Spleen Segmentation and Assessment in CT Images for Traumatic Abdominal Injuries

Abstract

Spleen segmentation is especially challenging as the majority of solid organs in the abdomen region have similar gray level range. Physician analysis of computed tomography (CT) images to assess abdominal trauma could be very time consuming and hence, automating this process can reduce time to treatment. The proposed method presented in this paper is a fully automated and knowledge based technique that employs anatomical information to accurately segment the spleen in CT images. The spleen detection procedure is proposed to locate the spleen in both healthy and injured cases. In the presence of hemorrhage and laceration, the edge merging technique is used. The accuracy of the method is measured by some criteria such as mis-segmented area, accuracy, specificity and sensitivity. The results show that the proposed spleen segmentation method performs well and outperforms other methods.

Current Applications and Future Impact of Machine Learning in Radiology

“…For instance, the extraction of incidental findings such as pulmonary and thyroid nodules (20)(21)(22) has been demonstrated to be possible with machine learning techniques. Further machine learning research has also been performed for detection of critical findings such as pneumothorax ( Fig 6), fractures, organ laceration, and stroke (23)(24)(25)(26)(27)(28)(29).…”

Automatized spleen segmentation in non-contrast-enhanced MR volume data using subject-specific shape priors

“…However, nearly all approaches are designed and tested for CT images, which makes a reliable result comparison with our framework hardly possible. Hence, we cannot compare our results with the work of Belgherbi and Bessaid (2012) and Soroushmehr et al (2015) that are designed for CT image data. Furthermore, the semi-automatic approaches of Belgherbi and Bessaid (2012) and Soroushmehr et al (2015) require user interaction and cannot be compared to our fully automatized framework concept.…”

“…A graph cut technique is then applied to segment spleen tissue in reconstructed images. In Soroushmehr et al (2015) the authors propose a slicewise thresholding method beginning with thresholded spleen regions in a region of interest (ROI) of the starting slice. Thresholded regions in the following slices that maximally superpose with the spleen regions of the previous slices are selected as new spleen regions.…”

Engineering and clinical use of artificial intelligence (AI) with machine learning and data science advancements: radiology leading the way for future

“…Additional ML studies that were carried out identified important findings such as pneumothoraces, fractures, organ lacerations, and stroke. 25 – 31 …”

“…Additional ML studies that were carried out identified important findings such as pneumothoraces, fractures, organ lacerations, and stroke. [25][26][27][28][29][30][31] Interpretation of scans is a frequent discussion area and the topic of contention as to whether such techniques could/would eventually fully replace radiologists. The interpretation of detected findings 'requires a high level of expert knowledge, experience, clinical judgment and correlation based on each clinical scenario'.…”