Preliminary assessment of an Artificial Intelligence algorithm based on MRI breast modelling with US fusion

Congress:

ECR 2020

Poster Number:

C-07119

Type:

Scientific Exhibit

Keywords:

Multicentre study, Experimental, Not applicable, Image verification, Image registration, Localisation, Imaging sequences, Computer Applications-Detection, diagnosis, Ultrasound, MR, Image manipulation / Reconstruction, Breast, Artificial Intelligence

Authors:

S. de Beni¹, S. D'Onofrio¹, M. Cereseto¹, A. Abate², R. Giovanazzi³, R. Corso³, A. martegani⁴, C. FACHINETTI⁴, G. Sakas⁵; ¹Esaote/IT, ²Lesmo/IT, ³Monza/IT, ⁴Como/IT, ⁵Darmstadt/DE

DOI:

10.26044/ecr2020/C-07119

DOI-Link:

https://dx.doi.org/10.26044/ecr2020/C-07119

Fig. 1: Clinical FI examination on US system with patient in supine position

Fig. 7: In vivo test of BreastNav FI technology, minimal mismatch assessed,...

Fig. 11: US Shear Wave Elastography evaluation (QElaXto 2D, Esaote) together with...

Purpose

EvaluationofanewUltrasound(US)technology,based on Artificial Intelligence (AI) algorithm, thatcorrelates and fusesMRI-prone withUS-supine BreastImaging (BI) examinations. In-vitroandin-vivotestswereperformed.

Methods and materials

Breast cancer represents the most common oncological disease for women and a multimodality approach, with a clinical exam and an advanced set of BI examinations, is mandatory as current medical standard. MRI and Contrast-enhanced MRI (CE-MRI) play a central role in BI, due to their very high sensitivity. In addition, technology evolution over the recent years has made breast US an essential component of the BI evaluation. In current praxis, when lesions detected on MRI are not visible/equivocal on mammogram and/or are not identified by...

Results

For in-vitro test, the mismatchonthereference amorphous lesion identified in prone MRI andtheonevisualizedinreal-time US, calculated between the real-time US probe position and the MRI set target,wasminimal (Fig.6). No further investigations or statistical analysis has been performed, since the phantom deformation is limited, compared to real tissue, therefore phantoms are not an adequate representative of the real breast organ deformation. Forin-vivotests,2 different investigation scenarios were performed. The first investigation was made to evaluate the BreastNav® feasibility and its AI algorithm performance on real patients. In a previous...

Conclusion

Integrated breast multimodal FI is starting to become possible, thanks to technology evolution. BreastNavtechnologyisaimedtospeedupthedetectionand characterization oflesions,visibleonly onMRI, also onreal-time US, thankstotheFIbetweenthetwo differentBI modalities and to manage the deformation induced by patient’s different positions between US and MRI, within clinically acceptable localization errors.

Personal information and conflict of interest

S. D'onofrio; Esaote/IT - Author at Esaote SpA - Employee at Esaote SpA S. de Beni; Esaote/IT - Author at Esaote SpA - Employee at Esaote SpA M. Cereseto; Esaote/IT - Author at Esaote SpA - Employee at Esaote SpA A. Abate; Lesmo/IT - Author at ASST Monza R. Giovanazzi; Monza/IT - Author at ASST Monza R. Corso; Monza/IT - Author at ASST Monza A. Martegani; Como/IT - Author at Ospedale Valduce (Como, IT) C. Fachinetti; Como/IT - Author at Ospedale Valduce (Como, IT) G....

References

Di Grezia et al. “State-of-the-Art in Integrated Breast Imaging”, Biomed Res Int. 2019; 2019:7596059, doi:10.1155/2019/7596059 Mann R.M. et al. “Breast MRI: State of the Art”, Radiology 2019; 292: 520-36,https://doi.org/10.1148/radiol.2019182947 Abate A. et al., “Tecnica Di Fusione Delle Immagini Risonanza Magnetica-Ecografia “Real Time” Per La Caratterizzazione Delle Lesioni Mammarie”, Poster AIS P13.1-2019 Mazzei M.A. et al. “Efficacy of Second-Look Ultrasound with MR Coregistration for Evaluating Additional Enhancing Lesions of the Breast: Review of the Literature”, Biomed Res Int. 2018; 2018:3896946,https://doi.org/10.1155/2018/3896946 Aribal E, et al. “Volume navigation...