Previously submitted to: JMIR Medical Informatics (no longer under consideration since Sep 18, 2026)
Date Submitted: Sep 17, 2026
Open Peer Review Period: Sep 18, 2026 - Sep 18, 2026
(closed for review but you can still tweet)
NOTE: This is an unreviewed Preprint
Warning: This is a unreviewed preprint (What is a preprint?). Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn (a note "no longer under consideration" will appear above).
Peer review me: Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period (in this case, a "Peer Review Me" button to sign up as reviewer is displayed above). All preprints currently open for review are listed here. Outside of the formal open peer-review period we encourage you to tweet about the preprint.
Citation: Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).
Final version: If our system detects a final peer-reviewed "version of record" (VoR) published in any journal, a link to that VoR will appear below. Readers are then encourage to cite the VoR instead of this preprint.
Settings: If you are the author, you can login and change the preprint display settings, but the preprint URL/DOI is supposed to be stable and citable, so it should not be removed once posted.
Submit: To post your own preprint, simply submit to any JMIR journal, and choose the appropriate settings to expose your submitted version as preprint.
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Title: AN EDGE-DEPLOYABLE LIGHTWEIGHT DEEP LEARNING DIAGNOSTIC FRAMEWORK FOR MULTI-CLASS LUNG CANCER CLASSIFICATION AND STAGING Abstract: Lung cancer is one of the main causes of cancer-related deaths worldwide and needs to be diagnosed in its early and clinical stages with the help of accurate and efficient diagnostic systems that can be used in clinical settings. In this study, a lightweight deep learning framework, called LiteSOS-Net, is proposed to classify lung cancer types and predict the TNM stage from CT images simultaneously. The proposed pipeline combines pre-processing, Attention U-Net segmentation, and a ShuffleNet-based classification backbone with the Osprey algorithm for efficient feature learning. The results of a substanti
ABSTRACT
Background:
Lung cancer is one of the leading causes of cancer-related deaths worldwide. Early and accurate diagnosis, along with reliable clinical-stage assessment, is essential for effective treatment planning and improved patient outcomes. However, conventional diagnostic approaches can be time-consuming and may have limitations in accurately classifying cancer types and stages. Therefore, there is a need for efficient and lightweight computer-aided diagnostic systems that can support clinical decision-making.
Objective:
To develop a lightweight deep learning framework, LiteSOS-Net, for simultaneous lung cancer type classification and TNM stage prediction from CT images, while achieving a suitable balance between diagnostic accuracy, computational efficiency, and real-time inference capability for resource-limited and edge healthcare environments.
Methods:
The proposed LiteSOS-Net framework integrates CT image preprocessing, Attention U-Net-based lung nodule segmentation, and a ShuffleNet-based classification backbone. The Osprey optimization algorithm is incorporated for efficient feature learning and classification. The framework was evaluated using three publicly available datasets: LIDC-IDRI, LUNA16, and IQ-OTH-NCCD. Performance was assessed using accuracy, sensitivity, specificity, macro-AUC, inference latency, and model size.
Results:
LiteSOS-Net achieved 98.47% accuracy and a macro-AUC of 0.9980 for six-class lung cancer type classification. For four-class stage prediction, the framework achieved 97.78% accuracy with a macro-AUC of 0.9903. The model demonstrated approximately 98.19% sensitivity and 98.79% specificity. In addition, the lightweight architecture achieved an inference latency as low as 22.1 ms, while quantization reduced the model size to approximately 6 MB, supporting real-time deployment on edge devices.
Conclusions:
The experimental results demonstrate that LiteSOS-Net provides accurate and computationally efficient lung cancer classification and stage prediction from CT images. By combining high diagnostic performance with low inference latency and a compact model size, the proposed framework shows potential for real-time and resource-efficient clinical decision-support applications, particularly in resource-limited healthcare settings.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.