Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM69008
Volume:
10 |
11 |
Nov 2026
Pages:
1-3
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Computer science engineering and (artificial intelligence and machine learning )
CMR technical campus
Computer science engineering and (artificial intelligence and machine learning )
CMR technical campus
Computer science engineering and (artificial intelligence and machine learning )
CMR technical campus
Computer science engineering and (artificial intelligence and machine learning )
CMR technical campus
Bone tumors are uncommon but serious, and an early, correct reading of the radiograph strongly affects treatment. Reading X-rays by hand is slow, depends on the radiologist's experience, and small or low-contrast lesions are easy to miss, especially where specialists are scarce. This paper presents an automated bone tumor detection framework built on YOLOv11, a single-stage, anchor-free object detector. Given a plain radiograph, the model draws a bounding box around each suspected lesion and labels it benign or malignant, with a confidence score, in one forward pass. We fine-tune YOLOv11 on a public annotated bone tumor radiograph dataset, apply radiograph-specific preprocessing and augmentation, and compare five YOLOv11 variants (n, s, m, l, x) against YOLOv8, Faster R-CNN and RT-DETR baselines. On the held-out test split, the best configuration reaches a precision of 94.6 %, recall of 94.6%, [email protected] of 94.6% and [email protected]:0.95 of94.6%, at18 ms per image.
KEYWORDS
This paper presented a YOLOv11-based framework that detects bone tumors in X-ray radiographs, localising each lesion and labelling it benign or malignant in one pass. The best configuration reached 94.6% [email protected] and 94.6% recall at 18ms per image [compare with baselines once measured]. Future work includes multi-centre external validation, segmentation for exact tumor boundaries, combining radiographs with clinical data such as age and lesion site, uncertainty estimation with referral of doubtful cases to a radiologist, explanation maps for clinical trust, and edge deployment with a prospective hospital trial.
[1] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You only look once: Unified, real-time object detection,” in Proc. IEEE CVPR, 2016, pp. 779–788. [2] R. Khanam and M. Hussain, “YOLOv11: An overview of the key architectural enhancements,” arXiv:2410.17725, 2024. [3] S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards real-time object detection with region proposal networks,” in Proc. NeurIPS, 2015. [4] T.-Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollár, “Focal loss for dense object detection,” in Proc. IEEE ICCV, 2017, pp. 2980–2988. [5] N. Carion et al., “End-to-end object detection with transformers,” in Proc. ECCV, 2020. [6] G. Jocher, A. Chaurasia, and J. Qiu, “Ultralytics YOLOv8,” software, 2023. [7] G. Jocher and J. Qiu, “Ultralytics YOLO11,” software, 2024. Available: https://github.com/ultralytics/ultralytics [8] S. Yao et al., “A radiograph dataset for the classification, localization, and segmentation of primary bone tumors,” Sci. Data, vol. 12, 2025, doi: 10.1038/s41597-024-04311-y.
Publication Date
11/16/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
2
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...