Glycation promotes pulp calcification in Type 2 diabetes rat model.

OBJECTIVES: Intrapulpal calcifications can occur in the dental pulp of patients with diabetes. We focused on the association between ectopic calcifications in the dental pulp and advanced glycation end products (AGEs) in Spontaneously Diabetic Torii (SDT)-fatty rats, an obese type 2 diabetic rat model, to determine the mechanism of calcification with pulp stone in the dental pulp. MATERIALS AND METHODS: Pathologic calcification in the dental pulp of SDT-fatty rats was observed using electron microscopy and immunohistochemical analysis. Moreover, mechanical analysis of periapical region of molar tooth against occlusal force was performed. RESULTS: In SDT-fatty rats, pathogenic pulpal calcifications occurred during blood glucose elevation after 6 weeks, and granular calcification was observed in the dental pulp after 11 weeks. Pentosidine, a major AGE, and the receptor for AGEs were strongly expressed in the dental pulp of SDT-fatty rats. S100A8, TNF-α, and IL-6 also showed positive response in the dental pulp of the SDT-fatty rat, which indicated pulpal inflammation. Blood flow disorder and hypoxic dental pulp cells were also observed. In silico simulation, strain from occlusal force concentrates on the root apex. CONCLUSIONS: Glycation makes blood vessels fragile, and occlusal forces damage the vessels mechanically. These are factors for intrapulpal calcification of diabetes.

Machine Learning for Multi-Vessel Coronary Artery Disease Prediction on Electrocardiogram Gated Single-Photon Emission Computed Tomography.

Background: Single-photon emission computed tomography (SPECT) encounters difficulties in diagnosing severe multi-vessel coronary artery disease (svMVD) because of balanced ischemia. We estimated the predictive value of electrocardiogram-gated SPECT for svMVD and improved it using machine learning (ML). Methods and results: We enrolled consecutive 335 patients (median age, 74 years; 255 men) who underwent adenosine stress-gated SPECT (99mTechnesium) and coronary angiography. svMVD was defined as three-vessel disease or left main tract stenosis. Predictive models were constructed using statistical and ML methods. Eighteen cases (5%) showed svMVD, and diabetes, summed stress score (SSS), and the max difference among segmental time of stroke volume per cardiac cycle (MDSV: a parameter of left ventricular [LV] end-systolic dyssynchrony) on adenosine stress were independent significant predictors. The area under the receiver operating characteristic curve (AUC) of SSS and MDSV on stress were 0.759 and 0.763, respectively. Conversely, the extra trees classifier and light gradient boosting machine had improved AUC values of 0.826 and 0.870, respectively, and the MDSV on stress and diabetes showed high feature values in the ML models. Conclusion: ML on SPECT helped to improve the diagnostic performance of svMVD and diabetes, and the parameters of LV dyssynchrony played essential roles in the ML predictive models.

Risk prediction of inappropriate implantable cardioverter-defibrillator therapy using machine learning.

We aimed to develop machine learning-based predictive models for identifying inappropriate implantable cardioverter-defibrillator (ICD) therapy. Our study included 182 consecutive cases (average age 62.2 ± 4.5 years, 169 men) and employed 14 non-deep learning models for prediction (hold-out method). These models utilized selected electrocardiogram parameters and clinical features collected after ICD implantation. From the feature importance analysis of the best ML model, we established easily calculable scores. Among the patients, 25 (13.7%) experienced inappropriate therapy, and we identified 16 significant predictors. Using recursive feature elimination with cross-validation, we reduced the features to six with high feature importance: history of atrial arrhythmia (Atr-arrhythm), ischemic cardiomyopathy (ICM), absence of diabetes mellitus (DM), lack of cardiac resynchronization therapy (CRT), V3 ST level at J point (V3 STJ), and V5 R-wave amplitudes (V5R amp). The extra-trees classifier yielded the highest area under receiver operating characteristics curve (AUROC; 0.869 on test data). Thus, the Cardi35 score was defined as [+ 5.5*Atr-arrhythm - 1.5*CRT + 1.0*V3STJ + 1.0*V5R - 1.0*ICM - 0.5*DM], which demonstrated a hazard ratio of 1.62 (P < 0.001). A cut-off value of the score + 5.5 showed high AUROC (0.826). The ML approach can yield a robust prediction model, and the Cardi35 score was a convenient predictor for inappropriate therapy.

Clinical Significance of Coronary Healed Plaques in Stable Angina Pectoris Patients Undergoing Percutaneous Coronary Intervention.

BACKGROUND: Coronary healed plaques (HPs) reportedly have high vulnerability or show advanced atherosclerosis and a risk of rapid plaque progression. However, the prognosis of stable angina pectoris (SAP) patients with HPs undergoing percutaneous coronary intervention (PCI) remains under-investigated.Methods and Results: We analyzed 417 consecutive lesions from SAP patients undergoing pre- and post-intervention optical coherence tomography (OCT) for which HPs were defined as having a layered appearance. We investigated the differences in clinical and lesion characteristics, and post-PCI outcomes between HPs and non-HPs. To account for differences in clinical characteristics, propensity score matching was performed between the groups. HPs were observed in 216 lesions (51.8%) in the total cohort. In the propensity-matched cohort (n=294), HPs had higher rates of angiographic-B2/C lesions (77.6% vs. 59.2%, P<0.001), OCT-lipid-rich plaques (40.8% vs. 25.9%, P=0.007), macrophages (78.2% vs. 44.2%, P<0.001), greater luminal area stenosis (73.5±11.0% vs. 71.5±10.3%, P=0.002), and a higher prevalence of post-stenting irregular tissue protrusion (45.1% vs. 14.7%, P<0.001) than non-HPs. In the total cohort, target lesion revascularization (TLR)-free survival was poorer for HPs (log-rank test 7.66; P=0.006), and Cox proportional hazards analysis showed HP as an independent predictor of TLR (hazard ratio, 5.98; 95% confidence interval, 1.72-20.82; P=0.005). CONCLUSIONS: In SAP patients, HPs had greater complexity of lesions and higher vulnerability, which may have contributed to the poorer post-PCI outcomes.

Machine learning of microvolt-level 12-lead electrocardiogram can help distinguish takotsubo syndrome and acute anterior myocardial infarction.

Background: Qualitative differences in 12-lead electrocardiograms (ECG) at onset have been reported in patients with takotsubo syndrome (TTS) and acute anterior myocardial infarction (Ant-AMI). We aimed to distinguish these diseases by machine learning (ML) approach of microvolt-level quantitative measurements. Methods: We enrolled 56 consecutive patients with sinus rhythm TTS (median age, 77 years; 16 men), and 1-to-1 random matching was performed based on age and sex of the patients. The ECG in the emergency room was evaluated using an automated system (ECAPs12c; Nihon-Koden). Statistical and ML predictive models for TTS were constructed using clinical features and ECG parameters. Results: Statistically significant differences were observed in 25 parameters; the V1 ST level at the J point (V1 STJ) showed the lowest P value (P < .001). V1 STJ ≤+18 µV showed the highest accuracy for TTS (0.773). The highest area under the receiver operating characteristic curve (AUROC) was shown in the aVR ST level at 1/16th of the preceding R-R interval after the J point (aVR STmid: 0.727). Conversely, the light gradient boosting machine (model_LGBM) and extra tree classifier (model_ET) indicated higher accuracy (model_LGBM: 0.842, model_ET: 0.831) and AUROC (model_LGBM: 0.868, model_ET 0.896) than other statistical models. V1 STJ had high feature importance and Shapley additive explanation values in the 2 ML models. Conclusion: ML applied to automated microvolt-level ECG measurements showed the possibility of distinguishing between TTS and Ant-AMI, which may be a clinically useful ECG-based discriminator.

Novel susceptibility loci for steroid-associated osteonecrosis of the femoral head in systemic lupus erythematosus.

Osteonecrosis of the femoral head (ONFH) involves necrosis of bone and bone marrow of the femoral head caused by ischemia with unknown etiology. Previous genetic studies on ONFH failed to produce consistent results, presumably because ONFH has various causes with different genetic backgrounds and the underlying diseases confounded the associations. Steroid-associated ONFH (S-ONFH) accounts for one-half of all ONFH, and systemic lupus erythematosus (SLE) is a representative disease underlying S-ONFH. We performed a genome-wide association study (GWAS) to identify genetic risk factors for S-ONFH in patients with SLE. We conducted a two-staged GWAS on 636 SLE patients with S-ONFH and 95 588 non-SLE controls. Among the novel loci identified, we determined S-ONFH-specific loci by comparing allele frequencies between SLE patients without S-ONFH and non-SLE controls. We also used Korean datasets comprising 148 S-ONFH cases and 37 015 controls to assess overall significance. We evaluated the functional annotations of significant variants by in silico analyses. The Japanese GWAS identified 4 significant loci together with 12 known SLE susceptibility loci. The four significant variants showed comparable effect sizes on S-ONFH compared with SLE controls and non-SLE controls. Three of the four loci, MIR4293/MIR1265 [odds ratio (OR) = 1.99, P-value = 1.1 × 10-9)], TRIM49/NAALAD2 (OR = 1.65, P-value = 4.8 × 10-8) and MYO16 (OR = 3.91, P-value = 4.9 × 10-10), showed significant associations in the meta-analysis with Korean datasets. Bioinformatics analyses identified MIR4293, NAALAD2 and MYO16 as candidate causal genes. MIR4293 regulates a PPARG-related adipogenesis pathway relevant to S-ONFH. We identified three novel susceptibility loci for S-ONFH in SLE.

Effects of advanced glycation end products on dental pulp calcification.

OBJECTIVE: The main aim of this study was to elucidate the effects of advanced glycation end products (AGEs) on the calcification of cultured rat dental pulp cells (RDPCs) and to investigate the crystallisation ability of glycated collagen. MATERIALS AND METHODS: AGEs were prepared via non-enzymatic glycation of a dish coated with type I collagen using dl-glyceraldehyde. To investigate the effects of AGEs on RDPCs, we performed WST-1 and lactate dehydrogenase assays; alkaline phosphatase, Alizarin Red S and immunohistochemical staining; and real-time quantitative reverse transcription PCR. In addition, we performed crystallisation experiments on glycated collagen. All microstructures were analysed using scanning electron microscopy/energy-dispersive X-ray spectroscopy and transmission electron microscopy/diffraction pattern analysis. RESULTS: AGEs did not affect the proliferation or differentiation of RDPCs, but enhanced the calcification rate and cytotoxicity. No major calcification-related genes or proteins were involved in these calcifications, and glycated collagen was found to exhibit a negative polarity and form calcium phosphate crystals. Cytotoxicity due to drastic changes in the concentration of pericellular ions led to dystrophic calcification, assumed to represent an aspect of diabetic pulp calcifications. CONCLUSION: Glycated collagen-containing AGEs provide a nurturing environment for crystallisation and have a significant effect on the early calcification of RDPCs.

Does a combination treatment of repetitive transcranial magnetic stimulation and occupational therapy improve upper limb muscle paralysis equally in patients with chronic stroke caused by cerebral hemorrhage and infarction?: A retrospective cohort study.

ABSTRACT: The clinical presentation of stroke is usually more severe in patients with intracerebral hemorrhage (ICH) than in those with cerebral infarction (CI); recovery of stroke-related muscle paralysis is influenced and limited by the type of stroke. To date, many patients have been treated by neurorehabilitation; however, the changes in the recovery of motor paralysis depending on the type of stroke, ICH or CI, have not been established. This study aimed to determine this difference in improvement of upper extremity paralysis using 2-week in-hospital NovEl intervention Using Repetitive transcranial magnetic stimulation combined with Occupational therapy (NEURO).We scrutinized the medical records of all patients with poststroke (ICH or CI) upper extremity muscle paralysis using Fugl-Meyer assessments (FMAs) who had been admitted to 6 hospitals between March 2010 and December 2018 for rehabilitation treatment. This was a multiinstitutional, open-label, retrospective cohort study without control patients. We evaluated the effects of NEURO on patients with CI and ICH by dividing them into 2 groups according to the type of stroke, after adjustment for age, sex, dominant hand, affected hand side, time since stroke, and prediction of recovery capacity in the upper extremity.The study included 1716 (CI [n = 876] and ICH [n = 840]) patients who had undergone at least 2 FMAs and had experienced stroke at least 6 months before. The type of stroke had no effect on the outcomes (changes in the FMA-upper extremity score, F[4,14.0] = 2.05, P = .09, partial η2 = 0.01). Patients from all 5 groups equally benefited from the treatment (improvement in FMA scores) according to the sensitivity analysis-stratified analysis (F = 0.08 to 1.94, P > .16, partial η2 < 0.001).We conclude that NEURO can be recommended for chronic stroke patients irrespective of the type of stroke.

Optical coherence tomography and coronary angioscopy assessment of healed coronary plaque components.

PURPOSE: Histopathological or intracoronary image assessment of healed plaques (HPs) has been reported. However, the lesion characteristics of HPs remains undetermined yet. We assessed the healed plaque components in patients with coronary artery lesions using multiple imaging modalities. METHODS: We enrolled 33 stable angina pectoris (SAP) patients with 36 native coronary culprit lesions with angiography severe stenosis and without severe calcification undergoing pre-intervention optical coherence tomography (OCT) and coronary angioscopy (CAS). HPs were defined as layered phenotype on OCT. Lesion morphologies and plaque characteristics of HPs were assessed using OCT and CAS. RESULTS: HPs were observed in 19 lesions (52.8%). HP lesions had higher frequent B2/C lesions (89.4% vs. 52.9%, p = 0.02), worse pre-PCI coronary flow (corrected thrombolysis in myocardial infarction count 21.6 ± 13.5 vs. 13.8 ± 6.2, p = 0.047) and greater lumen-area stenosis (79.6 ± 10.6% vs. 68.0 ± 21.6%, p = 0.047) than non-HP lesions. HP lesions had higher prevalence of OCT-thin-cap fibroatheroma (TCFA) (31.6% vs. 0.0%, p = 0.02), OCT-macrophage (89.5% vs. 41.2%, p = 0.004), and CAS-red thrombus (89.5% vs. 41.2%, p = 0.004) than non-HP lesions. The combination of 3 features including OCT-TCFA, macrophages, and CAS-red thrombus showed higher predictive valuer for HPs on OCT than each single variable. Post-PCI irregular tissue protrusion was more frequently observed in lesions with HPs than in those without (52.6% vs. 13.3%, p = 0.03). CONCLUSIONS: SAP lesions with HPs might have more frequent vulnerable plaques with intraplaque inflammation and thrombus than those without, suggesting that layered phenotype on OCT might reflect not only healing process but also potential risks for future coronary events.

Kawasaki disease with acute myocarditis in an adolescent.

A previously healthy 16-year-old adolescent was admitted to our hospital with fever, sore throat, diarrhoea, strawberry tongue, rashes on lower extremities, and pain in the neck, abdomen and joints. He was initially diagnosed with IgA vasculitis triggered by acute pharyngitis. Despite antibiotic treatment, he gradually developed multiorgan failure and cardiac shock. Cardiac catheterisation did not show any signs of coronary artery disease. Subsequently, he developed serious rhabdomyolysis and peripheral extremity necrosis, suggesting peripheral arteritis. Although blood culture results were negative, he received endotoxin adsorption therapy, but it was ineffective. Hence, we suspected Kawasaki disease (KD). We administered high-dose intravenous immunoglobulin therapy (2 g/kg), which was effective. He gradually recovered without major complications and was given ambulatory discharge 43 days after admission. Early diagnosis and administration of intensive care in adolescents with KD with acute myocarditis are critical but challenging.

Empyema with fistula successfully treated with a comprehensive approach including bronchial blocker and embolization receiving veno-venous extracorporeal membrane oxygenation.

BACKGROUND: Complicated empyema accompanied by bronchopleural fistula (BPF) has high mortality. The treatment strategy for severe respiratory failure due to empyema with BPF has yet to be established. CASE PRESENTATION: A 70-year-old man was brought to our hospital and diagnosed with right empyema, BPF (at bronchi B4-10), and secondary left pneumonia. We initiated drainage followed by veno-venous extracorporeal membrane oxygenation due to the severe hypoxia. First, the patient underwent endoscopic treatment with obstructive materials (known as endobronchial Watanabe spigot [EWS]) at B8-10, and was weaned off veno-venous extracorporeal membrane oxygenation on day 7. A secondary EWS was carried out at B4-6. A combination of medical treatments (drainage, antibiotics, nutritional therapy, and rehabilitation) improved his general condition. The patient was able to leave the hospital on foot. CONCLUSION: A comprehensive approach could explain the success of the medical treatment. The principal components are the repeated application of EWS as damage control.

Left ventricular end-systolic contractile entropy can predict cardiac prognosis in patients with complete left bundle branch block.

BACKGROUND: Several patients with complete left bundle branch block (CLBBB) show left ventricular (LV) dyssynchrony and poor cardiac prognosis. However, the prognostic value of LV end-systolic contractile entropy which was measured by single-photon emission computer tomography (SPECT) has not been elucidated in patients with CLBBB. METHODS AND RESULTS: We recruited consecutive 115 sinus-rhythm patients with CLBBB who underwent ECG-gated 201TlCl-SPECT. After 30 days of observation, finally 102 patients (75.2 ± 9.5 years, 62 male) were enrolled and observed retrospectively for a median of 671 days. Twenty-five patients fell into major cardiac events. Multivariate Cox regression analysis showed estimated glomerular filtration rate (eGFR) ≤ 39.35 mL/min and entropy ≥ 79% were significant and independent predictors for major cardiac events (hazard ratio: 4.256 and 7.587, P value = 0.006 and < 0.001, respectively). Machine learning (Random Forest method) revealed eGFR and entropy had higher feature importance than other predictors (0.140 and 0.138, respectively). Kaplan-Meyer curve analysis demonstrated that the group with entropy ≥ 79% and eGFR ≤ 39.36 mL/min had the worst cardiac prognosis (Logrank: P = 0.002). CONCLUSIONS: Left ventricular end-systolic contractile entropy predicts poor cardiac prognosis in patients with CLBBB, which may be more valuable than the other parameters of SPECT.

Association between inflammation in acute phase and early onset pneumonia in patients with out-of-hospital cardiac arrest treated with extracorporeal cardiopulmonary resuscitation.

BACKGROUND: Early onset pneumonia (EOP) in patients with cardiac arrest treated with targeted temperature management is a recently debated issue. We assessed the association between C-reactive protein (CRP) levels and development of EOP in patients treated with extracorporeal cardiopulmonary resuscitation (ECPR). METHODS AND RESULTS: We reviewed the data of all patients admitted to our hospital after out-of-hospital cardiac arrest treated with ECPR between April 2006 and April 2019 who survived for at least 48 h. We collected demographic data, cardiac arrest characteristics, prophylactic antibiotic use, and neurologic outcomes. Diagnosis of EOP was made based on clinical, radiological, and microbiological criteria. The primary endpoint was the association between the incidence of EOP and CRP levels from day 1 to day 4. A total of 55 patients were included, of which 20 developed EOP. CRP levels on days 3 and 4 were significantly elevated in patients who developed EOP (13.1 [11.8-21.1] mg/dL versus 11.6 [7.4-15.2] mg/dL, P = 0.005; and 19.0 [16.9-27.1] mg/dL versus 14.7 [7.4-21.2] mg/dL, P = 0.019, respectively). In the multivariable logistic regression model, the CRP level on day 3 was significantly associated with the development of EOP (odds ratio 1.22; 95% confidence interval 1.06-1.41; P = 0.001). CONCLUSIONS: Increased inflammation in acute phase was associated with development of EOP in patients treated with ECPR.

An endoscopic dilation method using the rendezvous approach for the treatment of severe anastomotic stenosis after rectal cancer surgery: a case report.

BACKGROUND: Postoperative anastomotic stenosis is a common complication in colorectal cancer patients (3-30%). Complete anastomotic stenosis is rare; however, when it occurs, almost all cases require surgical treatment. We herein report a case in which endoscopic dilation was effective for treating complete anastomotic stenosis after high anterior resection in a rectal cancer patient. CASE PRESENTATION: The patient was a 67-year-old man who underwent laparoscopic high anterior resection for rectal cancer (RS, T4a, N0, M0, Stage IIB (TNM Classification of Malignant Tumors)) in May 2018. The postoperative course was good and the patient was discharged on the 12th postoperative day. Subsequently adjuvant chemotherapy was initiated with oral uracil and tegafur plus leucovorin (UFT/LV); however, he complained of frequent defecation and melena after completion of the first course of chemotherapy. Thus, colonoscopy was performed, which revealed anastomotic stenosis. Endoscopic dilation was initially attempted, but failed. Thus, low anterior resection was performed with diverting colostomy. Four additional courses of chemotherapy were administered for 1 month after surgery. At 6 months after the second surgery, colonoscopy was performed, and complete anastomotic stenosis was pointed out again. The patient was successfully treated by endoscopic dilation using the rendezvous method. After this treatment, the lumen of the anastomotic site was observed to have narrowed again and endoscopic dilatation to treat anastomotic stenosis was repeated. In addition, he received local injection of steroids in anastomotic stenosis site. The lumen of anastomotic stenosis remained after the local injection of steroids and closure of colostomy was performed 9 months after the second operation. CONCLUSIONS: Endoscopic dilation using the rendezvous method was effective for treating anastomotic stenosis after colorectal surgery.

Prognostic value of transient conduction disturbance in out-of-hospital cardiac arrest.

AIM: A retrospective observational study to verify the impact of electrocardiograms (ECGs) following out-of-hospital cardiac arrest (OHCA) on mortality. METHODS: We retrospectively studied 101 OHCA patients who achieved a return of spontaneous circulation (ROSC) and survived for ≥3 h. Among them, 50 patients (66 ± 17 years; 22 male) were evaluated using 12-lead ECGs repeatedly and were included in the final analysis: immediately after ROSC (initial ECG) and after the initial evaluation in the emergency department (second ECG). Transient conduction disturbance (transient CD) was defined as a narrowing in QRS duration from the initial to second ECG of ≥18 ms. Multivariate Cox regression analyses were carried out to predict 90-day mortality following OHCA. RESULTS: Among 50 OHCA patients, 30 patients survived for 90 days. Thirty patients had initial ventricular fibrillation rhythm. Median emergency medical services response time and low-flow duration were 8 and 21 min, respectively. Multivariate analysis showed that the transient CD and low-flow duration were significant predictors of all-cause mortality (hazard ratio 16.55, 1.06; P = 0.001, 0.022, respectively). CONCLUSION: Transient CD is a powerful predictor of 90-day mortality in patients who survived 3 h after ROSC from OHCA.

Neuroleptic malignant syndrome in patients with COVID-19.

We report the first two cases of Coronavirus Disease 2019 (COVID-19) who were receiving intensive care including favipiravir, and were clinically diagnosed with neuroleptic malignant syndrome (NMS) to focus attention on NMS in COVID-19 management. Case 1: A 46-year-old-man with acute respiratory distress syndrome (ARDS) caused by COVID-19 infection was being administered favipiravir. Fentanyl, propofol, and rocuronium were also given. On day 3, midazolam administration was initiated for deep sedation. On day 5, his high body temperature increased to 41.2 °C, creatine kinase level elevated, and he developed tachycardia, tachypnea, altered consciousness, and diaphoresis. NMS was suspected, and supportive therapy was initiated. High-grade fever persisted for 4 days and subsided on day 9. Case 2: A 44-year-old-man with ARDS caused by COVID-19 infection was being treated with favipiravir. On day 5, risperidone was started for delirium. On day 7, his body temperature suddenly increased to 40.8 °C, his CK level elevated, and he developed tachycardia, tachypnea, altered consciousness, and diaphoresis. NMS diagnosis was confirmed, and both, favipiravir and risperidone were discontinued on day 8. On the same day, his CK levels decreased, and his body temperature normalized on day 9. Patients with COVID-19 infection frequently require deep sedation and develop delirium; therefore, more attention should be paid to the development of NMS in patients who are being administered such causative agents. The mechanism underlying the occurrence of NMS in COVID-19 patients treated with favipiravir remains unknown. Therefore, careful consideration of NMS development is necessary in the management of COVID-19 patients.

Convolutional neural network for classification of two-dimensional array images generated from clinical information may support diagnosis of rheumatoid arthritis.

This research aimed to study the application of deep learning to the diagnosis of rheumatoid arthritis (RA). Definite criteria or direct markers for diagnosing RA are lacking. Rheumatologists diagnose RA according to an integrated assessment based on scientific evidence and clinical experience. Our novel idea was to convert various clinical information from patients into simple two-dimensional images and then use them to fine-tune a convolutional neural network (CNN) to classify RA or nonRA. We semi-quantitatively converted each type of clinical information to four coloured square images and arranged them as one image for each patient. One rheumatologist modified each patient's clinical information to increase learning data. In total, 1037 images (252 RA, 785 nonRA) were used to fine-tune a pretrained CNN with transfer learning. For clinical data (10 RA, 40 nonRA), which were independent of the learning data and were used as testing data, we compared the classification ability of the fine-tuned CNN with that of three expert rheumatologists. Our simple system could potentially support RA diagnosis and therefore might be useful for screening RA in both specialised hospitals and general clinics. This study paves the way to enabling deep learning in the diagnosis of RA.