Brain tumour classification: a comprehensive systematic review on various constraints

Abstract

ABSTRACT Brain tumour classification is among the most challenging tasks in medical image analysis. The purpose of brain tumour classification is to produce a precise delineation of tumour areas. Successful early detection of brain tumours is essential to enhancing treatment outcomes and thus improving patient survival. Over the last several decades, many researchers have contributed significantly to the field of brain tumour classification. The level of user supervision and computing efficiency have both been factors in the clinical acceptability of classification systems. However, due to a lack of collaboration between doctors and researchers, practical applications are still limited, and clinicians still rely on manual tumour estimates. This research presents a comprehensive review of recent brain tumour classification methods. The purpose of this survey is to conduct an evaluation of 70 papers that deal with the classification of brain tumours. This survey includes a systematic analysis based on performance levels and related maximum achievements for each contribution. In addition, the chronological review and numerous implemented tools used in the evaluated works are examined. Finally, the survey highlights a number of research issues and flaws that researchers may find helpful in developing prospective studies on brain tumour classification.