1. A kidney tumor detection system comprising: • a CT image acquisition module (101) configured to capture computed tomography scans of renal regions; • a preprocessing unit (102) configured to rescale, normalize, and refine masks of CT images; • a deep learning segmentation module (103) comprising architectures selected from ConvNeXtV2FPN (103a), SegFormer (103b), and FastSCNN (103c), trained to identify kidney tumors; • a classification unit (104) configured to distinguish tumor boundaries and features; and • an evaluation module (105) configured to compute segmentation metrics including Dice coefficient, Intersection-over-Union (IoU), Precision, Recall, and ROC-based evaluation.

2. A method for detecting kidney tumors from CT images, comprising: • acquiring CT images using the CT image acquisition module (101); • preprocessing the images with rescaling, normalization, and mask refinement using the preprocessing unit (102); • segmenting kidney tumors using the deep learning segmentation module (103); • classifying tumor boundaries with the classification unit (104); and • evaluating segmentation performance using the evaluation module (105).

3. The system as claimed in Claim 1, wherein SegFormer (103b) achieves superior segmentation accuracy and clearer tumor boundaries compared to ConvNeXtV2FPN (103a) and FastSCNN (103c).

4. The system as claimed in Claim 1, wherein ConvNeXtV2FPN (103a) provides strong multi-scale feature fusion for tumor detection.

5. The system as claimed in Claim 1, wherein FastSCNN (103c) enables real-time inference with reduced computational overhead.

6. The method as claimed in Claim 2, wherein preprocessing (102) includes standardized normalization and removal of irrelevant background regions.

7. The system as claimed in Claim 1, wherein the evaluation module (105) outputs Dice and IoU scores for segmentation accuracy.

8. The system as claimed in Claim 1, wherein the classification unit (104) distinguishes tumor-specific texture variations across CT datasets.

9. The system as claimed in Claim 1, wherein the detection framework supports integration into clinical decision-support systems for near real-time workflows

Description:FIELD OF THE INVENTION
This invention relates to kidney tumor detection from computed tomography images using deep learning segmentation methods
BACKGROUND OF THE INVENTION
Accurate segmentation of kidney tumors in computed tomography (CT) images is essential for early diagnosis, precise treatment planning, and postoperative evaluation. Traditional manual segmentation methods are time-consuming, subjective, and prone to inter-observer variability.
SUMMARY OF THE INVENTION
This summary is provided to introduce a selection of concepts, in a simplified format, that are further described in the detailed description of the invention.
This summary is neither intended to identify key or essential inventive concepts of the invention and nor is it intended for determining the scope of the invention.
An automated kidney tumor segmentation framework using three modern deep learning architectures is suggested. They are ConvNextV2FPN, SegFormer, and FastSCNN. The system was trained and evaluated on publicly available CT datasets after applying standardized preprocessing steps, including rescaling, normalization, and mask refinement. Each model was assessed using quantitative metrics such as Dice coefficient, Intersection-over-Union (IoU), Precision, Recall, and ROC-based evaluation. Experimental results demonstrate that SegFormer consistently outperformed the other two architectures, achieving superior segmentation accuracy and producing clearer tumor boundaries, owing to its transformer-based global feature extraction and lightweight design. In contrast, ConvNextV2FPN delivered moderate performance with strong multi-scale feature fusion, while FastSCNN, designed for real-time inference, showed limitations in capturing fine tumor details. Overall, the findings indicate that transformer-driven architectures such as SegFormer offer significant advantages for kidney tumor segmentation on CT images, highlighting their potential for integration into clinical decision-support systems and future medical imaging workflows.
To further clarify advantages and features of the present invention, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
The illustrated embodiments of the subject matter will be understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and methods that are consistent with the subject matter as claimed herein, wherein:
FIGURE 1: SYSTEM ARCHITECTURE
FIGURE 2: DATAFLOW DIAGRAM
The figures depict embodiments of the present subject matter for the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DETAILED DESCRIPTION OF THE INVENTION
The detailed description of various exemplary embodiments of the disclosure is described herein with reference to the accompanying drawings. It should be noted that the embodiments are described herein in such details as to clearly communicate the disclosure. However, the amount of details provided herein is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure as defined by the appended claims.
It is also to be understood that various arrangements may be devised that, although not explicitly described or shown herein, embody the principles of the present disclosure. Moreover, all statements herein reciting principles, aspects, and embodiments of the present disclosure, as well as specific examples, are intended to encompass equivalents thereof.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a",” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and/or “including,” when used herein, specify the presence of stated features, integers, steps, operations, elements and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof.
It should also be noted that in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed concurrently or may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
In addition, the descriptions of "first", "second", “third”, and the like in the present invention are used for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Thus, features defining "first" and "second" may include at least one of the features, either explicitly or implicitly.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which example embodiments belong. It will be further understood that terms, e.g., those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The present invention relates to an automated kidney tumor detection framework that leverages deep learning segmentation methods to improve diagnostic accuracy and efficiency. The system begins with a CT image acquisition module (101) that captures high-resolution scans of renal regions. These images are processed by a preprocessing unit (102), which performs rescaling, normalization, and mask refinement to ensure consistent input quality and removal of irrelevant background data.
The processed images are then analyzed by a deep learning segmentation module (103), which incorporates modern architectures such as ConvNeXtV2FPN (103a), SegFormer (103b), and FastSCNN (103c). SegFormer demonstrates superior segmentation accuracy and clearer tumor boundaries due to transformer-based global feature extraction, ConvNeXtV2FPN provides strong multi-scale feature fusion, and FastSCNN offers lightweight real-time inference though with limitations in fine detail capture.
Following segmentation, a classification unit (104) identifies tumor boundaries and distinguishes tumor-specific texture variations. The system’s performance is validated by an evaluation module (105), which computes metrics including Dice coefficient, Intersection-over-Union (IoU), Precision, Recall, and ROC-based evaluation.
Experimental results confirm that SegFormer consistently outperforms other architectures, while ConvNeXtV2FPN and FastSCNN provide complementary strengths in multi-scale feature fusion and real-time inference. The integration of these modules enables robust detection of kidney tumors across diverse CT variations, reducing subjectivity and time delays associated with manual interpretation.
By combining CT imaging (101), preprocessing (102), advanced segmentation (103), classification (104), and evaluation (105), the invention provides a reliable, automated diagnostic tool that supports clinical decision-making, enhances diagnostic efficiency, and minimizes variability in tumor detection workflows.
An automated kidney tumor segmentation framework using three modern deep learning architectures is suggested. They are ConvNextV2FPN, SegFormer, and FastSCNN. The system was trained and evaluated on publicly available CT datasets after applying standardized preprocessing steps, including rescaling, normalization, and mask refinement. Each model was assessed using quantitative metrics such as Dice coefficient, Intersection-over-Union (IoU), Precision, Recall, and ROC-based evaluation. Experimental results demonstrate that SegFormer consistently outperformed the other two architectures, achieving superior segmentation accuracy and producing clearer tumor boundaries, owing to its transformer-based global feature extraction and lightweight design. In contrast, ConvNextV2FPN delivered moderate performance with strong multi-scale feature fusion, while FastSCNN, designed for real-time inference, showed limitations in capturing fine tumor details. Overall, the findings indicate that transformer-driven architectures such as SegFormer offer significant advantages for kidney tumor segmentation on CT images, highlighting their potential for integration into clinical decision-support systems and future medical imaging workflows.
Input Image Ground Truth FastSCNN ConvNextV2FPN SegFormer

While prior art in kidney tumor segmentation is dominated by heavy 3D U-Net/nnU-Net–based pipelines, our work introduces a unified automated framework that leverages three modern, diverse deep-learning architectures—ConvNeXtV2FPN, SegFormer, and the lightweight FastSCNN—to jointly enhance feature representation, efficiency, and robustness across the segmentation workflow.
 
, Claims:1. A kidney tumor detection system comprising:
• a CT image acquisition module (101) configured to capture computed tomography scans of renal regions;
• a preprocessing unit (102) configured to rescale, normalize, and refine masks of CT images;
• a deep learning segmentation module (103) comprising architectures selected from ConvNeXtV2FPN (103a), SegFormer (103b), and FastSCNN (103c), trained to identify kidney tumors;
• a classification unit (104) configured to distinguish tumor boundaries and features; and
• an evaluation module (105) configured to compute segmentation metrics including Dice coefficient, Intersection-over-Union (IoU), Precision, Recall, and ROC-based evaluation.
2. A method for detecting kidney tumors from CT images, comprising:
• acquiring CT images using the CT image acquisition module (101);
• preprocessing the images with rescaling, normalization, and mask refinement using the preprocessing unit (102);
• segmenting kidney tumors using the deep learning segmentation module (103);
• classifying tumor boundaries with the classification unit (104); and
• evaluating segmentation performance using the evaluation module (105).
3. The system as claimed in Claim 1, wherein SegFormer (103b) achieves superior segmentation accuracy and clearer tumor boundaries compared to ConvNeXtV2FPN (103a) and FastSCNN (103c).
4. The system as claimed in Claim 1, wherein ConvNeXtV2FPN (103a) provides strong multi-scale feature fusion for tumor detection.
5. The system as claimed in Claim 1, wherein FastSCNN (103c) enables real-time inference with reduced computational overhead.
6. The method as claimed in Claim 2, wherein preprocessing (102) includes standardized normalization and removal of irrelevant background regions.
7. The system as claimed in Claim 1, wherein the evaluation module (105) outputs Dice and IoU scores for segmentation accuracy.
8. The system as claimed in Claim 1, wherein the classification unit (104) distinguishes tumor-specific texture variations across CT datasets.
9. The system as claimed in Claim 1, wherein the detection framework supports integration into clinical decision-support systems for near real-time workflows