confidence: 99%

Experimental Demonstration of In‐Place Calibration for Time Domain Microwave Imaging System

Kwon

Lee

2018

Radio Science

View full text Add to dashboard Cite

In this study, the experimental demonstration of in‐place calibration was conducted using the developed time domain measurement system. Experiments were conducted using three calibration methods—in‐place calibration and two existing calibrations, that is, array rotation and differential calibration. The in‐place calibration uses dual receivers located at an equal distance from the transmitter. The received signals at the dual receivers contain similar unwanted signals, that is, the directly received signal and antenna coupling. In contrast to the simulations, the antennas are not perfectly matched and there might be unexpected environmental errors. Thus, we experimented with the developed experimental system to demonstrate the proposed method. The possible problems with low signal‐to‐noise ratio and clock jitter, which may exist in time domain systems, were rectified by averaging repeatedly measured signals. The tumor was successfully detected using the three calibration methods according to the experimental results. The cross correlation was calculated using the reconstructed image of the ideal differential calibration for a quantitative comparison between the existing rotation calibration and the proposed in‐place calibration. The mean value of cross correlation between the in‐place calibration and ideal differential calibration was 0.80, and the mean value of cross correlation of the rotation calibration was 0.55. Furthermore, the results of simulation were compared with the experimental results to verify the in‐place calibration method. A quantitative analysis was also performed, and the experimental results show a tendency similar to the simulation.

show abstract

SVM-based classification of breast tumour phantoms using a UWB radar prototype system

Cited by 11 publications

References 11 publications

Diagnosing Breast Cancer with Microwave Technology: Remaining Challenges and Potential Solutions with Machine Learning

Diagnosing Breast Cancer with Microwave Technology: Remaining Challenges and Potential Solutions with Machine Learning

Classification of breast tumor models with a prototype microwave imaging system

Experimental Demonstration of In‐Place Calibration for Time Domain Microwave Imaging System

Contact Info

Product

Resources

About