High-resolution CT findings of pulmonary infections in patients with hematologic malignancy: comparison between patients with or without hematopoietic stem cell transplantation

Adoption value of support vector machine algorithm-based computed tomography imaging in the diagnosis of secondary pulmonary fungal infections in patients with malignant hematological disorders

This study aimed to assess the feasibility of diagnosing secondary pulmonary fungal infections (PFIs) in patients with hematological malignancies (HM) using computerized tomography (CT) imaging and a support vector machine (SVM) algorithm. A total of 100 patients with HM complicated by secondary PFI underwent CT scans, and they were included in the training group. Concurrently, 80 patients with the same underlying disease who were treated at our institution were included in the test group. The types of pathogens among different PFI patients and the CT imaging features were compared. Radiomic features were extracted from the CT imaging data of patients, and a diagnostic SVM model was constructed by integrating these features with clinical characteristics. Aspergillus was the most common pathogen responsible for PFIs, followed by Candida, Pneumocystis jirovecii, Mucor, and Cryptococcus, in descending order of occurrence. Patients typically exhibited bilateral diffuse lung lesions. Within the SVM algorithm model, six radiomic features, namely the square root of the inverse covariance of the gray-level co-occurrence matrix (square root IV), the square root of the inverse covariance of the gray-level co-occurrence matrix, and small dependency low gray-level emphasis, significantly influenced the diagnosis of secondary PFIs in patients with HM. The area under the curve values for the training and test sets were 0.902 and 0.891, respectively. Therefore, CT images based on the SVM algorithm demonstrated robust predictive capability in diagnosing secondary PFIs in conjunction with HM.

New trend in artificial intelligence-based assistive technology for thoracic imaging

Although there is no solid agreement for artificial intelligence (AI), it refers to a computer system with intelligence similar to that of humans. Deep learning appeared in 2006, and more than 10 years have passed since the third AI boom was triggered by improvements in computing power, algorithm development, and the use of big data. In recent years, the application and development of AI technology in the medical field have intensified internationally. There is no doubt that AI will be used in clinical practice to assist in diagnostic imaging in the future. In qualitative diagnosis, it is desirable to develop an explainable AI that at least represents the basis of the diagnostic process. However, it must be kept in mind that AI is a physician-assistant system, and the final decision should be made by the physician while understanding the limitations of AI. The aim of this article is to review the application of AI technology in diagnostic imaging from PubMed database while particularly focusing on diagnostic imaging in thorax such as lesion detection and qualitative diagnosis in order to help radiologists and clinicians to become more familiar with AI in thorax.

Feasibility of thoracic CT in assessing anemia for aplastic anemia patients undergoing allogeneic hematopoietic stem cell transplantation

BACKGROUND

Anemia is an important clinical symptom for aplastic anemia (AA) patients who are suffered with peripheral pancytopenia.

OBJECTIVE

To evaluate the accuracy of diagnosing anemia with non-invasive chest computed tomography (CT) for AA patients.

METHODS

The CT attenuation of left ventricular (LV) cavity and interventricular septum (IVS) on unenhanced thoracic CT images of AA patients are retrospectively analyzed, including 84 AA patients in pre-transplant and 1-month (n = 82), 2-month (n = 72), 3-month (n = 75), 6-month (n = 74) and 12-month (n = 70) followed patients in post-transplant. The difference (IVS-LV) and ratio (LV/IVS) of the CT attenuation between LV cavity and interventricular septum are calculated. Serum hemoglobin is estimated within 24 hours of CT imaging. The CT attenuations of IVS-LV and LV/IVS are correlated with hemoglobin, and their variation tendency is analyzed during the treatment of a-HSCT. A receiver operating characteristic (ROC) curve analysis is then performed for the diagnosis of anemia.

RESULTS

The CT attenuations of IVS-LV and LV/IVS well correlate with hemoglobin (r = -0.618 and 0.628, respectively, P <  0.001). The variation tendency of IVS-LV and LV/IVS is similar to that of hemoglobin with opposite directions during one-year follow-up of a-HSCT. When a threshold of CT attenuation of IVS-LV and LV/IVS is set at 11.5HU and 0.77, respectively, both the sensitivity and specificity in diagnosing anemia are good (74.7% and 73.8% in CT attenuation of IVS-LV; 77.4% and 70.4% in LV/LVS, respectively).

CONCLUSIONS

Both CT attenuation of LV/IVS and IVS-LV had similar accuracy in diagnosing anemia for AA patients. The non-invasive chest CT can offer a new possibility to complementarily evaluate anemia for AA patients in the diagnostic radiology reports.