Regional lung densities in alpha-1 antitrypsin deficiency compared to predicted values.

BACKGROUND: We developed a method to calculate a standard score for lung tissue mass derived from CT scan images from a control group without respiratory disease. We applied the method to images from subjects with emphysema associated with alpha-1 antitrypsin deficiency (AATD) and used it to study regional patterns of differential tissue mass. METHODS: We explored different covariates in 76 controls. Standardization was applied to facilitate comparability between different CT scanners and a standard Z-score (Standard Mass Score, SMS) was developed, representing lung tissue loss compared to normal lung mass. This normative data was defined for the entire lungs and for delineated apical, central and basal regions. The agreement with DLCO%pred was explored in a data set of 180 patients with emphysema who participated in a trial of alpha-1-antitrypsin augmentation treatment (RAPID). RESULTS: Large differences between emphysematous and normal tissue of more than 10 standard deviations were found. There was reasonable agreement between SMS and DLCO%pred for the global densitometry (κ = 0.252, p < 0.001), varying from κ = 0.138 to κ = 0.219 and 0.264 (p < 0.001), in the apical, central and basal region, respectively. SMS and DLCO%pred correlated consistently across apical, central and basal regions. The SMS distribution over the different lung regions showed a distinct pattern suggesting that emphysema due to severe AATD develops from basal to central and ultimately apical regions. CONCLUSIONS: Standardization and normalization of lung densitometry is feasible and the adoption of the developed principles helps to characterize the distribution of emphysema, required for clinical decision making.

Automatic quantification of tenosynovitis on MRI of the wrist in patients with early arthritis: a feasibility study.

OBJECTIVES: Tenosynovitis (inflammation of the synovial lining of the sheath surrounding tendons) is frequently observed on MRI of early arthritis patients. Since visual assessment of tenosynovitis is a laborious task, we investigated the feasibility of automatic quantification of tenosynovitis on MRI of the wrist in a large cohort of early arthritis patients. METHODS: For 563 consecutive early arthritis patients (clinically confirmed arthritis ≥ 1 joint, symptoms < 2 years), MR scans of the wrist were processed in three automatic stages. First, super-resolution reconstruction was applied to fuse coronal and axial scans into a single high-resolution three-dimensional image. Next, 10 extensor/flexor tendon regions were segmented using atlas-based segmentation and marker-based watershed. A measurement region of interest (ROI) was defined around the tendons. Finally, tenosynovitis was quantified by identifying image intensity values associated with tenosynovial inflammation using fuzzy clustering and measuring the fraction of voxels with these characteristic intensities within the measurement ROI. A subset of 60 patients was used for training and the remaining 503 patients for validation. Correlation between quantitative measurements and visual scores was assessed through Pearson correlation coefficient. RESULTS: Pearson correlation between quantitative measurements and visual scores across 503 patients was r = 0.90, p < 0.001. False detections due to blood vessels and synovitis present within the measurement ROI contributed to a median offset from zero equivalent to 13.8% of the largest measurement value. CONCLUSION: Quantitative measurement of tenosynovitis on MRI of the wrist is feasible and largely consistent with visual scores. Further improvements in segmentation and exclusion of false detections are warranted. KEY POINTS: • Automatic measurement of tenosynovitis on MRI of the wrist is feasible and largely consistent with visual scores. • Blood vessels and synovitis in the vicinity of evaluated tendons can contribute to false detections in automatic measurements. • Further improvements in segmentation and exclusion of false detections are important directions of future work on the path to a robust quantification framework.

Automatic quantification of bone marrow edema on MRI of the wrist in patients with early arthritis: A feasibility study.

PURPOSE: To investigate the feasibility of automatic quantification of bone marrow edema (BME) on MRI of the wrist in patients with early arthritis. METHODS: For 485 early arthritis patients (clinically confirmed arthritis of one or more joints, symptoms for less than 2 years), MR scans of the wrist were processed in three automatic stages. First, super-resolution reconstruction was applied to fuse coronal and axial scans into a single high-resolution 3D image. Next, the carpal bones were located and delineated using atlas-based segmentation. Finally, the extent of BME within each bone was quantified by identifying image intensity values characteristic of BME by fuzzy clustering and measuring the fraction of voxels with these characteristic intensities within each bone. Correlation with visual BME scores was assessed through Pearson correlation coefficient. RESULTS: Pearson correlation between quantitative and visual BME scores across 485 patients was r=0.83, P<0.001. CONCLUSIONS: Quantitative measurement of BME on MRI of the wrist has the potential to provide a feasible alternative to visual scoring. Complete automation requires automatic detection and compensation of acquisition artifacts. Magn Reson Med 79:1127-1134, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Intravenous augmentation treatment and lung density in severe α1 antitrypsin deficiency (RAPID): a randomised, double-blind, placebo-controlled trial.

BACKGROUND: The efficacy of α1 proteinase inhibitor (A1PI) augmentation treatment for α1 antitrypsin deficiency has not been substantiated by a randomised, placebo-controlled trial. CT-measured lung density is a more sensitive measure of disease progression in α1 antitrypsin deficiency emphysema than spirometry is, so we aimed to assess the efficacy of augmentation treatment with this measure. METHODS: The RAPID study was a multicentre, double-blind, randomised, parallel-group, placebo-controlled trial of A1PI treatment in patients with α1 antitrypsin deficiency. We recruited eligible non-smokers (aged 18-65 years) in 28 international study centres in 13 countries if they had severe α1 antitrypsin deficiency (serum concentration <11 µM) with a forced expiratory volume in 1 s of 35-70% (predicted). We excluded patients if they had undergone, or were on the waiting list to undergo, lung transplantation, lobectomy, or lung volume-reduction surgery, or had selective IgA deficiency. We randomly assigned patients (1:1; done by Accovion) using a computerised pseudorandom number generator (block size of four) with centre stratification to receive A1PI intravenously 60 mg/kg per week or placebo for 24 months. All patients and study investigators (including those assessing outcomes) were unaware of treatment allocation throughout the study. Primary endpoints were CT lung density at total lung capacity (TLC) and functional residual capacity (FRC) combined, and the two separately, at 0, 3, 12, 21, and 24 months, analysed by modified intention to treat (patients needed at least one evaluable lung density measurement). This study is registered with ClinicalTrials.gov, number NCT00261833. A 2-year open-label extension study was also completed (NCT00670007). FINDINGS: Between March 1, 2006, and Nov 3, 2010, we randomly allocated 93 (52%) patients A1PI and 87 (48%) placebo, analysing 92 in the A1PI group and 85 in the placebo group. The annual rate of lung density loss at TLC and FRC combined did not differ between groups (A1PI -1·50 g/L per year [SE 0·22]; placebo -2·12 g/L per year [0·24]; difference 0·62 g/L per year [95% CI -0·02 to 1·26], p=0·06). However, the annual rate of lung density loss at TLC alone was significantly less in patients in the A1PI group (-1·45 g/L per year [SE 0·23]) than in the placebo group (-2·19 g/L per year [0·25]; difference 0·74 g/L per year [95% CI 0·06-1·42], p=0·03), but was not at FRC alone (A1PI -1·54 g/L per year [0·24]; placebo -2·02 g/L per year [0·26]; difference 0·48 g/L per year [-0·22 to 1·18], p=0·18). Treatment-emergent adverse events were similar between groups, with 1298 occurring in 92 (99%) patients in the A1PI group and 1068 occuring in 86 (99%) in the placebo group. 71 severe treatment-emergent adverse events occurred in 25 (27%) patients in the A1PI group and 58 occurred in 27 (31%) in the placebo group. One treatment-emergent adverse event leading to withdrawal from the study occurred in one patient (1%) in the A1PI group and ten occurred in four (5%) in the placebo group. One death occurred in the A1PI group (respiratory failure) and three occurred in the placebo group (sepsis, pneumonia, and metastatic breast cancer). INTERPRETATION: Measurement of lung density with CT at TLC alone provides evidence that purified A1PI augmentation slows progression of emphysema, a finding that could not be substantiated by lung density measurement at FRC alone or by the two measurements combined. These findings should prompt consideration of augmentation treatment to preserve lung parenchyma in individuals with emphysema secondary to severe α1 antitrypsin deficiency. FUNDING: CSL Behring.

Lung Function and CT Densitometry in Subjects with alpha-1-Antitrypsin Deficiency and Healthy Controls at 35 Years of Age.

Alpha-1-antitrypsin (AAT) deficiency is a genetic risk factor for pulmonary emphysema. In 1972-74 all 200,000 Swedish new-born infants were screened for AAT deficiency. The aim of the present study was to investigate whether the PiZZ and PiSZ individuals identified by this screening have signs of emphysema and the role of smoking in this, compared with a random sample of control subjects at 35 years of age. The study participants underwent complete pulmonary function tests (PFT) and CT densitometry. The fifteenth percentile density (PD15) and the relative area below -910 HU (RA-910) were analyzed. Fifty-four PiZZ, 21 PiSZ and 66 PiMM control subjects participated in the study. No significant differences were found in lung function between the never-smoking AAT-deficient and control subjects. The 16 PiZZ ever-smokers had significantly lower carbon monoxide transfer coefficient (KCO) than the 20 PiSZ never-smokers (p = 0.014) and the 44 PiMM never-smokers (p = 0.005). After correction for the CT derived lung volume, the PiZZ ever-smokers had significantly lower PD15 (p = 0.046) than the ever-smoking controls. We conclude that 35-year-old PiZZ and PiSZ never-smokers have normal lung function when compared with never-smoking control subjects. The differences in KCO and CT densitometric parameters between the PiZZ ever-smokers and the control subjects may indicate early signs of emphysema.

Deep learning in rheumatological image interpretation.

Artificial intelligence techniques, specifically deep learning, have already affected daily life in a wide range of areas. Likewise, initial applications have been explored in rheumatology. Deep learning might not easily surpass the accuracy of classic techniques when performing classification or regression on low-dimensional numerical data. With images as input, however, deep learning has become so successful that it has already outperformed the majority of conventional image-processing techniques developed during the past 50 years. As with any new imaging technology, rheumatologists and radiologists need to consider adapting their arsenal of diagnostic, prognostic and monitoring tools, and even their clinical role and collaborations. This adaptation requires a basic understanding of the technical background of deep learning, to efficiently utilize its benefits but also to recognize its drawbacks and pitfalls, as blindly relying on deep learning might be at odds with its capabilities. To facilitate such an understanding, it is necessary to provide an overview of deep-learning techniques for automatic image analysis in detecting, quantifying, predicting and monitoring rheumatic diseases, and of currently published deep-learning applications in radiological imaging for rheumatology, with critical assessment of possible limitations, errors and confounders, and conceivable consequences for rheumatologists and radiologists in clinical practice.

MRI of diffuse-type tenosynovial giant cell tumour in the knee: a guide for diagnosis and treatment response assessment.

Tenosynovial giant cell tumour (TGCT) is a rare soft-tissue tumour originating from synovial lining of joints, bursae and tendon sheaths. The tumour comprises two subtypes: the localised-type (L-TGCT) is characterised by a single, well-defined lesion, whereas the diffuse-type (D-TGCT) consists of multiple lesions without clear margins. D-TGCT was previously known as pigmented villonodular synovitis. Although benign, TGCT can behave locally aggressive, especially the diffuse-type. Magnetic resonance imaging (MRI) is the modality of choice to diagnose TGCT and discriminate between subtypes. MRI can also provide a preoperative map before synovectomy, the mainstay of treatment. Finally, since the arrival of colony-stimulating factor 1-receptor inhibitors, a novel systemic therapy for D-TGCT patients with relapsed or inoperable disease, MRI is key in assessing treatment response. As recurrence after treatment of D-TGCT occurs more often than in L-TGCT, follow-up imaging plays an important role in D-TGCT. Reading follow-up MRIs of these diffuse synovial tumours may be a daunting task. Therefore, this educational review focuses on MRI findings in D-TGCT of the knee, which represents the most involved joint site (approximately 70% of patients). We aim to provide a systematic approach to assess the knee synovial recesses, highlight D-TGCT imaging findings, and combine these into a structured report. In addition, differential diagnoses mimicking D-TGCT, potential pitfalls and evaluation of tumour response following systemic therapies are discussed. Finally, we propose automated volumetric quantification of D-TGCT as the next step in quantitative treatment response assessment as an alternative to current radiological assessment criteria.

Automated quantification of the pulmonary vasculature in pulmonary embolism and chronic thromboembolic pulmonary hypertension.

The shape and distribution of vascular lesions in pulmonary embolism (PE) and chronic thromboembolic pulmonary hypertension (CTEPH) are different. We investigated whether automated quantification of pulmonary vascular morphology and densitometry in arteries and veins imaged by computed tomographic pulmonary angiography (CTPA) could distinguish PE from CTEPH. We analyzed CTPA images from a cohort of 16 PE patients, 6 CTEPH patients, and 15 controls. Pulmonary vessels were extracted with a graph-cut method, and separated into arteries and veins using deep-learning classification. Vascular morphology was quantified by the slope (α) and intercept (ß) of the vessel radii distribution. To quantify lung perfusion defects, the median pulmonary vascular density was calculated. By combining these measurements with densities measured in parenchymal areas, pulmonary trunk, and descending aorta, a static perfusion curve was constructed. All separate quantifications were compared between the three groups. No vascular morphology differences were detected in contrast to vascular density values. The median vascular density (interquartile range) was -567 (113), -452 (95), and -470 (323) HU, for the control, PE, and CTEPH group. The static perfusion curves showed different patterns between groups, with a statistically significant difference in aorta-pulmonary trunk gradient between the PE and CTEPH groups (p = 0.008). In this proof of concept study, not vasculature morphology but densities differentiated between patients of three groups. Further technical improvements are needed to allow for accurate differentiation between PE and CTEPH, which in this study was only possible statistically by measuring the density gradient between aorta and pulmonary trunk.

Randomised controlled trial for emphysema with a selective agonist of the γ-type retinoic acid receptor.

Palovarotene is an oral γ-selective retinoid agonist. In animal emphysema models, palovarotene reduced inflammation, promoted structural repair and functional improvement. REPAIR (Retinoid treatment of Emphysema in Patients on the α(1)-antitrypsin International Registry), was an investigator-initiated, double-blind, placebo-controlled randomised study to assess the safety and efficacy of 5 mg·day(-1) palovarotene given for 1 year to 262 patients with severe α(1)-antitrypsin deficiency and emphysema confirmed by computed tomography. Change in volume-adjusted 15th percentile point lung density from baseline in 1 year was the primary end-point; functional end-points were also regularly assessed. We randomly assigned 133 and 129 patients to placebo or palovarotene, respectively. Both groups were well matched for all baseline characteristics, including respiratory medications. 88% and 85% of patients completed 1 year of treatment with placebo and palovarotene, respectively. Palovarotene was generally well tolerated. In the study completers population, the placebo-corrected difference of lung density was -0.45 HU at week 28 (p=0.64) and -0.25 HU at week 52 (p=0.94). A nonsignificant treatment difference in most functional parameters of the lung in favour of the drug was observed over time suggesting potential pharmacological effects of palovarotene. Palovarotene 5 mg·day(-1) over 1 yr failed to show a significant benefit on lung density in moderate-to-severe emphysema secondary to severe α(1)-antitrypsin deficiency.

Automatic radiographic quantification of hand osteoarthritis; accuracy and sensitivity to change in joint space width in a phantom and cadaver study.

OBJECTIVE: To validate a newly developed quantification method that automatically detects and quantifies the joint space width (JSW) in hand radiographs. Repeatability, accuracy and sensitivity to changes in JSW were determined. The influence of joint location and joint shape on the measurements was tested. METHODS: A mechanical micrometer set-up was developed to define and adjust the true JSW in an acrylic phantom joint and in human cadaver-derived phalangeal joints. Radiographic measurements of the JSW were compared to the true JSW. Repeatability, systematic error (accuracy) and sensitivity (defined as the smallest detectable difference (SDD)) were determined. The influence of joint position on the JSW measurement was assessed by varying the location of the acrylic phantom on the X-ray detector with respect to the X-ray beam and the influence of joint shape was determined by using morphologically different human cadaver joints. RESULTS: The mean systematic error was 0.052 mm in the phantom joint and 0.210 mm in the cadaver experiment. In the phantom experiments, the repeatability was high (SDD = 0.028 mm), but differed slightly between joint locations (p = 0.046), and a change in JSW of 0.037 mm could be detected. Dependent of the joint shape in the cadaver hand, a change in JSW between 0.018 and 0.047 mm could be detected. CONCLUSIONS: The automatic quantification method is sensitive to small changes in JSW. Considering the published data of JSW decline in the normal and osteoarthritic population, the first signs of OA progression with this method can be detected within 1 or 2 years.

Rapid fall in lung density following smoking cessation in COPD.

INTRODUCTION: Whether smoking-induced lung inflammation subsides after smoking cessation is currently a matter of debate. We used computed tomography (CT) to evaluate the effect of smoking cessation on lung density in patients with COPD. MATERIAL AND METHODS: Thirty-six patients quit smoking out of 254 current smokers with COPD who were followed with annual CT and lung function tests (LFT) for 2?4 years as part of a randomised placebo-controlled trial of the effect of inhaled budesonide on CT-lung density. Lung density was expressed as the 15th percentile density (PD15) and relative area of emphysema below -910 HU (RA-910). From the time-trends in the budesonide and placebo groups the expected CT-lung densities at the first visit after smoking cessation were calculated by linear regression and compared to the observed densities. RESULTS: Following smoking cessation RA-910 increased by 2.6% (p = 0.003) and PD15 decreased by -4.9 HU (p = 0.0002). Furthermore, changes were larger in the budesonide group than the placebo group (PD15: -7.1 vs -2.8 HU. RA-910 3.7% vs 1.7%). These differences were, however, not statistically significant. The LFT parameters (FEV(1) and diffusion capacity) were not significantly influenced by smoking cessation. CONCLUSION: Inflammation partly masks the presence of emphysema on CT and smoking cessation results in a paradoxical fall in lung density, which resembles rapid progression of emphysema. This fall in density is probably due to an anti-inflammatory effect of smoking cessation.

MRI-based 3D retinal shape determination.

OBJECTIVE: To establish a good method to determine the retinal shape from MRI using three-dimensional (3D) ellipsoids as well as evaluate its reproducibility. METHODS AND ANALYSIS: The left eyes of 31 volunteers were imaged using high-resolution ocular MRI. The 3D MR-images were segmented and ellipsoids were fitted to the resulting contours. The dependency of the resulting ellipsoid parameters on the evaluated fraction of the retinal contour was assessed by fitting ellipsoids to 41 different fractions. Furthermore, the reproducibility of the complete procedure was evaluated in four subjects. Finally, a comparison with conventional two-dimensional (2D) methods was made. RESULTS: The mean distance between the fitted ellipsoids and the segmented retinal contour was 0.03±0.01 mm (mean±SD) for the central retina and 0.13±0.03 mm for the peripheral retina. For the central retina, the resulting ellipsoid radii were 12.9±0.9, 13.7±1.5 and 12.2±1.2 mm along the horizontal, vertical and central axes. For the peripheral retina, these radii decreased to 11.9±0.6, 11.6±0.4 and 10.4±0.7 mm, which was accompanied by a mean 1.8 mm posterior shift of the ellipsoid centre. The reproducibility of the ellipsoid fitting was 0.3±1.2 mm for the central retina and 0.0±0.1 mm for the peripheral retina. When 2D methods were used to fit the peripheral retina, the fitted radii differed a mean 0.1±0.1 mm from the 3D method. CONCLUSION: An accurate and reproducible determination of the 3D retinal shape based on MRI is provided together with 2D alternatives, enabling wider use of this method in the field of ophthalmology.

Evaluation of intraocular lens position and retinal shape in negative dysphotopsia using high-resolution magnetic resonance imaging.

PURPOSE: To assess potential relationships of intraocular lens (IOL) position and retinal shape in negative dysphotopsia (ND). SETTING: Department of Ophthalmology, Leiden University Medical Center, Leiden, the Netherlands. DESIGN: Case-control study. METHODS: High-resolution ocular magnetic resonance imaging (MRI) scans were performed in patients with ND and pseudophakic controls, and subsequently used to determine the displacement and tilt of the in-the-bag IOL about the pupil and iris. In addition, anterior segment tomography was used to assess the iris-IOL distance. Furthermore, the retinal shape was quantified from the MRI scans by fitting an ellipse to the segmented inner boundary of the retina. Both the IOL position and retinal shape were compared between groups to assess their potential role in the etiology of ND. RESULTS: In total, 37 patients with ND and 26 pseudophakic controls were included in the study. The mean displacement and tilt of the IOL were less than 0.1 mm and 0.5 degrees, respectively, in both groups and all directions. The corresponding mean iris-IOL distance was 1.1 mm in both groups. Neither of these values differed statistically significantly between groups (all P values >.6). The retinal shape showed large variations but was not statistically significantly different between the groups in both the left-right (P = .10) and the anterior-posterior (P = .56) directions. CONCLUSIONS: In this study, the in-the-bag IOL position and retinal shape did not statistically significantly differ between patients with ND and the general pseudophakic population. Given the large variation in retinal shape between subjects, however, it could still be an important factor in a multifactorial origin of ND.

Therapeutic efficacy of α-1 antitrypsin augmentation therapy on the loss of lung tissue: an integrated analysis of 2 randomised clinical trials using computed tomography densitometry.

BACKGROUND: Two randomised, double-blind, placebo-controlled trials have investigated the efficacy of IV alpha-1 antitrypsin (AAT) augmentation therapy on emphysema progression using CT densitometry. METHODS: Data from these similar trials, a 2-center Danish-Dutch study (n = 54) and the 3-center EXAcerbations and CT scan as Lung Endpoints (EXACTLE) study (n = 65), were pooled to increase the statistical power. The change in 15th percentile of lung density (PD15) measured by CT scan was obtained from both trials. All subjects had 1 CT scan at baseline and at least 1 CT scan after treatment. Densitometric data from 119 patients (AAT [Alfalastin® or Prolastin®], n = 60; placebo, n = 59) were analysed by a statistical/endpoint analysis method. To adjust for lung volume, volume correction was made by including the change in log-transformed total lung volume as a covariate in the statistical model. RESULTS: Mean follow-up was approximately 2.5 years. The mean change in lung density from baseline to last CT scan was -4.082 g/L for AAT and -6.379 g/L for placebo with a treatment difference of 2.297 (95% CI, 0.669 to 3.926; p = 0.006). The corresponding annual declines were -1.73 and -2.74 g/L/yr, respectively. CONCLUSIONS: The overall results of the combined analysis of 2 separate trials of comparable design, and the only 2 controlled clinical trials completed to date, has confirmed that IV AAT augmentation therapy significantly reduces the decline in lung density and may therefore reduce the future risk of mortality in patients with AAT deficiency-related emphysema. TRIAL REGISTRATION: The EXACTLE study was registered in ClinicalTrials.gov as 'Antitrypsin (AAT) to Treat Emphysema in AAT-Deficient Patients'; ClinicalTrials.gov Identifier: NCT00263887.

Prediction of lung function and lung density of young adults who had bronchopulmonary dysplasia.

COPD risk is jointly determined by fetal lung development, lung growth rate and lung growth duration leading to the maximally attained level of lung function in early adulthood. Bronchopulmonary dysplasia (BPD) is considered a developmental arrest of alveolarisation. Long-term outcome studies of adult survivors born before the introduction of surfactant therapy ("old BPD") showed impaired lung function. We aimed to predict adult lung function and lung density in a cohort of premature infants born in the surfactant era, representing "new BPD". We studied a cohort of young adults born between 1987 and 1998, with (n=36) and without (n=28) BPD, treated in a single centre. Their perinatal characteristics and pulmonary function in infancy were studied by regression analysis for correlation with adult lung function and tissue lung density, all expressed by z-scores, at a mean age of 19.7±1.1 and 21±2.2 years, respectively. Although BPD adults had on average lower forced expiratory volume in 1 s (zFEV1)/forced vital capacity (FVC) and zFEV1 than those without, 55% of the BPD group had zFEV1/FVC values above the lower limit of normal (LLN). Moreover, above LLN values of diffusing capacity of the lung for carbon monoxide (zD LCO) was present in 89% of BPD adults and lung density in 71%. Only higher oxygen supply (F IO2) at 36 weeks post-conception of BPD subjects had a trend with lower zFEV1 (B=-6.4; p=0.053) and lower zD LCO (B=-4.1; p=0.023) at adulthood. No statistically significant predictors of new BPD were identified.

Artificial intelligence in detecting early RA.

To prevent chronicity of Rheumatoid Arthritis (RA) by early treatment, detecting inflammatory signs in an early phase is essential. Since Magnetic Resonance Imaging (MRI) of the wrist, hand and foot can detect inflammation before it is clinically detectable, this modality may play an important role in achieving very early diagnoses. By collecting large amounts of MRI data from healthy controls and patients with arthralgia suspicious for progression to RA, patterns can be studied that are most specific for early development of RA. Furthermore, MRI can be used as outcome parameter for randomized placebo-controlled trials on early RA treatment, by detecting subtle changes in image intensities originating from natural progression or treatment effects. Very large amounts of MRI data, however, make manual quantification impractical and the coarse scale used in visual scoring systems (i.e. whole values between 0 and 3) limits its sensitivity to detect changes that are likely to be very subtle in such an early phase. In recent years, advances in artificial intelligence and especially 'deep learning' in interpreting medical images have shown that -in specific areas- a computerized analysis can outperform human observers. Therefore, research has been initiated into applying these artificial intelligence techniques to the quantification of early RA from MRI data. In this paper, an overview is given on the background and history of artificial intelligence, with a special focus on recent developments in 'deep learning', and how these techniques could be applied to detect subtle inflammatory changes in MRI data.

Identifying MRI-detected inflammatory features specific for rheumatoid arthritis: two-fold feature reduction maintains predictive accuracy in clinically suspect arthralgia patients.

OBJECTIVE: MRI-detected inflammation is considered of diagnostic value for rheumatoid arthritis (RA), but its evaluation involves a time-consuming scoring of 61 joint-level features. It is not clear, however, which of these features are specific for RA and whether evaluating a subset of specific features is sufficient to differentiate RA patients. This study aimed to identify a subset of RA-specific features in a case-control setting and validate them in a longitudinal cohort of arthralgia patients. METHODS: The difference in frequency of MRI-detected inflammation (bone marrow edema, synovitis, and tenosynovitis) between 199 RA patients and 193 controls was studied in 61 features across the wrist, metacarpophalangeal, and metatarsophalangeal joints. A subset of RA-specific features was obtained by applying a cutoff on the frequency difference while maximizing discriminative performance. For validation, this subset was used to predict arthritis development in 225 clinically suspect arthralgia (CSA) patients. Diagnostic performance was compared to a reference method that uses the complete set of 61 features normalized for inflammation levels in age-matched controls. RESULTS: Subset of 30 features, mainly (teno)synovitis, was obtained from the case-control setting. Validation in CSA patients yielded an area of 0.69 (95% CI: 0.59-0.78) under the ROC curve and a positive predictive value (PPV) of 31%, compared to 0.68 (95% CI: 0.60-0.77) and 29% PPV of the reference method with 61 features. CONCLUSION: Subset of 30 MRI-detected inflammatory features, dominated by (teno)synovitis, offers a considerable reduction of scoring efforts without compromising accuracy for prediction of arthritis development in CSA patients.