Nidhi Prabhakar

Content-based image retrieval for interstitial lung diseases using classification confidence

Content Based Image Retrieval (CBIR) system could exploit the wealth of High-Resolution Computed Tomography (HRCT) data stored in the archive by finding similar images to assist radiologists for self learning and differential diagnosis of Interstitial Lung Diseases (ILDs). HRCT findings of ILDs are classified into several categories (e.g. consolidation, emphysema, ground glass, nodular etc.) based on their texture like appearances. Therefore, analysis of ILDs is considered as a texture analysis problem. Many approaches have been proposed for CBIR of lung images using texture as primitive visual content. This paper presents a new approach to CBIR for ILDs. The proposed approach makes use of a trained neural network (NN) to find the output class label of query image. The degree of confidence of the NN classifier is analyzed using Naive Bayes classifier that dynamically takes a decision on the size of the search space to be used for retrieval. The proposed approach is compared with three simple distance based and one classifier based texture retrieval approaches. Experimental results show that the proposed technique achieved highest average percentage precision of 92.60% with lowest standard deviation of 20.82%.

Comparison of CT Colonography with Conventional Colonoscopy in Patients with Ulcerative Colitis

RATIONALE AND OBJECTIVE Patients with ulcerative colitis require recurrent conventional colonoscopy (CC) to define the extent of the disease. Computed tomography (CT) colonography (CTC) can be used as an alternative technique for studying the colon in these patients. The purpose of the study was to compare the findings of CTC to CC in patients with ulcerative colitis. MATERIALS AND METHODS Twenty patients proven to have ulcerative colitis on biopsy and in clinical remission state were enrolled in the study. They underwent CTC and CC within 1 week of each test. The investigators performing CTC and CC were blinded to the findings of each other. The chi-square test, kappa test, sensitivity, and specificity were used to compare the findings on CTC and CC. In addition, patient acceptance for both the procedures was compared. RESULTS Sensitivity and specificity on CTC for detecting granular appearance were 81.0% and 73.8%, respectively, and for pseudopolyps were 82.1% and 84.5%, respectively. Good correlation was seen between CTC and CC for detection of granular appearance and pseudopolyps. Loss of haustral folds, wall thickening, pericolonic vascularity, and pericolonic lymph nodes seen on CTC were found to correlate with intraluminal findings seen on CC. Patient acceptance for CTC was better than that for CC. CONCLUSIONS CTC can be used for evaluating patients with ulcerative colitis who are in remission.

Wavelet based rotation invariant texture feature for lung tissue classification and retrieval

This paper evaluates the performance of recently proposed rotation invariant texture feature extraction method for the classi¯cation and retrieval of lung tissues a®ected with Interstitial Lung Diseases (ILDs). The method makes use of principle texture direction as the reference direction and extracts texture features using Discrete Wavelet Transform (DWT). A private database containing high resolution computed tomography (HRCT) images belonging to ¯ve category of lung tissue is used for the experiment. The experimental result shows that the texture appearances of lung tissues are anisotropic in nature and hence rotation invariant features achieve better retrieval as well as classi¯cation accuracy.

Hypothalamic glioma masquerading as craniopharyngioma

Hypothalamic glioma account for 10-15% of supratentorial tumors in children. They usually present earlier (first 5 years of age) than craniopharyngioma. Hypothalamic glioma poses a diagnostic dilemma with craniopharyngioma and other hypothalamic region tumors, when they present with atypical clinical or imaging patterns. Neuroimaging modalities especially MRI plays a very important role in scrutinizing the lesions in the hypothalamic region. We report a case of a hypothalamic glioma masquerading as a craniopharyngioma on imaging along with brief review of both the tumors.

Role of perfusion CT in the evaluation of adnexal masses

Abstract The objective of this study was to evaluate the role of perfusion computed tomography (PCT) in differentiating benign from malignant adnexal masses. Twenty patients, each of pathologically proven malignant and benign adnexal masses who had undergone PCT on 64–slice CT scanner, were included in the study. The PCT parameters, viz. blood volume (BV), blood flow (BF), permeability surface index area (PS) and time to maximum of the tissue residue function (Tmax) of the adnexal masses were calculated. Statistical analysis to study the association between PCT parameters and histopathological diagnosis was done. In the malignant group, the mean PS, BV and BF values were elevated. The mean Tmax of the benign lesions was higher compared to that of the malignant lesions. There was a significant statistical difference in the PCT parameters between the malignant and benign groups (p value = .001). PCT can be a useful tool for differentiating benign and malignant adnexal masses. Impact statement What is already known on this subject? It is not always possible to distinguish benign from malignant adnexal lesions despite the application of various imaging techniques. Perfusion CT (PCT) is an imaging technique with which we can obtain both the morphological and functional information of tumours. Perfusion-based imaging enables us to objectively evaluate the neovascularity in a lesion. This helps in differentiating the benign lesions from aggressive malignant lesions. What do the results of this study add? The PCT parameters, viz. blood volume (BV), blood flow (BF), permeability surface index area (PS) and time to maximum of the tissue residue function (Tmax) were calculated from adnexal masses on a 64–multi-slice CT scanner and correlated with their histopathological diagnoses. The values of the mean PS, BV and BF values were significantly higher in the malignant adnexal masses. The mean Tmax in the benign masses was more compared to that of the malignant lesions. Significant statistical difference was seen in PCT parameters between malignant and benign groups. What are the implications of these findings for clinical practice and/or further research? PCT can be a useful tool for differentiating benign from malignant adnexal masses. However, more collaborative research and robust validation are imperative to further evaluate this innovative evolving technique.

B/L Basal Ganglia Lesions in a Child Leading to a Diagnosis of Glucose-6-Phosphate Dehydrogenase Deficiency

Bilateral basal ganglia lesions are a common non-specific finding seen in many diseases. One of the differential diagnoses for it, in a child, is kernicterus occurring due to hyperbilirubinemia. Glucose-6-phosphate dehydrogenase (G6PD) deficiency is a common cause of severe hyperbilirubinemia. A 1-year old child presented to the hospital with history of generalized dystonia in the previous 3 days. MRI showed evidence of symmetrical lesions in bilateral globus pallidus, which were hyperintense on T2/FLAIR and isointense on T1. Patient’s blood test revealed G6PD deficiency. Hence, a diagnosis of G6PD deficiency leading to kernicterus was made. In a child, the diseases that may affect the basal ganglia symmetrically and bilaterally include kernicterus, hypoxia, carbon monoxide poisoning, hypoglycemia, inherited metabolic and dysmyelinating disorders like Leigh disorder, Canavan and Krabbe, Neurofibromatosis, Herpes encephalitis, congenital HIV infection, manganese poisoning and extrapontine myelinolysis. Important causes of kernicterus are Rh incompatibility, ABO incompatibility, sepsis, hemolytic anaemia and G6PD deficiency. G6PD deficiency leading to kernicterus should be considered a differential diagnosis of bilateral basal ganglia lesions in children. Proper elicitation of history with appropriate blood biochemical tests will help in arriving at a proper diagnosis.

An experimental study of interstitial lung tissue classification in HRCT images using ANN and role of cost functions

In this paper, we investigate the effect of the error criteria used during a training phase of the artificial neural network (ANN) on the accuracy of the classifier for classification of lung tissues affected with Interstitial Lung Diseases (ILD). Mean square error (MSE) and the cross-entropy (CE) criteria are chosen being most popular choice in state-of-the-art implementations. The classification experiment performed on the six interstitial lung disease (ILD) patterns viz. Consolidation, Emphysema, Ground Glass Opacity, Micronodules, Fibrosis and Healthy from MedGIFT database. The texture features from an arbitrary region of interest (AROI) are extracted using Gabor filter. Two different neural networks are trained with the scaled conjugate gradient back propagation algorithm with MSE and CE error criteria function respectively for weight updation. Performance is evaluated in terms of average accuracy of these classifiers using 4 fold cross-validation. Each network is trained for five times for each fold with randomly initialized weight vectors and accuracies are computed. Significant improvement in classification accuracy is observed when ANN is trained by using CE (67.27%) as error function compared to MSE (63.60%). Moreover, standard deviation of the classification accuracy for the network trained with CE (6.69) error criteria is found less as compared to network trained with MSE (10.32) criteria.

A study of retrieval accuracy of pulmonary nodules based on external attachment

In this paper, retrieval accuracy of different types of pulmonary nodules is studied. The trainee radiologists could enrich their knowledge using the visual information of the retrieved nodules. In the proposed retrieval system, the pulmonary nodules are segmented using a semi-automated technique. Several 3D features are explored to improve the performance of the proposed retrieval system. A set of relevant shape and texture features is determined for efficient representation of the nodules in the feature space. The proposed CBIR system is validated on a data set of 542 nodules of Lung Image Database Consortium and Image Database Resource Initiative (LIDC/IDRI). The nodules with composite rank of malignancy “1”, “2” are considered as benign and “4”, “5” are considered as malignant. Considering top five retrieved images, the precision of the proposed retrieval system are 84.76%, 80.75%, and 80.34% for well-circumscribed, juxta-pleural, and juxtavascular nodules, respectively.

Mucinous Noncystic Colloid Carcinoma of the Pancreas with Solitary HepaticMetastasis: A Rare Presentation

Mucinous noncystic (colloid) carcinoma (MNCC) of the pancreas is considered an uncommon variant of ductal adenocarcinoma of the pancreas with specific histopathological and molecular features. We report the CT findings of large area of hypoattenuation and intratumoral calcification in a pathologically proven case of mucinous non-cystic carcinoma of the pancreas. MNCC of the pancreas is a distinct entity that needs to be radiologically differentiated from the ductal adenocarcinoma as usually it has more favorable outcome when compared to the rapidly fatal course of ductal adenocarcinomas. But in our index case, it was associated with a large solitary hepatic metastasis with similar calcific foci, which to the best of our knowledge has not been described in the English literature till date.c

Content-Based Image Retrieval (CBIR) Based Computer-Aided Diagnosis (CAD) in Evaluation of Lung Nodules: A New Tool for Self-Learning and To Assist Radiologists in Diagnosing Lung Cancer

Lung cancer is the leading cause of cancer related deaths in general population. Early diagnosis of malignant pulmonary nodule, can improve 5-year survival rate of lung cancer by upto 80%. There is increase in incidentally detected pulmonary nodules with the increased usage of diagnostic imaging modalities especially computed tomography (CT) of chest. Most often, physicians and trainee doctors have to depend on the experienced radiologists to confidently label these nodules as benign or malignant, thereby raising a need for some method, which could help them in self-learning and also could assist radiologists in ruling out malignancy with good certainty and confidence.