Radiomics-based machine learning (ML) classifier for detection of type 2 diabetes on standard-of-care abdomen CTs: a proof-of-concept study

Imaging abnormalities of the pancreas in diabetes: implications for diagnosis and treatment

PURPOSE OF REVIEW

Radiographic imaging of the pancreas has drawn recent interest as pancreas volume may serve as a biomarker in identifying the likelihood of diabetes development, subtyping diabetes, and identifying prognostic indicators of poor ultimate outcomes. In this review, the role of pancreas imaging is discussed in various forms of diabetes including type 1 diabetes (T1D), type 2 diabetes (T2D), and diabetes of the exocrine pancreas, particularly diabetes following acute or chronic pancreatitis.

RECENT FINDINGS

Recent literature of quantitative pancreatic imaging correlating with various forms of diabetes was reviewed. Imaging-derived pancreas volumes are lower in individuals with diabetes, in particular those with T1D. Additionally, morphologic changes, enhancement characteristics, fat content, and MRI signal changes have been observed in different diabetes subtypes. These characteristics, as well as potential confounding variables, are reviewed. Additionally, future areas of research in MRI, CT radiomics, and pancreatitis-related imaging predictors of diabetes are discussed.

SUMMARY

Increased understanding of pancreas imaging features which predict diabetes and gauge prognosis has the potential to identify at-risk individuals and will become increasingly important in diabetes care. This article reviews the current knowledge of common pancreas imaging features as well as future directions of ongoing research in diabetes imaging.

Radiomics analysis of pancreas based on dual-energy computed tomography for the detection of type 2 diabetes mellitus

Objective

To utilize radiomics analysis on dual-energy CT images of the pancreas to establish a quantitative imaging biomarker for type 2 diabetes mellitus.

Materials and methods

In this retrospective study, 78 participants (45 with type 2 diabetes mellitus, 33 without) underwent a dual energy CT exam. Pancreas regions were segmented automatically using a deep learning algorithm. From these regions, radiomics features were extracted. Additionally, 24 clinical features were collected for each patient. Both radiomics and clinical features were then selected using the least absolute shrinkage and selection operator (LASSO) technique and then build classifies with random forest (RF), support vector machines (SVM) and Logistic. Three models were built: one using radiomics features, one using clinical features, and a combined model.

Results

Seven radiomic features were selected from the segmented pancreas regions, while eight clinical features were chosen from a pool of 24 using the LASSO method. These features were used to build a combined model, and its performance was evaluated using five-fold cross-validation. The best classifier type is Logistic and the reported area under the curve (AUC) values on the test dataset were 0.887 (0.73-1), 0.881 (0.715-1), and 0.922 (0.804-1) for the respective models.

Conclusion

Radiomics analysis of the pancreas on dual-energy CT images offers potential as a quantitative imaging biomarker in the detection of type 2 diabetes mellitus.

Assessing the robustness of a machine-learning model for early detection of pancreatic adenocarcinoma (PDA): evaluating resilience to variations in image acquisition and radiomics workflow using image perturbation methods

PURPOSE

To evaluate robustness of a radiomics-based support vector machine (SVM) model for detection of visually occult PDA on pre-diagnostic CTs by simulating common variations in image acquisition and radiomics workflow using image perturbation methods.

METHODS

Eighteen algorithmically generated-perturbations, which simulated variations in image noise levels (σ, 2σ, 3σ, 5σ), image rotation [both CT image and the corresponding pancreas segmentation mask by 45° and 90° in axial plane], voxel resampling (isotropic and anisotropic), gray-level discretization [bin width (BW) 32 and 64)], and pancreas segmentation (sequential erosions by 3, 4, 6, and 8 pixels and dilations by 3, 4, and 6 pixels from the boundary), were introduced to the original (unperturbed) test subset (n = 128; 45 pre-diagnostic CTs, 83 control CTs with normal pancreas). Radiomic features were extracted from pancreas masks of these additional test subsets, and the model's performance was compared vis-a-vis the unperturbed test subset.

RESULTS

The model correctly classified 43 out of 45 pre-diagnostic CTs and 75 out of 83 control CTs in the unperturbed test subset, achieving 92.2% accuracy and 0.98 AUC. Model's performance was unaffected by a three-fold increase in noise level except for sensitivity declining to 80% at 3σ (p = 0.02). Performance remained comparable vis-a-vis the unperturbed test subset despite variations in image rotation (p = 0.99), voxel resampling (p = 0.25-0.31), change in gray-level BW to 32 (p = 0.31-0.99), and erosions/dilations up to 4 pixels from the pancreas boundary (p = 0.12-0.34).

CONCLUSION

The model's high performance for detection of visually occult PDA was robust within a broad range of clinically relevant variations in image acquisition and radiomics workflow.

The Present and Future of Artificial Intelligence-Based Medical Image in Diabetes Mellitus: Focus on Analytical Methods and Limitations of Clinical Use

Artificial intelligence (AI)-based diagnostic technology using medical images can be used to increase examination accessibility and support clinical decision-making for screening and diagnosis. To determine a machine learning algorithm for diabetes complications, a literature review of studies using medical image-based AI technology was conducted using the National Library of Medicine PubMed, and the Excerpta Medica databases. Lists of studies using diabetes diagnostic images and AI as keywords were combined. In total, 227 appropriate studies were selected. Diabetic retinopathy studies using the AI model were the most frequent (85.0%, 193/227 cases), followed by diabetic foot (7.9%, 18/227 cases) and diabetic neuropathy (2.7%, 6/227 cases). The studies used open datasets (42.3%, 96/227 cases) or directly constructed data from fundoscopy or optical coherence tomography (57.7%, 131/227 cases). Major limitations in AI-based detection of diabetes complications using medical images were the lack of datasets (36.1%, 82/227 cases) and severity misclassification (26.4%, 60/227 cases). Although it remains difficult to use and fully trust AI-based imaging analysis technology clinically, it reduces clinicians' time and labor, and the expectations from its decision-support roles are high. Various data collection and synthesis data technology developments according to the disease severity are required to solve data imbalance.