IAES International Journal of Artificial Intelligence (IJ-AI) Vol. 13, No. 3, September 2024, pp. 3339~3349 ISSN: 2252-8938, DOI: 10.11591/ijai.v13.i3.pp3339-3349  3339 Journal homepage: http://ijai.iaescore.com SANAS-Net: spatial attention neural architecture search for breast cancer detection Melwin D'souza 1 , Ananth Prabhu Gurpur 1 , Varuna Kumara 2 1 Department of Computer Science and Engineering, Sahyadri College of Engineering and Management, Mangaluru, India 2 Department of Electronics and Communication Engineering, Moodlakatte Institute of Technology, Kundapura, India Article Info ABSTRACT Article history: Received Nov 16, 2023 Revised Feb 13, 2024 Accepted Feb 28, 2024 The utilization of mammography images plays a vital role in the prompt detection and treatment of breast cancer. Breast imaging techniques aid medical professionals in assessing the dimensions, morphology, and spatial orientation of breast lesions, facilitating the differentiation between benign and malignant conditions. Breast tissue can vary widely in terms of density, composition, and structure, leading to complexities in distinguishing between benign and malignant conditions. The primary contribution of this paper is the proposal of a spatial attention-based neural architecture search network (SANAS-Net) technique that incorporates a spatial attention mechanism, enabling the model to learn and prioritize key regions within mammograms (MMs). Multi-head attention is employed within the transformer blocks to effectively capture a wide range of spatial relations and feature interactions. Global contextual information was integrated into the transformer blocks by means of introducing positional embeddings. Several practical studies have been undertaken to verify the effectiveness of our methodology in identifying fully attentive networks that exhibit good performance in distinguishing between malignant and benign breast cancer cases. The experimental study reached a test accuracy of 89.95%, which is way higher than previously proposed algorithms for mammography image- based breast cancer detection. Keywords: Breast cancer detection Deep learning Mammography Neural architecture search Spatial attention This is an open access article under the CC BY-SA license. Corresponding Author: Melwin D'souza Department of Computer Science and Engineering, Sahyadri College of Engineering and Management Mangaluru, India Email: mellumerdy@gmail.com 1. INTRODUCTION Breast cancer is a prevalent and highly lethal malignancy that affects women on a global scale. The condition has the potential to impact females across all age groups, although the likelihood of occurrence escalates with advancing age and in the presence of specific genetic and environmental influences. The timely identification of breast cancer is a crucial measure in mitigating its advancement and enhancing the likelihood of effective intervention and long-term survival. Women can effectively discover any changes or abnormalities in their breasts promptly by engaging in routine self-examinations, undergoing mammography and clinical breast examinations, and maintaining awareness of the signs and symptoms associated with breast cancer [1]. The timely identification of a medical condition can also contribute to a decrease in the necessity for more intrusive and forceful treatment methods, such as chemotherapy, radiation therapy, or mastectomy. Hence, it is imperative that women get education and empowerment to assume responsibility for their breast health and promptly seek medical intervention upon detecting any potentially concerning alterations. The identification of breast cancer at an early stage is not only a pivotal measure for managing