Journal: Computer Methods and Programs in Biomedicine 

Autori: Andrea Berti (a, b, c), Camilla Scapicchio (b, d), Chiara Iacconi (e), Charlotte Marguerite Lucille Trombadori (f), Maria Evelina Fantacci (b, d), Alessandra Retico (b), Sara Colantonio (a)

Affiliazioni:

a) Institute of Information Science and Technologies (ISTI) - National Research Council of Italy (CNR), Via Giuseppe Moruzzi, 1, Pisa, 56127, Italy, IT
b) National Institute for Nuclear Physics (INFN) section of Pisa, Largo Bruno Pontecorvo, 3/Edificio C, Pisa, 56127, Italy, IT
c) Department of Information Engineering, University of Pisa, Via Girolamo Caruso, 16, Pisa, 56122, Italy, IT
d) Department of Physics, University of Pisa, Largo Bruno Pontecorvo, 3, Pisa, 56127, Italy, IT
e) Department of radiology, breast imaging UOSD, Piazza Monzoni 1, Carrara, 54033, Italy, IT
f) Department of Diagnostic Imaging, Oncological Radiotherapy and Haematology, Fondazione Policlinico Universitario A. Gemelli IRCCS, Largo Agostino Gemelli, 8, Rome, 00168, Italy, IT

Abstract:

Background and Objective:
Breast cancer is the most common cancer among women worldwide, making early detection through breast screening crucial for improving patient outcomes. Digital Breast Tomosynthesis (DBT) is an advanced radiographic technique that enhances clarity over traditional mammography by compiling multiple X-ray images into a 3D reconstruction, thereby improving cancer detection rates. However, the large volume of data generated by DBT poses a challenge for timely analysis. This study aims to introduce a transparent AI system that expedites the analysis of DBT scans while ensuring interpretability.

Methods:
The study employs a two-stage deep learning process. The first stage uses state-of-the-art Neural Network (NN) models, specifically YOLOv5 and YOLOv8, to detect lesions within the scans. An ensemble method is also explored to enhance detection capabilities. The second stage involves classifying the identified lesions using ProtoPNet, which leverages prototypical part learning to distinguish between benign and cancerous lesions. The system is designed to facilitate clear interpretability in decision-making, which is crucial for medical diagnostics.

Results:
The AI system's performance, validated by expert radiologists, demonstrates competitive recall rates. The evaluation metrics include precision, sensitivity, and specificity values, highlighting the system's potential for future clinical relevance. Despite challenges such as dataset limitations and the need for more accurate ground truth annotations, the approach shows significant advancement in applying AI to DBT scans.

Conclusions:
This study contributes to the growing field of AI in breast cancer screening by emphasizing the need for systems that are not only accurate but also transparent and interpretable. The proposed AI system marks a significant step forward in the timely and accurate analysis of DBT scans, with potential implications for improving early breast cancer detection and patient outcomes.