A hybrid CNN-Random Forest algorithm for bacterial spore segmentation and classification in TEM images.

We present a new approach to segment and classify bacterial spore layers from Transmission Electron Microscopy (TEM) images using a hybrid Convolutional Neural Network (CNN) and Random Forest (RF) classifier algorithm. This approach utilizes deep learning, with the CNN extracting features from images, and the RF classifier using those features for classification. The proposed model achieved 73% accuracy, 64% precision, 46% sensitivity, and 47% F1-score with test data. Compared to other classifiers such as AdaBoost, XGBoost, and SVM, our proposed model demonstrates greater robustness and higher generalization ability for non-linear segmentation. Our model is also able to identify spores with a damaged core as verified using TEMs of chemically exposed spores. Therefore, the proposed method will be valuable for identifying and characterizing spore features in TEM images, reducing labor-intensive work as well as human bias.

Early Results of Mechanochemical Ablation by Flebogrif and Radiofrequency Ablation in Treating Primary Varicose Veins of Lower Limb.

Introduction Chronic venous insufficiency is a common condition that leads to the development of incompetent great or short saphenous veins (GSV or SSV) resulting in varicose vein development. Conservative management is initially employed for its treatment; however, the varicosities that do not respond to conservative management may require intervention by surgery or endovenous routes. Radiofrequency ablation (RFA) and mechanochemical ablation (MOCA) are the two most common endovascular techniques available for the management of incompetent GSV or SSV. Clarivein and Flebogrif are two devices to treat incompetent GSV or SSV by MOCA. Mechanical ablation is provided by their flexible cutting elements and chemical ablation is provided by polidocanol or sodium tetradecyl sulfate (STS). RFA uses radiofrequency waves to treat venous insufficiency. Therefore, the aim of this study was to determine the early treatment outcome results for incompetent GSV or SSV treated with RFA or MOCA by Flebogrif. Materials and methods This was a retrospective cross-sectional study undertaken at the Radiology Department of Indus Hospital and Health Network. Electronic Medical Records of all the patients who underwent RFA or MOCA for GSV or SSV for venous insufficiency from January 2021 to December 2021 were included. Both male and female patients aged 18 years and above diagnosed with venous insufficiency having Clinical, Etiologic, Anatomic, and Pathophysiologic (CEAP) scores of >1 were included. Statistical Package for Social Sciences (SPSS) v 22 (IBM Corp., Armonk, NY) was used for data entry and analysis. Results 137 patients were included in the present study with a mean age of 53.8 ± 12.1 years. Pre-procedure CEAP score was C3 in 59 (84.4%), C4 in four (5.7%), and C6 in seven (10.0%) patients in patients who underwent RFA, and it was successful in 69 (98.6%) patients. Pre-procedure CEAP score was C3 in 62 (92.5%), C4 in two (3.0%), and C6 in three (4.5%) patients who underwent MOCA, and it was successful in 59 (88.1%) patients. Pain was the most frequent complication observed in both RFA and MOCA. Conclusion RFA has a high success rate as compared to MOCA by Flebogrif in treating incompetent GSV or SSVs.

Comparison of the Treatment Response of Drug-Eluting Bead Transarterial Chemoembolization and Conventional Transarterial Chemoembolization in Patients With Hepatocellular Carcinoma.

Introduction Hepatocellular carcinoma (HCC) is a common primary hepatic cancer. Its early diagnosis can aid in its treatment by curative means such as surgery or ablation. Advanced-stage diagnosis limits these treatment options, and such cases can be treated with transarterial chemoembolization (TACE). Conventional transarterial chemoembolization (cTACE) and drug-eluting bead transarterial chemoembolization (DEB-TACE) are usually used, and follow-up response is evaluated using modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria. This study was done to compare the treatment response of cTACE and DEB-TACE in patients with HCC. Materials and methods A retrospective review of electronic medical records of all patients diagnosed with HCC from January 2021 to August 2022 who underwent cTACE or DEB-TACE was undertaken at the Department of Interventional Radiology, Indus Hospital and Health Network. Both male and female patients aged 18 years or above with Child-Pugh class A and B were included. DEB-TACE or cTACE was performed by a fellowship-trained interventional radiologist, and the response was evaluated at six weeks follow-up using mRECIST criteria. Results A total of 129 patients were included in this study, with a mean age of 54.1 ± 10.8 years. The mean size of HCC was 3.1 ± 1.7 cm. Seventy-eight (60.5%) patients underwent cTACE, and 51 (39.5%) underwent DEB-TACE. Out of the 78 patients who underwent cTACE, complete response (CR) was found in 28 (35.9%), partial response (PR) was found in 33 (42.3%), stable disease (SD) was found in 12 (15.4%), and progressive disease (PD) was found in five (6.4%) patients. Of the 51 patients who underwent DEB-TACE, CR was found in 13 (25.5%), PR was found in 20 (39.2%), SD was found in 11 (21.6%), and PD was found in seven (13.7%) patients. Conclusion The response rate of TACE in the form of complete or partial response was higher with a lower frequency of stable or progressive disease. cTACE has a high response rate as compared to DEB-TACE.

Empiric Transcatheter Arterial Embolization for Massive or Recurrent Gastrointestinal Bleeding: Ten-year Experience from a Single Tertiary Care Center.

Purpose In patients with massive or recurrent gastrointestinal bleeding (GIB) which is not amenable to endoscopic therapy, angiographic interventions are often employed. We report our ten-year experience of empiric transcatheter arterial embolization (TAE) for patients with massive or recurrent GIB. Methods All patients who had undergone empiric TAE at our hospital between March 2004 and June 2015 were identified using the institutional radiology information system. A retrospective chart review was performed using a structured pro forma. Technical success rate, 30-day clinical success rate, 30-day mortality rate, and rate of procedural complications were computed. Statistical analysis was performed using Statistical Package for Social Sciences (SPSS) version 20. Results A total of 32 patients had undergone empiric TAE for GIB during the study period. The median age of subjects was 56 years and two-thirds of them were male (68.7%). Gastroduodenal (n=24), ileocolic (n=3), left gastric (n=2), right gastroepiploic (n=1), and branches of superior and middle rectal arteries (n=1) were embolized using microcoils (n=25), polyvinyl alcohol particles (n=25), and gelatin sponge (n=3)--either alone or in combination. Technical and 30-day clinical success rates were 96.9% (31/32) and 71.9% (23/32), respectively. The 30-day mortality rate for our cohort was 21.9% (7/32). One patient developed re-bleeding at two days after the initial procedure and required repeat embolization. Coil migration (n=3) and access site hematoma (n=1) were the observed procedural complications. Conclusion Empiric TAE can be a useful treatment option for selected patients with massive or recurrent GIB that is not amenable to endoscopic therapy.

Accuracy of CT chest without oral contrast for ruling out esophageal perforation using fluoroscopic esophagography as reference standard: a retrospective study.

PURPOSE: Esophageal perforation has a high mortality rate. Fluoroscopic esophagography (FE) is the procedure of choice for diagnosing esophageal perforation. However, FE can be difficult to perform in seriously ill patients. METHODS: We retrospectively reviewed charts and scans of all patients who had undergone thoracic CT (TCT) without oral contrast and FE for suspicion of esophageal perforation at our hospital between October, 2010 and December, 2015. Scans were interpreted by a single consultant radiologist having > 5 years of relevant experience. Statistical analysis was performed using SPSS version 20. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of TCT were computed using FE as reference standard. RESULTS: Of 122 subjects, 106 (83%) were male and their median age was 42 [inter-quartile range (IQR) 29-53] years. Esophageal perforation was evident on FE in 15 (8%) cases. Sensitivity, specificity, PPV and NPV of TCT for detecting esophageal perforation were 100, 54.6, 23.4 and 100%, respectively. When TCT was negative (n = 107), an alternative diagnosis was evident in 65 cases. CONCLUSION: Thoracic computed tomography (TCT) had 100% sensitivity and negative predictive value for excluding esophageal perforation. FE may be omitted in patients who have no evidence of mediastinal collection, pneumomediastinum or esophageal wall defect on TCT. However, in the presence of any of these features, FE is still necessary to confirm or exclude the presence of an esophageal perforation.

Diagnostic validity of ultrasonography in thyroid nodules.

OBJECTIVE: To determine the diagnostic validity of thyroid ultrasound in differentiating between benign and malignant thyroid nodules. METHODS: The cross-sectional study was conducted at Aga Khan University Hospital, Karachi, from August 2011 to January 2013, and comprised all patients of either gender with thyroid nodules referred for ultrasound thyroid and fine needle aspiration cytology. Ultrasonography was performed by radiologists and ultrasound parameters were assessed and compared with cytology results in all nodules. Diagnostic validity of each ultrasound feature was calculated. RESULTS: Of the 101 patients, 81(80%) were female. The overall mean age was 43±13 years (range: 15-73 years). On histocytopathology, 96(95%) nodules were benign and 5(4.9%) were malignant. The sensitivity and specificity of ultrasound features in predicting malignancy were calcification 80% and 68%; hypoechogenecity 80% and 52%; ill-defined lobulated margin 40% and 96%; solid 80% and 40%; taller than wider 50% and 63%. Each ultrasound feature had negative predictive value ranging from 95% to 98% in malignant nodules. CONCLUSIONS: Identification of calcification, hypoechogenecity and solid with ill-defined margins in a thyroid nodule on ultrasound was helpful in suspecting thyroid malignancy and warranted urgent diagnostic fine needle aspiration cytology.