Real-Time Object Detector for Medical Diagnostics (RTMDet): A High-Performance Deep Learning Model for Brain Tumor Diagnosis

• Published in: Bioengineering (Basel) Vol. 12; no. 3; p. 274
• Main Authors: Bakhtiyorov, Sanjar, Umirzakova, Sabina, Musaev, Musabek, Abdusalomov, Akmalbek, Whangbo, Taeg Keun
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 11.03.2025; MDPI
• Subjects: Accuracy; Artificial intelligence; Artificial neural networks; Automation; Brain; Brain cancer; Brain research; brain tumor detection; Brain tumors; Care and treatment; Classification; Clinical outcomes; computational efficiency; Datasets; Deep learning; Design; Detectors; Diagnosis; Diagnostic equipment (Medical); Efficiency; Innovations; Machine learning; medical diagnostics; Medical imaging; Neural networks; neuro-oncology; Neuroimaging; Oncology; Optimization algorithms; Optimization techniques; Patients; Real time; Tumors; computational efficiency; medical diagnostics; neuro-oncology; brain tumor detection
• ISSN: 2306-5354; 2306-5354
• DOI: 10.3390/bioengineering12030274

Background: Brain tumor diagnosis requires precise and timely detection, which directly impacts treatment decisions and patient outcomes. The integration of deep learning technologies in medical diagnostics has improved the accuracy and efficiency of these processes, yet real-time processing remains a challenge due to the computational intensity of current models. This study introduces the Real-Time Object Detector for Medical Diagnostics (RTMDet), which aims to address these limitations by optimizing convolutional neural network (CNN) architectures for enhanced speed and accuracy. Methods: The RTMDet model incorporates novel depthwise convolutional blocks designed to reduce computational load while maintaining diagnostic precision. The effectiveness of the RTMDet was evaluated through extensive testing against traditional and modern CNN architectures using comprehensive medical imaging datasets, with a focus on real-time processing capabilities. Results: The RTMDet demonstrated superior performance in detecting brain tumors, achieving higher accuracy and speed compared to existing CNN models. The model’s efficiency was validated through its ability to process large datasets in real time without sacrificing the accuracy required for a reliable diagnosis. Conclusions: The RTMDet represents a significant advancement in the application of deep learning technologies to medical diagnostics. By optimizing the balance between computational efficiency and diagnostic precision, the RTMDet enhances the capabilities of medical imaging, potentially improving patient outcomes through faster and more accurate brain tumor detection. This model offers a promising solution for clinical settings where rapid and precise diagnostics are critical.