Fluorescence optical imaging feature selection with machine learning for differential diagnosis of selected rheumatic diseases.

Background and objective: Accurate and fast diagnosis of rheumatic diseases affecting the hands is essential for further treatment decisions. Fluorescence optical imaging (FOI) visualizes inflammation-induced impaired microcirculation by increasing signal intensity, resulting in different image features. This analysis aimed to find specific image features in FOI that might be important for accurately diagnosing different rheumatic diseases. Patients and methods: FOI images of the hands of patients with different types of rheumatic diseases, such as rheumatoid arthritis (RA), osteoarthritis (OA), and connective tissue diseases (CTD), were assessed in a reading of 20 different image features in three phases of the contrast agent dynamics, yielding 60 different features for each patient. The readings were analyzed for mutual differential diagnosis of the three diseases (One-vs-One) and each disease in all data (One-vs-Rest). In the first step, statistical tools and machine-learning-based methods were applied to reveal the importance rankings of the features, that is, to find features that contribute most to the model-based classification. In the second step machine learning with a stepwise increasing number of features was applied, sequentially adding at each step the most crucial remaining feature to extract a minimized subset that yields the highest diagnostic accuracy. Results: In total, n = 605 FOI of both hands were analyzed (n = 235 with RA, n = 229 with OA, and n = 141 with CTD). All classification problems showed maximum accuracy with a reduced set of image features. For RA-vs.-OA, five features were needed for high accuracy. For RA-vs.-CTD ten, OA-vs.-CTD sixteen, RA-vs.-Rest five, OA-vs.-Rest eleven, and CTD-vs-Rest fifteen, features were needed, respectively. For all problems, the final importance ranking of the features with respect to the contrast agent dynamics was determined. Conclusions: With the presented investigations, the set of features in FOI examinations relevant to the differential diagnosis of the selected rheumatic diseases could be remarkably reduced, providing helpful information for the physician.

Features Found in Indocyanine Green-Based Fluorescence Optical Imaging of Inflammatory Diseases of the Hands.

Rheumatologists in Europe and the USA increasingly rely on fluorescence optical imaging (FOI, Xiralite) for the diagnosis of inflammatory diseases. Those include rheumatoid arthritis, psoriatic arthritis, and osteoarthritis, among others. Indocyanine green (ICG)-based FOI allows visualization of impaired microcirculation caused by inflammation in both hands in one examination. Thousands of patients are now documented and most literature focuses on inflammatory arthritides, which affect synovial joints and their related structures, making it a powerful tool in the diagnostic process of early undifferentiated arthritis and rheumatoid arthritis. However, it has become gradually clear that this technique has the potential to go even further than that. FOI allows visualization of other types of tissues. This means that FOI can also support the diagnostic process of vasculopathies, myositis, collagenoses, and other connective tissue diseases. This work summarizes the most prominent imaging features found in FOI examinations of inflammatory diseases, outlines the underlying anatomical structures, and introduces a nomenclature for the features and, thus, supports the idea that this tool is a useful part of the imaging repertoire in rheumatology clinical practice, particularly where other imaging methods are not easily available.

3D-printed tissue-simulating phantoms for near-infrared fluorescence imaging of rheumatoid diseases.

SIGNIFICANCE: Fluorescence imaging of rheumatoid diseases with indocyanine green (ICG) is an emerging technique with unique potential for diagnosis and therapy. Device characterization, monitoring of the performance, and further developments of the technique require tissue-equivalent fluorescent phantoms of high stability with appropriate anatomical shapes. AIM: Our investigations aim at the development of a three-dimensional (3D) printing technique to fabricate hand and finger models with appropriate optical properties in the near-infrared spectral range. These phantoms should have fluorescence properties similar to ICG, and excellent photostability and durability over years. APPROACH: We modified a 3D printing methacrylate photopolymer by adding the fluorescent dye Lumogen IR 765 to the raw material. Reduced scattering and absorption coefficients were adjusted to values representative of the human hand by incorporating titanium dioxide powder and black ink. The properties of printed phantoms of various compositions were characterized using UV/Vis and fluorescence spectroscopy, and time-resolved measurements. Photostability and bleaching were investigated with a hand imager. For comparison, several phantoms with ICG as fluorescent dye were printed and characterized as well. RESULTS: The spectral properties of Lumogen IR 765 are very similar to those of ICG. By optimizing the concentrations of Lumogen, titanium dioxide, and ink, anatomically shaped hand and vessel models with properties equivalent to in vivo investigations with a fluorescence hand imager could be printed. Phantoms with Lumogen IR 765 had an excellent photostability over up to 4 years. In contrast, phantoms printed with ICG showed significant bleaching and degradation of fluorescence over time. CONCLUSIONS: 3D printing of phantoms with Lumogen IR 765 is a promising method for fabricating anatomically shaped fluorescent tissue models of excellent stability with spectral properties similar to ICG. The phantoms are well-suited to monitor the performance of hand imagers. Concepts can easily be transferred to other fluorescence imaging applications of ICG.

Quantitative assessment of synovitis in patients with rheumatoid arthritis using fluorescence optical imaging.

INTRODUCTION: To prospectively evaluate quantitative assessment of fluorescence optical imaging (FOI) for differentiation of synovitic from non-synovitic joints in patients suffering from rheumatoid arthritis (RA). METHODS: FOI of the hands was performed in patients with active RA, and a stratified quantitative fluorescence readout (FLRO) of 3 phases (1-120 s; 121-240 s; 241-360 s) was generated for 5 individual joints of the clinical predominant hand (carpal joint, metacarpophalangeal and proximal interphalangeal joints of digits II & III). To dissect the effect of the overall perfusion of the hand from the perfusion due to synovitis, a fluorescence ratio (FLRA) was additionally calculated, dividing each FLRO by the readout of the eponychium of digit II. The mean FLRO and FLRA were compared between joints with absent vs. present synovitis determined by clinical examination, grayscale, color Doppler ultrasonography, or magnetic resonance imaging (MRI). RESULTS: The analysis for 90 individual joints from 18 patients yielded FLRO ranging from 4.4 to 49.0 × 10(3), and FLRAs ranging from 0.37 to 2.27. Overall, the analyses based on the FLRA revealed a higher discrimination than the analyses related to the FLRO, demonstrating most significant differences in phases 2 and 3. A sensitivity of 26/39 (67%) and a specificity of 31/40 (77%) were calculated for FLRA of phase 3 using a cut-off value of more than 1.2 to detect MRI-confirmed synovitis with FOI. CONCLUSIONS: FOI has a potential for visualizing synovitis in subjects with RA. For adequate FOI interpretation, quantitative analysis should be based on the novel FLRA calculated for phases 2 and 3.