Biomedical Signal Processing and Control 87 (2024) 105474 Available online 20 September 2023 1746-8094/© 2023 Elsevier Ltd. All rights reserved. Gradient bald vulture optimization enabled multi-objective Unet++ with DCNN for prostate cancer segmentation and detection Jayashree Rajesh Prasad a, * , Rajesh Shardanand Prasad a , Amol Dhumane d , Nihar Ranjan c , Mubin Tamboli b a Computer Science and Engineering, School of Computing, MIT Art Design and Technology University, Pune, Pune, Maharashtra 412201, India b Computer Engineering, Pimpri Chinchwad College of Engineering, Pune, Maharashtra, India c Information Technology, JSPM’s Rajarshi Shahu College of Engineering, Pune, Maharashtra, India d Computer Engineering, Symbiosis Institute of Technology, Pune, Maharashtra, India A R T I C L E INFO Keywords: Prostate cancer Deep convolution neural network U-Net++ Non-local means flter MRI ABSTRACT Prostate cancer (PCa) represents the general type of cancer and is considered the third leading reason of death worldwide. As a combined part of computer-aided detection (CAD) applications, magnetic resonance imaging (MRI) is extensively studied for the precise detection of PCa. However, various issues rely on MRI, which includes the complexity of interpretation and increased time. Thus, deep learning-based tumor detection and segmen- tation methods attempt to be imperative techniques for radiologists to execute their tasks more precisely. The objective is to present Gradient Bald vulture optimization (GBVO)-based Deep Convolution Neural Networks (DCNN) with U-Net++ for segmenting and detecting prostate cancer. Initially, image pre-processing is done using Non-Local Means (NLM) flter. After pre-processing, image segmentation is carried out using the proposed optimized multi-objective Unet++, where the objective function in Unet++ is modifed using pixel-wise cross entropy and the Jaccard coeffcient. In addition, the training of Unet++ is done using the newly designed Gradient Bald Eagle Optimization (GBEO), which is a combination of Stochastic Gradient Descent (SGD) and Bald Eagle optimization (BEO). Finally, cancer detection is done using DCNN. DCNN is trained using GBVO, which is the integration of the African Vultures Optimization Algorithm (AVOA) and GBEO. The proposed method out- performed state-of-the-art techniques. The developed method achieved the highest accuracy of 0.916, a false negative ratio (FNR) of 0.104, a false positive ratio (FPR) of 0.100, and a negative predictive value (NPV) of 0.903. 1. Introduction PCa is termed a major reason for cancer death among men in the contemporary world. PCa represents the frequent appearance of cancer among males in the United States. In 2017, it is considered the third leading reason of death from cancer with 161,360 new cases [1]. PCa represents the second most general malignancy in men throughout the world that accounts for 1,276,106 new cases and leading to 358,989 deaths in 2018 [23]. Although prostate cancer represents the most common disease, it can be treated if detected earlier. The survival rate becomes high if treated early because of a slow evolution of the disease [4]. Thus, effectual observation and earlier detection represent the key to enhanced survival of patients. MRI is considered an effective imaging modality that is extensively utilized for discovering prostate cancer. The Multiparametric MRI depends on Diffusion-Weighted Imaging (DWI), which is extensively becoming the benchmark of prostate cancer diag- nosis in radiology wherein the region under the Receiver Operating Characteristic Curve (ROC) alters from 0.69 to 0.81 for radiologists discovering PCa [5]. The most common technique to interpret images is known as Prostate Imaging-Reporting and Data System (PI-RADS), which is devised for radiologists, but the remaining problems with inter- observer unpredictability consider the PI-RADS scheme [6]. The issue of treating PCa represents how to discover and differentiate indolent PCa using clinically tested PCa [7]. Tesla-based multi-parametric MRI (3 T mp-MRI) [8] offers a strong integration of functional and anatomical data of PCa and acquires a major role in diagnosing the PCa by mini- mizing the redundant biopsies [4] and adding diagnosis options in vigorous surveillance and focal treatment [910]. * Corresponding author. E-mail address: jayashree.prasad@mituniversity.edu.in (J. Rajesh Prasad). Contents lists available at ScienceDirect Biomedical Signal Processing and Control journal homepage: www.elsevier.com/locate/bspc https://doi.org/10.1016/j.bspc.2023.105474 Received 26 December 2022; Received in revised form 26 August 2023; Accepted 12 September 2023