Can the generalizability issue of artificial intelligence be overcome? Pneumothorax detection algorithm.

The generalizability of artificial intelligence (AI) models is a major issue in the field of AI applications. Therefore, we aimed to overcome the generalizability problem of an AI model developed for a particular center for pneumothorax detection using a small dataset for external validation. Chest radiographs of patients diagnosed with pneumothorax (n = 648) and those without pneumothorax (n = 650) who visited the Ankara University Faculty of Medicine (AUFM; center 1) were obtained. A deep learning-based pneumothorax detection algorithm (PDA-Alpha) was developed using the AUFM dataset. For implementation at the Health Sciences University (HSU; center 2), PDA-Beta was developed through external validation of PDA-Alpha using 50 radiographs with pneumothorax obtained from HSU. Both PDA algorithms were assessed using the HSU test dataset (n = 200) containing 50 pneumothorax and 150 non-pneumothorax radiographs. We compared the results generated by the algorithms with those of physicians to demonstrate the reliability of the results. The areas under the curve for PDA-Alpha and PDA-Beta were 0.993 (95% confidence interval (CI): 0.985-1.000) and 0.986 (95% CI: 0.962-1.000), respectively. Both algorithms successfully detected the presence of pneumothorax on 49/50 radiographs; however, PDA-Alpha had seven false-positive predictions, whereas PDA-Beta had one. The positive predictive value increased from 0.525 to 0.886 after external validation (p = 0.041). The physicians' sensitivity and specificity for detecting pneumothorax were 0.585 and 0.988, respectively. The performance scores of the algorithms were increased with a small dataset; however, further studies are required to determine the optimal amount of external validation data to fully address the generalizability issue.

Interstitial lung disease associated with chronic liver disease.

It is important to make the differential diagnosis of restrictive changes associated with hepatic hydrothorax or hepatopulmonary syndrome seen in the later stages of chronic liver diseases and restrictive changes associated with interstitial lung disease. Lymphocytic interstitial pneumonia (LIP) is in the rare idiopathic interstitial pneumonia subgroup of interstitial lung diseases. LIP is a rare disease, and its incidence is unknown. LIP is characterized by infiltration of the alveolar interstitium with lymphocytes, plasma cells, and histiocytes. The etiology of LIP includes idiopathic causes, rheumatological diseases, immune deficiencies, viral infections, and drug-related causes. Chronic liver diseases are also rarely included in the etiology of LIP. A 75-year-old male patient who was followed up for liver cirrhosis presented with dyspnea. He had hypoxemia in the arterial blood gas. In the thorax and abdominal computed tomography, irregular reticulations in bilateral lungs, ground-glass opacities, and scattered air cysts in both lung parenchyma, chronic liver parenchymal disease, splenomegaly, chronic portal vein thrombosis were determined. Clinical and radiological changes in the patient were evaluated in favor of interstitial lung disease. Although histopathological diagnosis could not be made, the patient whose radiological pattern was compatible with LIP was evaluated together with clinical findings and was accepted as lymphocytic interstitial pneumonia. He was evaluated in terms of diseases that could cause LIP. He was accepted as LIP due to chronic liver disease. Although histopathological examination is the gold standard for the diagnosis, a biopsy could not be performed in our case. Radiological and clinical findings were considered sufficient for the diagnosis of LIP. Chronic viral hepatitis and cirrhosis are also present in the etiology of LIP. Our case is presented as an example in the literature because it is a case of LIP due to chronic liver disease, and it is rare.

Successful treatment of COVID-19 infection in a patient with tracheostomy.

Coronavirus disease 2019 (COVID-19) has been demonstrated to be the cause of emerging atypical pneumonia. In patients with tracheostomy, coronavirus hypothetically coexists with well-known bacterial agents. A 61-year-old male patient with tracheostomy was admitted to the hospital with dyspnea, fever and increased tracheal secretions. Laboratory findings revealed lymphopenia and elevated C-reactive protein and procalcitonin levels. Chest computed tomography showed consolidation areas and ground-glass opacities more prominent in subpleural areas. Although; two consecutive RT-PCR analyses of combined nasopharengeal/oropharengeal swabs were found to be negative for SARS-CoV-2 RNA, positivity was reported for endotracheal aspirate (ETA) sample. Significant growth of Pseudomonas aeruginosa and Stenotrophomonas maltophilia was detected in the bacterial culture of ETA sample. In conclusion, clinical samples for SARS-CoV-2 should be obtained through the lower respiratory tract, if possible and if upper airway samples are negative. To the best our knowledge, our paper is the first report of the patient with tracheostomy who was treated successfully for COVID-19.