Automating classification of osteoarthritis according to Kellgren-Lawrence in the knee using deep learning in an unfiltered adult population.

BACKGROUND: Prevalence for knee osteoarthritis is rising in both Sweden and globally due to increased age and obesity in the population. This has subsequently led to an increasing demand for knee arthroplasties. Correct diagnosis and classification of a knee osteoarthritis (OA) are therefore of a great interest in following-up and planning for either conservative or operative management. Most orthopedic surgeons rely on standard weight bearing radiographs of the knee. Improving the reliability and reproducibility of these interpretations could thus be hugely beneficial. Recently, deep learning which is a form of artificial intelligence (AI), has been showing promising results in interpreting radiographic images. In this study, we aim to evaluate how well an AI can classify the severity of knee OA, using entire image series and not excluding common visual disturbances such as an implant, cast and non-degenerative pathologies. METHODS: We selected 6103 radiographic exams of the knee taken at Danderyd University Hospital between the years 2002-2016 and manually categorized them according to the Kellgren & Lawrence grading scale (KL). We then trained a convolutional neural network (CNN) of ResNet architecture using PyTorch. We evaluated the results against a test set of 300 exams that had been reviewed independently by two senior orthopedic surgeons who settled eventual interobserver disagreements through consensus sessions. RESULTS: The CNN yielded an overall AUC of more than 0.87 for all KL grades except KL grade 2, which yielded an AUC of 0.8 and a mean AUC of 0.92. When merging adjacent KL grades, all but one group showed near perfect results with AUC > 0.95 indicating excellent performance. CONCLUSION: We have found that we could teach a CNN to correctly diagnose and classify the severity of knee OA using the KL grading system without cleaning the input data from major visual disturbances such as implants and other pathologies.

Artificial intelligence for the classification of fractures around the knee in adults according to the 2018 AO/OTA classification system.

BACKGROUND: Fractures around the knee joint are inherently complex in terms of treatment; complication rates are high, and they are difficult to diagnose on a plain radiograph. An automated way of classifying radiographic images could improve diagnostic accuracy and would enable production of uniformly classified records of fractures to be used in researching treatment strategies for different fracture types. Recently deep learning, a form of artificial intelligence (AI), has shown promising results for interpreting radiographs. In this study, we aim to evaluate how well an AI can classify knee fractures according to the detailed 2018 AO-OTA fracture classification system. METHODS: We selected 6003 radiograph exams taken at Danderyd University Hospital between the years 2002-2016, and manually categorized them according to the AO/OTA classification system and by custom classifiers. We then trained a ResNet-based neural network on this data. We evaluated the performance against a test set of 600 exams. Two senior orthopedic surgeons had reviewed these exams independently where we settled exams with disagreement through a consensus session. RESULTS: We captured a total of 49 nested fracture classes. Weighted mean AUC was 0.87 for proximal tibia fractures, 0.89 for patella fractures and 0.89 for distal femur fractures. Almost ¾ of AUC estimates were above 0.8, out of which more than half reached an AUC of 0.9 or above indicating excellent performance. CONCLUSION: Our study shows that neural networks can be used not only for fracture identification but also for more detailed classification of fractures around the knee joint.

Argemone mexicana extract alleviates gastrointestinal disorders by stimulating muscarinic receptors and blocking voltage-gated L-type calcium channels

Objective: To investigate the pharmacological potential of Argemone mexicana in treating constipation and emesis by using in vitro and in vivo models. Methods: The spasmogenic and spasmolytic effects were evaluated on isolated rabbit jejunum fragments loaded in a tissue organ bath. The response was recorded with an isotonic transducer attached with Power Lab Data Acquisition System. The laxative and antiemetic activities were assessed in BALB-c mice and poultry chicks challenged with carbamylcholine and copper sulphate stimulated emesis, respectively. Results: The total phenolic and total flavonoids contents of the extract were (267.75 ± 5.77) mg GAE/g and (73.86 ± 6.01) mg QE/g, respectively. Argemone mexicana extract exerted spasmogenic effect on isolated rabbit jejunum segments with an EC50 value of 0.016 mg/mL, which was blocked by atropine (0.3 μM). Argemone mexicana extract exerted spasmolytic effect in atropine treated jejunum fragments with an EC50 value of 2.185 mg/mL. Furthermore, Argemone mexicana extract relaxed potassium (80 mM)-induced contractions (EC50: 9.07 mg/mL), similar to a standard drug verapamil. The calcium channel blocker activity was confirmed by a rightward shift of concentration-response curve of calcium in the presence of Argemone mexicana extract (1-5 mg/mL) and verapamil (0.1-1 μM). In addition, the extract increased the distance travelled by a charcoal in the gastrointestinal tract and exhibited antiemetic effect on copper sulphate induced emesis in chicks. Conclusions: Argemone mexicana shows cholinergic agonist and calcium channel blocker activities, as well as antiemetic effect. It may be used as a potential agent for treating gastrointestinal disorders.

Tylophora hirsuta L. leaf extract attenuates alloxan-induced diabetes in mice by suppressing oxidative stress and α-Amylase

Objective: To evaluate the antidiabetic potential of leaf extracts of Tylophora hirsuta (T. hirsuta). Methods: The methanolic and ethyl acetate extracts of T. hirsuta leaves were analyzed by high pressure liquid chromatography. In vitro antioxidant activity was determined by ferric ion reduction, 1, 1-diphenyl-2-picrylhydrazyl, and hydrogen peroxide scavenging methods. In vitro alpha amylase (α-Amylase) inhibitory activity of the plant extracts was assessed. In vivo antidiabetic potential was determined in alloxan-induced diabetic mice to assess glycated hemoglobin (HbA1c), oral glucose tolerance, serum amylase, lipid profile, fasting blood glucose, and body weight. Histopathological lesions of the pancreas, liver and kidney were observed. Oxidative stress biomarkers such as superoxide dismutase, catalase and peroxidase were also determined. Results: Quercetin, chlorogenic acid, p-coumaric acid, and m-coumaric acid were found in the plant extracts. The methanolic plant extract exhibited higher in vitro antioxidant activities than the ethyl acetate extract. Moreover, methanolic plant extract exhibited (83.90±1.56)% α-Amylase inhibitory activity at 3.2 mg/ mL concentration. Animal study showed that the methanolic extract of T. hirsuta improved the levels of fasting blood glucose, HbA1c, serum α-Amylase, lipid profile, liver function biomarkers, and kidney functions of diabetic mice. Moreover, the methanolic extract ameliorated diabetes-related oxidative stress by increasing superoxide dismutase and catalase activities and decreasing peroxidase and malondialdehyde levels. Histopathological examination showed that the plant extract had improved the integrity of pancreatic islets of Langerhans and reduced the pathological lesions in the liver and kidney of diabetic mice. Conclusions: The methanolic extract of T. hirsuta exhibits pronounced antidiabetic activity in mice through reduction of oxidative stress. The plant extract has several natural antioxidants such as phenolic acids. T. hirsuta extract could serve as a nutraceutical for managing diabetes mellitus.

Argemone mexicana extract alleviates gastrointestinal disorders by stimulating muscarinic receptors and blocking voltage-gated L-type calcium channels

Objective: To investigate the pharmacological potential of Argemone mexicana in treating constipation and emesis by using in vitro and in vivo models. Methods: The spasmogenic and spasmolytic effects were evaluated on isolated rabbit jejunum fragments loaded in a tissue organ bath. The response was recorded with an isotonic transducer attached with Power Lab Data Acquisition System. The laxative and antiemetic activities were assessed in BALB-c mice and poultry chicks challenged with carbamylcholine and copper sulphate stimulated emesis, respectively. Results: The total phenolic and total flavonoids contents of the extract were (267.75 ± 5.77) mg GAE/g and (73.86 ± 6.01) mg QE/g, respectively. Argemone mexicana extract exerted spasmogenic effect on isolated rabbit jejunum segments with an EC50 value of 0.016 mg/mL, which was blocked by atropine (0.3 μM). Argemone mexicana extract exerted spasmolytic effect in atropine treated jejunum fragments with an EC50 value of 2.185 mg/mL. Furthermore, Argemone mexicana extract relaxed potassium (80 mM)-induced contractions (EC50: 9.07 mg/mL), similar to a standard drug verapamil. The calcium channel blocker activity was confirmed by a rightward shift of concentration-response curve of calcium in the presence of Argemone mexicana extract (1-5 mg/mL) and verapamil (0.1-1 μM). In addition, the extract increased the distance travelled by a charcoal in the gastrointestinal tract and exhibited antiemetic effect on copper sulphate induced emesis in chicks. Conclusions: Argemone mexicana shows cholinergic agonist and calcium channel blocker activities, as well as antiemetic effect. It may be used as a potential agent for treating gastrointestinal disorders.

Tylophora hirsuta L. leaf extract attenuates alloxan-induced diabetes in mice by suppressing oxidative stress and α-Amylase

Objective: To evaluate the antidiabetic potential of leaf extracts of Tylophora hirsuta (T. hirsuta). Methods: The methanolic and ethyl acetate extracts of T. hirsuta leaves were analyzed by high pressure liquid chromatography. In vitro antioxidant activity was determined by ferric ion reduction, 1, 1-diphenyl-2-picrylhydrazyl, and hydrogen peroxide scavenging methods. In vitro alpha amylase (α-Amylase) inhibitory activity of the plant extracts was assessed. In vivo antidiabetic potential was determined in alloxan-induced diabetic mice to assess glycated hemoglobin (HbA1c), oral glucose tolerance, serum amylase, lipid profile, fasting blood glucose, and body weight. Histopathological lesions of the pancreas, liver and kidney were observed. Oxidative stress biomarkers such as superoxide dismutase, catalase and peroxidase were also determined. Results: Quercetin, chlorogenic acid, p-coumaric acid, and m-coumaric acid were found in the plant extracts. The methanolic plant extract exhibited higher in vitro antioxidant activities than the ethyl acetate extract. Moreover, methanolic plant extract exhibited (83.90±1.56)% α-Amylase inhibitory activity at 3.2 mg/ mL concentration. Animal study showed that the methanolic extract of T. hirsuta improved the levels of fasting blood glucose, HbA1c, serum α-Amylase, lipid profile, liver function biomarkers, and kidney functions of diabetic mice. Moreover, the methanolic extract ameliorated diabetes-related oxidative stress by increasing superoxide dismutase and catalase activities and decreasing peroxidase and malondialdehyde levels. Histopathological examination showed that the plant extract had improved the integrity of pancreatic islets of Langerhans and reduced the pathological lesions in the liver and kidney of diabetic mice. Conclusions: The methanolic extract of T. hirsuta exhibits pronounced antidiabetic activity in mice through reduction of oxidative stress. The plant extract has several natural antioxidants such as phenolic acids. T. hirsuta extract could serve as a nutraceutical for managing diabetes mellitus.

Artificial intelligence for analyzing orthopedic trauma radiographs.

Background and purpose - Recent advances in artificial intelligence (deep learning) have shown remarkable performance in classifying non-medical images, and the technology is believed to be the next technological revolution. So far it has never been applied in an orthopedic setting, and in this study we sought to determine the feasibility of using deep learning for skeletal radiographs. Methods - We extracted 256,000 wrist, hand, and ankle radiographs from Danderyd's Hospital and identified 4 classes: fracture, laterality, body part, and exam view. We then selected 5 openly available deep learning networks that were adapted for these images. The most accurate network was benchmarked against a gold standard for fractures. We furthermore compared the network's performance with 2 senior orthopedic surgeons who reviewed images at the same resolution as the network. Results - All networks exhibited an accuracy of at least 90% when identifying laterality, body part, and exam view. The final accuracy for fractures was estimated at 83% for the best performing network. The network performed similarly to senior orthopedic surgeons when presented with images at the same resolution as the network. The 2 reviewer Cohen's kappa under these conditions was 0.76. Interpretation - This study supports the use for orthopedic radiographs of artificial intelligence, which can perform at a human level. While current implementation lacks important features that surgeons require, e.g. risk of dislocation, classifications, measurements, and combining multiple exam views, these problems have technical solutions that are waiting to be implemented for orthopedics.

Factors of Transferability for a Generic ConvNet Representation.

Evidence is mounting that Convolutional Networks (ConvNets) are the most effective representation learning method for visual recognition tasks. In the common scenario, a ConvNet is trained on a large labeled dataset (source) and the feed-forward units activation of the trained network, at a certain layer of the network, is used as a generic representation of an input image for a task with relatively smaller training set (target). Recent studies have shown this form of representation transfer to be suitable for a wide range of target visual recognition tasks. This paper introduces and investigates several factors affecting the transferability of such representations. It includes parameters for training of the source ConvNet such as its architecture, distribution of the training data, etc. and also the parameters of feature extraction such as layer of the trained ConvNet, dimensionality reduction, etc. Then, by optimizing these factors, we show that significant improvements can be achieved on various (17) visual recognition tasks. We further show that these visual recognition tasks can be categorically ordered based on their similarity to the source task such that a correlation between the performance of tasks and their similarity to the source task w.r.t. the proposed factors is observed.

Iron overload in allogeneic hematopoietic stem cell transplant recipients.

CONTEXT: Patients who undergo hematopoietic stem cell transplant are at an increased risk of developing iron overload. OBJECTIVES: To describe the effect of hepatic iron overload on hematopoietic stem cell transplant recipients and to validate the utility of histologic scoring system of iron granules in the liver. DESIGN: Records of 154 post allogeneic hematopoietic stem cell transplant patients were reviewed. Forty-nine patients underwent liver biopsy. Histologic hepatic iron overload was defined as a score of 2 or greater (scale, 0-4). RESULTS: Twenty-eight of 49 patients (57%) evaluated by liver biopsy had hepatic iron overload; 17 had moderate to severe hepatic iron overload (score, 3 or 4). In multivariate analysis, a significant correlation was discovered between hepatic iron overload and the number of transfusions (P < .001), posttransplant serum ferritin levels (P=.004), lactate dehydrogenase levels (P=.03), and the development of blood stream infections (P= .02). There was no correlation between hepatic iron overload and abnormal liver function test results. While 37 patients (76%) died after receiving a transplant, mortality was not influenced by hepatic iron overload but was significantly higher in older patients, in patients with lower serum albumin levels, higher serum bilirubin levels, and higher clinical grade of acute graft-versus-host disease (P=.04, P=.001, P=<.001, and P .004, respectively). CONCLUSIONS: Hepatic iron overload is commonly identified in hematopoietic stem cell transplant patients and can be accurately diagnosed by liver biopsy. In addition, hepatic iron overload has been identified in patients receiving as few as 25 units of packed red blood cells, with elevated posttransplant serum ferritin levels, and with blood stream infections.

Significance of complement split product C4d in ABO-compatible liver allograft: diagnosing utility in acute antibody mediated rejection.

Diagnosis of liver allograft antibody-mediated rejection (AMR) is difficult and requires a constellation of clinical, laboratory and histologic features that support the disease and exclude other causes. Histologic features of AMR may intermix with those of biliary obstruction, preservation/reperfusion injury, and graft ischemia. Tissue examination for complement degradation product 4d (C4d) has been proved to support this diagnosis in other allografts. For this reason, we conducted a retrospective review of all ABO compatible/identical re-transplanted liver patients with primary focus on identifying AMR as a possible cause of graft failure and to investigate the utility of C4d in liver allograft specimens. We reviewed 193 liver samples obtained from 53 consecutive ABO-compatible re-transplant patients. 142 specimens were stained with C4d. Anti-donor antibody screening and identification was determined by Luminex100 flow cytometry. For the study analysis, patients were stratified into 3 groups according to time to graft failure: group A, patients with graft failure within 0-7 days (n=7), group B within 8-90 days (n=13) and C >90 days (n=33). Two patients (3.7%) met the diagnostic criteria of acute AMR. Both patients experienced rapid decline of graft function with presence of donor specific antibodies (DSA), morphologic evidence of humoral rejection and C4d deposition in liver specimens. C4d-positive staining was identified in different medical liver conditions i.e., acute cellular rejection (52%), chronic ductopenic rejection (50%), recurrent liver disease (48%), preservation injury (18%), and hepatic necrosis (54%). Univariate analysis showed no significant difference of C4d-positive staining among the 3 patients groups, or patients with DSA (P>.05). In conclusion, AMR after ABO-compatible liver transplantation is an uncommon cause of graft failure. Unlike other solid organ allografts, C4d-positive staining is not a rugged indicator of humoral rejection, thus, interpretation should be done with caution to avoid diagnostic dilemmas.

Naproxen-induced liver injury.

BACKGROUND: Nonsteroidal anti-inflammatory drugs (NSAIDs) have been reported to induce liver injury. Patterns of the injury usually range from mild elevations of liver enzymes to sometimes severe fulminant hepatic failure. Likewise, naproxen is a propionic acid derivative NSAID that was introduced in 1980 and has been available as an over-the-counter medication since 1994, but has rarely been reported to cause liver injury. METHODS: We treated a 30-year-old woman with jaundice and intractable pruritus that developed shortly after taking naproxen. We reviewed the medical history and liver histopathology of the patient as well as all previously published case reports of naproxen-associated liver toxicity in the English language literature. RESULTS: The liver biochemical profile of the patient revealed a mixed cholestasis and hepatitis pattern. Consecutive liver biopsies demonstrated focal lobular inflammation, hepatocyte drop-out, and a progressive loss of the small interlobular bile ducts (ductopenia). The biopsy performed two years after onset of the disease showed partial recovery of a small number of bile ducts; however, 10 years passed before the biochemical profile returned to near normal. CONCLUSIONS: Naproxen-associated liver toxicity remains a rare entity, but should be considered in any patient presenting with cholestasis shortly after its use. Liver injury is most commonly seen in a mixed pattern characterized by cholestasis and hepatitis. The resulting liver damage may take years to resolve.

Ductal carcinoma in situ of the breast: MR imaging findings with histopathologic correlation.

Ductal carcinoma in situ (DCIS) is a noninvasive malignancy that is commonly encountered at routine breast imaging. It may be a primary tumor or may be seen in association with other focal higher-grade tumors. Early detection is important because of the large proportion of DCIS that can progress to invasive carcinoma. The extent of DCIS involvement is frequently underestimated at mammography, which can reliably help detect only calcified DCIS; consequently, magnetic resonance (MR) imaging evaluation can alter the course of treatment. Seven biopsy-proved cases of DCIS were evaluated with T2-weighted MR imaging sequences, as well as T1-weighted sequences performed both before and after contrast material administration. The signal intensity and enhancement patterns of the tumors were analyzed, and the findings were correlated with the relevant underlying histopathologic features. Common enhancement patterns of DCIS include clumped linear-ductal enhancement, clumped focal enhancement, and masslike enhancement. The most common enhancement distribution pattern is segmental, followed by focal, diffuse, linear-ductal, and regional patterns. At T2-weighted MR imaging, DCIS is typically isointense relative to breast parenchyma; less commonly, it is hypointense or hyperintense. The use of MR imaging in the evaluation of DCIS is controversial, and many questions remain with regard to treatment and management. However, breast MR imaging can be extremely useful in the preoperative diagnosis and evaluation of DCIS when used in conjunction with other imaging modalities.

Small-duct primary sclerosing cholangitis with hepatocellular carcinoma requiring liver transplantation.

BACKGROUND: Primary sclerosing cholangitis (PSC) is a chronic progressive cholestatic liver disease, which usually affects young adults and is diagnosed by cholangiography. On a few occasions, the disease either starts in or exclusively involves the small intrahepatic bile ducts, referred to as small-duct PSC. METHODS: A 31-year-old man presented with severe hematemesis secondary to liver cirrhosis. Over a course of 8 years, his liver decompensated and required an orthotopic liver transplantation. In this report we discuss his disease presentation, course of management, and the post-transplantation course of management, and review the morphologic diagnosis, and differential diagnosis of the disease with large-duct type and other diseases that involve small intrahepatic bile ducts. RESULTS: The patient's explanted liver showed changes of PSC affecting only the small- and medium-sized bile ducts in addition to three incidental nodules of hepatocellular carcinoma. CONCLUSIONS: Small-duct PSC has a substantially better prognosis than the large-duct type, with less chance of developing cirrhosis and an equal risk for developing hepatocellular carcinoma, but no increased risk for developing cholangiocarcinoma. Treatment seems to help relieve the symptoms but not necessarily improve survival. Liver transplantation remains the ultimate cure.

Frequent expression of C4d in hepatic graft-versus-host disease: potential clue for diagnosis and distinguishing acute and chronic form.

BACKGROUND: Graft-versus-host disease (GVHD), a common complication of hematopoietic stem cell transplant, is generally regarded to develop through cell-mediated immune response following activation of helper T cells. Since production of antibodies is also mediated by helper T cells, the role of humoral immunity in GVHD is questioned and has not yet been explored in clinical practice. We conducted a pilot study to evaluate the role of antibody production in hepatic H-GVHD and whether it can distinguish acute and chronic forms. RESULTS: C4d expression was increased in portal vessels and hepatic sinusoids of patients with histological proven evidence of GVHD 11/16 (P=0.007). Patients classified as chronic GVHD were statistically more likely to have C4d expression in the portal vasculature and liver sinusoids (P=0.011). CONCLUSION: Humoral activation seems to play a role in pathophysiology of hepatic, especially chronic GVHD.